Ever been involved in a website redesign? Aaron and Michael have. Quite a few, as it turns out. In today’s episode, they dig into their bag of experience to share warning signs for redesigns done wrong, advice for how to find success, and techniques to set you up for future success.
- 7 Bad Reasons to Redesign Your Website
- Accessible CSS Generated Content
- ARIA in HTML
- How to Champion a Successful Website Redesign Project
- How to Use Google Optimize to Run A/B Tests for Free
- Why Every Website Redesign Project Needs a Strategy
- Your 2019 Website Redesign Checklist
The following is a machine-generated transcript of this episode. It will contain errors until it has been reviewed and edited, and we apologize for the difficulty that may cause for screen readers. Do you want to help us speed up our transcribing process? Consider sponsoring an episode.
Hey everybody, you’re listening to the drunken UX podcast. This is episode number 42. Or we’re going to be talking about surviving your own little special corner of website redesign. Hell, I’m your host, Michael Fienen.
I’m your other host, Aaron. Now how you doing, Michael?
I’m doing all right this evening. It’s we’re feeling good. We’re feeling spicy. I’m a spicy boy.
This is the episode where we have the answer to the life universe, everything.
I were supposed to come up with something like super clever for that.
I’ve literally had 48 hours.
That is six more hours than you needed for the number so
Oh, that was all right. Very good. I was very witty, and not off the deep end yet.
Folks, if you like other witty things, be sure run by our sponsors over at New cloud.com slash drunken UX. You can check them out for all your mapping needs, interactive maps, map illustrations, printed maps, all of those kinds of things. Check them out. Let them know that we sent you over so they have fun with us.
What do you got there? Michael?
I am drinking cider boys. This is the Mad bark. I am taking a different tack tonight. So this is the apple cinnamon hard cider is kind of like almost like a an apple cobbler. Kind of no interesting. It’s it’s pretty sweet. But it’s not super sugary. Which is nice because I’ve had some of these before like or if you’ve ever had a like I like what is it? Not your dad’s root beer or something like that.
I have not had that but I am interested in
it’s good. Okay, it’s root Berry and sweet. But there’s so much sugar in it. It literally a bottle of it makes my muscles hurt. There’s so much sugar in it. What’s
the alcohol base? Is it rum?
That the generic stuff is just filler that it’s it’s just alcohol.
I I had a I had some cracking mixed with like proper root beer like not like, you know, AW, or barks or whatever. But like actual, like the legitimate root beer that has that real like a little bit of The Wintergreen flavor of Christmas. It’s pretty good. It’s a night the two flavors play nicely together.
I like me some cracking so I’ll try that up. These are nice though. They’re like say they’re easy drinking. They’re very simple. From like an ingredient standpoint. And when I because I the first time I had one, I’m like, oh god, I’m going to pay for this later. I tasted how sweet it was. And I was like, man, there’s a ton of sugar in this. And I mean there is but it’s not just like corn syrup, sugar.
Spice fish, though that that should be their label. So glad You’ll pay for this twice.
Maybe not the spirit they’re going for but
are waiting tonight. I know.
It’s late. What do you got over there? I’ve got a
after. Last episode you were talking about having basil Hayden. And I was like, oh, man, I want some of that. And then I realized I had just enough for one more class left in my bottle. So that’s what I’ve got tonight. I’ve got it. I poured it on the rocks. Because it’s very warm here.
Did you muddle have like a nice fresh basil leaf into it?
No, I I hadn’t thought to do that. Is that good? Oh, because of the name seems appropriate? It actually To me it has kind of an herbal Lee mellow to it. That would I serve actually quite well. You’re probably right. My basil plant actually just died recently. But next time I get a bottle of this, I will also get another basil plant. write that out. Where we some beta leaves, fresh ones
that let me know how it is then I won’t ruin mine. If folks if you want to know more about what we’re drinking, what we’re doing, where we are where we go. And be sure to follow us on all of the social media wherever you want to track us down. We are on Twitter and Facebook at slash drunk. And you can also hit us up on Instagram at slash drunken UX podcast. You can chat with us anywhere you want to but mostly just from Slack, that’s drunken ux.com slash slack and it will get you right into our channel. And you can give us topic suggestions, drink suggestions,
drink suggestions, especially especially I have We’ve had a few recommendations from people in the Slack channel so far. And they’ve all been awesome.
So we have been very, very susceptible suggestion on this show when it comes to like what we’re drinking. So I think that’s fair. In fact, we haven’t had a Vesper Martini in a while. So I’m have to cue that back up.
I just heard a friend of mine from Ithaca. also works in higher ed. He recommended Oh, he was drinking a Western Martini at a bar one night. But he said that you actually can’t make a real Vesper Martini anymore because the
you can’t get a little a blank.
Yes. Oh, no, no, you can’t get not will live longer, but you can’t get the original thing that they would use. It was like something with quinine. Like it had a finite flavor to it. And so like little a block is like the closest, gotcha modern.
I can’t find Lily block either. So yeah. So I want to talk first, before we jump into redesign stuff. There was a article that came across my radar, that was kind of cool. Because it gets into a couple things that we’ve talked about at length, one being CSS one being accessibility. And what this is, is over, it’s a Eric Eggert firstname.lastname@example.org. And he was writing this article on the way you can use CSS content to do design that is also accessible. Hmm, this is a weird thing. So stick with me for a second because I had to kind of lay it out. I have to paint you a word picture as it were. Okay.
So you know, how you can use the content, property and CSS to append things to elements? Yeah, so yeah, for instance, if you are making a link that goes away, you can use after content to put in like a little SVG icon that, you know, like Wikipedia does with a little right arrow, right? Or a pin something like the letters PDF to a link, if it’s a download for a P
Yeah, I remember doing that in like the mid 2000s. That was a hot hot thing back then.
And people use CSS content for a lot of stuff, especially like presentational elements, like em dashes or rose
CSS before when use on a list item when you want like a like a special graphic. That is
right. Yeah. So so we’ve got this, this property we can use for inserting content into into a layout. Now, for a long time, browsers had no clue what to do with that. That content wasn’t in the page, it wasn’t in the DOM. So browsers just ignored it, which means screen readers just skipped over it. I’m not great, especially like in the case of, you know, if you’ve got a link, and you’re attaching something, you know, like, say, external link or something next to it, that that would mean somebody using a screen reader would miss out on that, and wouldn’t know that extra context before they interacted with it.
Okay, well, you’re supposed to, you’re supposed to put it in the title.
But, of course, yes, you’re supposed to. But yeah, we know that. That doesn’t always happen, especially when you start getting into design. Right? This is where this starts to kind of take occur. People were using that to their advantage in typography. So for instance, have you seen like when people will have a sentence or a header or something like that? Yeah. And they will use before and after content in CSS to duplicate that text. And they’ll offset like up and left and down and right. And they’ll like, do like that sort of staggered font effect, you know?
Yeah, I actually, that’s interesting. I didn’t know how they did that.
Yeah. Like it or like a streaky font or something. Yeah, they’ll they’ll do that by layering after content after content after content, whatever. Cool. And it worked, because screen readers wouldn’t pay any attention to that, except now they do. Browsers started supporting that because they realized people were putting actual content in those things, because that’s what it’s there for. is they is the issue that the screen readers weren’t interpreting the CSS before. And now they are or is the issue that the screen meters haven’t changed, but the browsers are changing how they interpret it. That’s a good question that I think is maybe a distinction without a difference. Okay.
Well, it’s just the Who do you ask, like, who do you blame? But like, who do you Who would you have to go to resolve this?
I think that if I remember the way it was phrased in the article, I think it was at least framed as the browser’s began interpreting the content into the the DOM tree that then gets sent to the screen readers. When that happens. I may be wrong about that. But like I said, I, it doesn’t exactly matter, because all that matters is the outcome was that, if you did and the example they use as a header with it duplicated once before, once after, which meant that header was in there three times. And so a screen reader would hit it, and it would read it three times as a result. Yeah, yeah. Even though the first two times were totally for nothing but presentation. They weren’t there to actually be content.
This is this is weird, because I I’m not blaming the screen readers for this, the screen readers are just doing their job. This is why I asked that question earlier. I think the browsers are not handling this, right. This is done in the CSS, right? Like it’s a CSS pseudo property.
It’s Yeah, it’s done in CSS, but it is literally called content. Yeah. So I mean, it’s it is meant to display stuff. The Yeah, I think there is obviously an argument over Is that the right place to put genuine content versus like little bits of presents stuff? That’s certainly debatable. But they found a really novel workaround to this. Okay, that turns out to be super elegant. Okay. So remember, when we were talking about the future of CSS back in was that episode 39? Right, or forgotten? 4040? At 240 was a 40 4040. sounds right. So if you go back and listen to that episode, we talked a little bit about CSS functions, and how we have a lot of them already, people just don’t use them a lot. Right? There’s a function called attribute. ATR. Okay. And what it does is it will get the value of an attribute and let you do something with it. Okay. So what they the solution that they found to this was, there is an aria attribute, Aria label.
Okay, yeah. And I do know that one, you can put
a label on anything for the most part, anything at sports a title and has content inside of it, you can put an Oreo label in an Oreo label is read instead of the content that is inside of it. That’s right. It’s it is considered superior for the screen reader. So that if for some reason, you do need to give the screen reader more context, perhaps than what is, you know, maybe, like, if you’re using visual cues, for instance, you may translate those in the words for the label. But what they found was they could do their header, use the label attribute, which in this case, would just say the same thing as the content inside only once. Yeah, it wouldn’t matter, because the screen reader is going to read it the same way either way, but then they as the value for the content property, they just say, you know, attribute Are you label, and the fact that there was now before and after content, no longer matters, because our label takes over for all of rights, right? So now they took this, this thing that was broken, people were using to something that broke, you know, because it was, quote, unquote, fixed, you know, they fixed the bug, quote, unquote, to using another workaround to get the same effect. But now it’s all like, it’s more technically appropriate as a solution. Now,
the CSS function is using, it’s still using before and after, yeah, you know, but the content is being like, put in place via the attribute, Aria label, it’s,
yeah, they’re still using before and after. But it’s not read three times, because Aria label exists. Got it. And so that, that simple fact just means that the screen reader can look at it, and it sees, hey, we’ve got content with before and after. But we were also given an aria label. And we are told to trust the label above all else.
Right. So that’s what gets read. So not specifically related to that, but related to Aereo. Able, we’ve been doing an accessibility playbook at my job. And so I’ve had to like research a lot about Aria tags and things and when to use them how to use them. And so there’s, there’s already a label, and it’s already labeled by Yep. And then there’s like, just a label tag proper. And so I wrote down the specifics of their day. I’m hopefully I get these, right. So if you have an aria label, you do not use Are you able to buy also, you can’t use them together,
right? Because the label applies to the node that it is on.
Yes, yes. And so but it also if you have a formal label tag, like the HTML label element, you also wouldn’t use either of those, right? Because the label has its own associating,
right? Because then you get into the ARIA roles, right? Right, you’d have an aria role navigation, but you don’t have to say, Are your role navigation on the nav element because its intrinsic value you is navigation,
there’s so the W three Aria specification, it’s, it’s a dense read. But it’s really interesting if you can take the time and sit down and actually like, kind of pour through it. It has, it has some tables in it that will say things like, for these kinds of elements, these are the roles you can apply. And with those roles, these are the ARIA tags, excuse me, that you can use with it. My hats off to accessibility experts This is there’s like so much information to learn in here. But it’s it’s cool stuff. If you if you work with the content side of things highly recommend taking a gander to that,
what’s a good it’s a good excuse for why learning accessibility can be quite useful to making you know, making things better and making your content better in the world run. So if you want to check out run by the blog, I will have a link to it in the show notes over at drunken UX com they’ve got that example there that I was talking about that walks you through it so if our explanation and description of it was not good enough, it’s a short read. It’s actually it’s very quick to get through and it makes a lot of sense when you see it done in space, especially the example like understanding why somebody would do the thing I was describing maybe hard gear, but when you see it in practice, it makes a lot of sense that sort of 70s effect of layered bond and you know when and you’ll think about you’ll see you’ve seen it other places and you’ll know you’ve seen another play that
they use though in the state of the state of CSS it wasn’t 19
I don’t even remember
on I wish I think that’s how they did it this see on the CSS has kind of a Are you know what I think using drop shadow nevermind i lied all the tricks there they I think they use literally every CSS thing possible to do that like 90 style logo though
and we forget them
I put a link for the the Rei w three document in the show notes it’s it’s very if you’ve ever read like an RFC or anything like it’s like that level of specificity. But yeah, check it out.
can’t hurt to begin. Yeah, so this week we want to talk though at in depth about website redesigns. And this is a think of this like a fireside chat. I got I got my cider boys. Aaron’s got his basil Hayden. We’re going to sit back and dig into our wealth of experience here and talk to you about the redesigns we’ve gone through the challenges we faced and and what you should be kind of on the lookout for what warning signs there are something bad maybe about to happen. I see him counting on his fingers in the video go off way back. I remember in last season, when we talked about the first websites we built.
Yeah, I don’t believe anything professionally speaking like where I’ve been paid on a W two salary to work for someone. I have worked on a minimum of seven full site redesigns. Oh god. And four of them. Were at one employer or no, no, three of them were at one employer. Three of them were at another and then one of them was earlier one. That’s not including ones I’ve done personally for my own sites.
Okay, I’ll I’ll exclude those. I will go the same way. You did. Yeah, redesigns that I was involved with where I was paid to be involved. I think my number is eight. Yep. Now some of those were, you know, some of those were places I worked. Some of those I was hired as a consultant to drive forum. Some of those, you know, I’ve been working of a team. Some of them, I was the only person. So the some of these sites were small, some of them were big. There’s there, you know, I like the salt and pepper to my experience there because I feel like I really have kind of done everything as many wrong ways as you can possibly do it so that you know how to put it right. Next time.
Let me say, Okay, so the very first redesign I ever did was when I worked for municipality, and I redid their, okay, I both redesign the site, but I also built a CMS, this is before WordPress had pages. So WordPress was just posts. So WordPress wasn’t really an option for this,
you’re going back to like, 1.5 or so yeah,
yeah, before 2.0. But But I have to say like, so I was, I was learning PHP and my SQL, I built the entire site to be like a database driven dynamic content, I built a CMS backend for for maintaining, like articles, essentially. And my favorite part of this, and I’m really sad that they don’t have this anymore. But my favorite part was that I took the entire entire corpus of the city code, like all the laws that apply to the citizens of this city. And I put it up into a searchable like full text searchable database with like topic tags and everything. So if you wanted to know, oh, what laws apply to bicycles, or what laws apply to parking, you can go on the site and a little search box, you type, you know, you can do like a Boolean search, because that was supported with full text search just in my school then. And you can search for like bicycles, minus, I don’t know, unicycle or whatever. And it will show you all of the individual parts of the city code that apply to that. It was a really cool, so you bet in Unicode, that’s what you did.
You built the
website? Did didn’t Unicode exist, then? I don’t know. I was like, No idea. thousand four, I think, or right around there. Anyways, it was I, in hindsight, I don’t recommend building a CMS.
Step one, you know, you’re on the road to a bad redesign. Your boss says, Hey, why don’t you just build the CMS?
Well, okay, so like, it’s some irony that like, you know, I did that. But then a later job, when I worked in higher ed, we kind of built a CMS, one of my colleagues, he helped build, it was actually really cool. The entire editor was all like, like, you’re viewing the site, you click on an area of the dock of the page, and then you can edit it, if it’s an editable region, it was really cool. And our people really liked it. And if you’re isn’t it not being used as because marketing was decided to go a different direction? or whatever, I don’t know, there’s some kind of like, higher up political reasons. That didn’t happen. But I mean, there’s sort of building a CMS, but it was like, way better than the one that I built. So I think that it’s okay.
So it’s, that’s a good chance to kind of start with I think, maybe the first step, which is, how, what are the bad reasons to jump into a redesign? And I think that’s, you know, let’s start with the bad so we can kind of work towards the good. And end on a high note.
I think a bad reason would be because I want to learn PHP.
I mean, that wasn’t exactly my reason for doing the one. But if it’s not,
yeah, I’m not saying I’ve never done a project where I’ve been like, I’d love to learn that. Let’s build something in it. But I’ve done that. One of the big ones. And I’ve heard this myself, and I don’t love it is we don’t like it anymore. We don’t like our side anymore, or somebody doesn’t like our side anymore.
It’s somebody not like a random public person. But like, somebody with decision making influence, right?
Yeah, the idea of somebody not liking something, that’s a bad reason to commit to a redesign, especially one person or one great, you know, depending on your site, and some of this, and I’ve been a warn folks now, like some of what we’re going to say tonight, is going to fluctuate and and I’ll try to, I will try to clarify what I’m saying something about, like a big site versus a small site. So for instance, if it’s your personal site, if it’s your blog, or your you know, any different, you don’t like it anymore, that yeah, that is very different from being at a company with a site that serves, you know, 10s of thousands of page views a day and has hundreds of thousands of users. And you know, as a marketing board that comes in as they don’t like it, that’s a very different thing.
If it’s your site representing you, or something that you maintain, and you’re the person who is responsible for doing the labor to make any changes to it, then you can pick any reason you want for doing a redesign, because you won’t pick a stupid one, because you’re going to have to be the one to do the work. Yeah.
Accountability helps, I’ll tell you that.
I would file I had this as a separate bullet item, but I think it falls under this one, the make it pop thing. I anyone, and I’m sure some of you out, there might be groaning right now, anyone who’s ever worked in web has heard that phrase, like needed to make it pop. And I don’t know, who first said that. But
that goes all the way back to the print design days. Clearly, that goes that goes way back to all kinds of marketing when everybody needed to stand out. Everybody was trying to, you know, get that full page ad and, you know, whatever comic book was coming out next,
right. I I read comic, I think I think the reason that make it pop is such a frustrating phrase is because the people who are saying or asking for it to be made to pop, generally don’t know specifically what that means. Or they don’t know what it requires to achieve. Popping, but
it doesn’t mean anything. It’s an empty state. And and that’s the same way with I don’t like it anymore. I don’t care if you don’t like it anymore. liking it isn’t a measurable quantity. And nor do I know what to do to make you like it now. Right? Right. If you say, hey, when you come to our site, you can’t even tell what our brand is. You know, you can’t, Jr, you know, you don’t? Yeah, our identity is hidden, because all we use are blacks and whites and grays. Sure, that I can work with. And that becomes justification to actually start redesigning something, because now I have an actionable thing. I understand the complaint. Yeah. And that can be hard.
And I would say also, if you’ve done, you know, user surveys, and, you know, people have explicitly said that the site feels boring, or uninteresting, or they just have trouble finding any content, because they just, it just all bleeds together. You know, if that’s actual feedback you’re getting from real users who have no dog in the fight, so to speak, then I mean, that’s legitimate and something to consider.
But then that’s grounds for user testing. And then when they say that, you can ask them, What do you mean by x? Right? Because it’s still even then like, while it’s still indicative of a problem, you don’t know what the problem is, with those kinds of amorphous phrases. So anytime you hear those, those should be signals to say, well, let’s stop and identify what is the actual need? What is the actual issue?
I think that goes into another one that I think you and I have both heard a bunch, which is the like, it’s time. Yeah, like, as if there’s like an expiration date when you do a site. But I mean, I’ve definitely heard that before, where it’s, typically upper management will make these requests. And we’ll just say, like, you know, we haven’t redesigned it in a while it’s time to do it. I don’t think it’s wrong. But I think that it’s incomplete. The mistake isn’t saying that it’s time the mistake is that you’re not saying what the like, what’s the underlying problem? Like, you know, okay, like, it looks stale, like, it’s time, as it’s like, is a 14 months and magical number? Or is it that like, you know, the the industry, our industry, our competitors, are all kind of looking like more temporary, and we look kind of like five years ago, which could be exactly what it is. But like, you need to have more specificity. And just saying alone, like, it’s just, it’s time for redesign isn’t enough? Because then you don’t know like, what the criteria are for success? How do you know when it is when the time is up?
Your website is not toothbrush?
or oil in your car?
Yeah, it’s Yeah, it’s not the kind of thing that yeah, it’s been six months, it’s been 3000 miles time to go.
Like that we we say this, but later on in the show, we’re going to talk about how content and code both rot. So like,
if there is no clock, that’s the important part. Yeah. And with that, to what you were just saying to this idea of, like, you know, what does your business cycle on? Do your customers start to look around or whatever? Those are good questions. But the general idea of just, well, we just need to be competitive. Our websites not competitive anymore. Okay. Why? What does that mean? Why is it not competitive? What is it not doing? And then when you tell me what it’s not doing, then we can look at dressing that than that.
So that could it could mean, our sales are down, like we’re converting less, we’re getting fewer visitors, which is resulting in fewer sales. I mean, like, if you have a measurable conversion that you can track, and you’re seeing less of that, that that’s totally legitimate reason,
and you’ll find it rare that the reason like a conversion drops isn’t because your design stayed the same. That’s not how variables work. You know, if your design can get old and stale, that is absolutely true. But the idea that something that was generally effective, stops being effective is indicative of a different kind of problem.
Okay, so you can have to do a shot now zone I mentioned Steve Krug, but
damn it. I don’t have shots. I’m drinking beer tonight that must have to chug the rest of this. You realize how many bottles of this I’m going to go through? This has to do a step
in Don’t make me think Steve Krug talks, he uses Amazon heavily as an example of like, really solid usability. And and I’m going to use it here as kind of a contra example. Because I think that the period when Krug wrote that book, Amazon site was actually like, very usable, it was in a very good place. And I don’t think the current iteration of the site is I think it’s a hot mess. And I think that that’s probably the redesign probably came around because they were trying to be more competitive, or because it was time or because they wanted to, I don’t even know, sell more stuff or something. But I just I think that’s Yeah,
Amazon’s in a weird case, too. And I that may even be worthy of its own episode at some point, because I’m sitting here thinking about it. And like, you know it for as much as you you will pull examples from a place like Amazon, and you know, that will talk here a little bit about the follow the leader mentality. The reality to me is that Amazon has up there above the rules, basically, like, yeah, and Google is another one, like some of these companies, they actually get to violate what we otherwise consider to be sort of the laws of UX.
Certainly an outlier. Yeah. I mean, like you look at like, when you talk about like a Site Reliability Engineering, a site like Twitter is there, they’re in their own category, because they have so much like traffic velocity, that it’s just, your site will never have the level of velocity Twitter has
ever Jenny generating more leads, though, is that that is a tactical chain. Yeah, that is a singular thing that you attack that you don’t resolve by redesigning your site, you deal with it by re architect in your lead generation process. And that may mean business logic changes as much as website changes.
I think related to that one, too, is also when people look at things like page hits. And then they’re equating that with overall site performance. And I’m not saying that the number of page heads doesn’t bear some utility. It’s useful when you look at like the overall, when you look at your year over year cadence, you know, these days a week, we have more users these months of the year, we have more or less users, that’s useful. It’s good to know that. But the number of users doesn’t necessarily equate sales specifically, I guess, unless you’re making revenue from AD displays. that would that would be that would be different. But if you’re if you’re doing student like recruitment, or sales or whatever else, like you would actually look at your actual conversions, rather than just purely like a single, indirect stat like site hits
yet. And anytime somebody brings up something like we need more traffic, that should be an immediate red flag under any kind of web project, because single dimension metrics are garbage, and they’re useless. If somebody says, We need more traffic, or we need to decrease bounce rate, or we need to improve the time on site, hmm, no, you don’t have to do any of those things.
The time on site always boggles me, because do do you really expect users to just like, you know, sign on to the site just hanging out? Yeah, like, Hey, guys, like, what’s up? Let’s hang out here. And
single dimensionality is used a lot, because it’s so simple to convey. But it is ultimately a throwaway statistic. And the only way to make those those metrics valuable is to give them a second dimension, generally, compared to some sort of conversion, or some kind of goal that the website has. Because what you want to know is, what is our ratio of traffic to conversions, you know, are the people who spend more time on the site more likely to buy, you may discover that the people who come to your site and spend 30 seconds there are the most likely to buy your thing on impulse. And then if they spend eight minutes on your page, they never want to convert. So right higher time on site in that situation, turns out to be a tear, reason to do a redesign, because it’s counterintuitive to what you actually should be doing. Right? There’s an even then like some of that stuff, then you have asked that question of why why do you know people on this site shorter times by more,
when you’re, when you’re faced with any of these as a reason to do a website, you should do the five wise know, get really get down to kind of the nut of the issue? Because a lot of the reasons that we kind of get superficially thrown out there. They’re not elucidated the problem enough. So you have to kind of question your way into it.
You’re still using big words, you need to drink more of your bourbon. Don’t lose the dating, kind of fancy kind of person, are you Oh, it just
fell out of my mouth.
That’s like 18 syllables, man. Well, if you’ll allow me, I want to go on a rant. This, partly goes back to it. Talk, I get my god eight years ago. And I’m going to apologize in advance because I’m going to quote numbers from this talk that I did their old numbers. So take them with a grain of salt, I did kind of try to scrounge around a little bit, see if I could find newer ones. But I was unsuccessful. However, I think, you know, when I went back and was thinking about it, I’m not sure that there’s a good reason why in this time frame, these numbers would have changed a ton. So I’m going to share them and you couldn’t agree or disagree with me. And if I find better data, I’ll be happy to share it. This premise is based on the idea that full redesigns are just wildly inefficient. They are a terrible way to go about improving your website. And that’s true. In the best of circumstances, it becomes even worse if you’re starting from one of the bad premises that we’ve already talked about. So if you’re already coming in with one of these, like weird, vague assessments, why you need a site redesign, this makes it that much worse. So this data goes back. It’s a study that was done by HubSpot back in 2011. And what they found, they talked to a bunch of marketers about website redesigns. And what they found was a third of marketers were not happy with the last redesign, they were part of the 33% ish, only half, or just under half actually finished and launched on time. Because that never happens. I mean, it’s it really is a flip of a coin, whether or not you’re actually going to finish this thing when you’re supposed to only 24% would have qualified themselves as extremely happy. So less than a quarter of them feel like they came in and got what they really wanted out of that deal. And the the stat that I always love is 68% of marketers had been in a redesign in a year or less, even though about a third of them say you should only do it every two to three years. So there’s a very big gap in the practice what you preach section of the choir for that. I mean, you’re looking at almost three quarters of the marketers surveyed had redesign a site in less than a year. Unknown Speaker Mm hmm.
That that is demonstrating a frightening cycle of churn. Because a lot of marketers, especially new ones, when they come into a company, they tend to feel that need to like make their mark, right? You gotta Yeah, you got to put your stamp on things and and rebrand or re message and do all this. And a lot of them see Oh, he designed this way to do that.
That should all stuff should be added to the bad reasons for reading. Yeah, yeah, because I’m a new hire,
because I’m a new hire. But that happens in a lot of places. And not just in this industry, like a lot of industries have that that sort of problem of somebody new comes in with power. And they have to kind of establish themselves. So they’re looking for that, that flag planting moment. Unfortunately, for marketing, they’re looking for very forward looking visible thing for that. And the website makes for an easy target and a lot of cases. Yeah, what they found was in the redesigns that were done, they spent roughly between $54,000 and $69,000 to complete that redesign. And it generally took on average about 5.1 months to do that whole project.
Now, it’s about 11 11,000 a month.
Yeah. So to put that into context, and this is just one of those kind of a maybe don’t redesign, you’re basically saying you could go out and hire a full time web developer for about $127,000 a year or two really great web developers for a year or three pretty okay, web developers for a year. You can empower yourself to do a lot with that money in that time. If that’s what you’re after.
I think just as the devil’s advocate here for business, I budgetary issue would be that the spending a one time cost on a contract like that would be allocated differently than allocating payroll money. Because there’s additional, like, overhead and everything that comes in with that. However, you could hire them as contract employees and see how that works. Or, you know, if you’re doing an honor, if you’re doing redesigns on a regular basis, you really should have more web people in the end. Because you absolutely can do them in house.
Yeah. So the the magic word that you said was one time? Mm hmm. And the reality is that most places don’t go through this problem one time.
Yeah. Speaking of experience,
they do not they go through frequent and it becomes very quickly apparent that Yeah, okay, yeah, maybe you’re not spending all that money in total every year. But I think you can definitely make the case for the value of a redesign, not generally being worth it. Yeah, and let’s say I’m, I’m passionate about this idea. I think redesigns generally aren’t good use of time, money resources, especially when money is being spent. If the differences between sending that money away, or bringing people into the fold that can help you, you know, long term, investing in yourself is always the better, better value on that. And keeping in mind that when you do a full website redesign, it’s never just design, never just design, it’s always including supplemental stuff, whether that means you’re doing content writing, you’re changing your IA, you’re doing the new CMS implementation, you’re swapping out other hardware, other supplemental technologies, like you your email marketing system, all of these things at scale, and increase the efficiency of the project. Because what happens is, everybody goes, Well, we’re throwing the baby out with the bathwater, till we might as well do these other things too, even though those other things aren’t always immediately related to the value of the project.
I will also say that, if you are a middle, like a middle sized organization, and so I’m saying like a large sized organization would be something like a really, really big company like Google or Amazon, middle sized would be like most higher ed institutions like that size, or smaller. If you’re doing a redesign, you can separate the redesign itself from the execution of the redesign, provided that the designers, you go to our web savvy, if they give you you know, the HTML and CSS layouts with components and everything else, you can hand them off to your on staff, web people, and then they can like execute and implement it in place, or in your staging environment, or whatever. And that will save you a lot over time. Because the whatever you’re paying your web people, it’s probably less than what you’d be paying these contractor or the firm to do it for you. And a lot of that is really just like, it just you have to throw time at it. You know, applying a new template, restructuring content, moving files around just takes time. And it’s not difficult. The difficult stuff is in making the design itself.
It’s I mean, and this all goes back to why I’m so like, opposed to this idea of the redesign, because no matter how you slice it, it’s an expensive process. Yeah, it’s you’re either sending money away, or you’re spending your own, you know, salary basis on it. But no matter how you cut it, it’s a huge sort of almost like a black hole, because you generally will get something out of it in the end, but it just has such a huge margin of error, go back to that stat 49% finished and launch on time.
You know, that’s, that’s so hard, like, how do you how do you know when, like, how can you possibly set a target date for when a work that includes unknown duration, like pieces, I just that every redesign I’ve worked on has had like a deadline. But none of them have ever made the deadline.
The deadlines are arbitrary, right? They’re all just kind of picked out of thin air and they’re set at the very onset of a project rather than waiting, you know, and and doing a way through. They, they tend to be politically arbitrary.
They’re they’re set. And this is something that might was my experience from higher ed, at least, they tend to be set by people who are like, Oh, well, obviously, we want the website done before students return right in the fall. And it’s like, Okay, well, we understand that, let’s set a target two months before that maybe like right after commencement, because it will probably take longer.
So if if you want to do this and do it right, you know, if you’re in a in a bind, and you know, redesign is coming, and you want to at least try to set yourself up for success. There are some things you can do, though, and these are the How to weather the storm kind of points. First and foremost. And one of the most important things to any project, let alone a redesign is make sure you know what your key performance indicators are before you start. That’s when if you’ve ever heard somebody throw it around the phrase KPI. KPI is the thing that matters. What is the thing you will measure that will tell you what is working and what isn’t? The reason you need to know it before you redesign is because you need it baseline. Yeah, you can’t redesign and then say, hey, how is this thing doing if that thing didn’t even exist beforehand.
It’s it’s how you, it’s how you protect yourself as as the team doing the work. For afterwards, you can say, based on these KPIs that we were in this data were given beforehand. This is how we’re performing now. And so the redesign was this much successful.
If the number of leads you get in the day is considered a KPI you know that? That goes back to this this argument, right? Well, we need to generate more leads. Okay, well, that’s a goal. And it’s a goal we can measure. That is one KPI among many that we would consider in a redesign. So we know what is our number? Is it five? Is it 50? Whatever it is, how much do you want to increase that? What do you think the capacity is can be increased by 20% 8%? Let’s set a goal. And then I do identify why the changes you’re going to make in the redesign should affect that. If they say we were getting 10 a day, we think we can get 12 a day. And we think these changes will do it. That’s a very well defined kind of strategy, then because now you can go back, you can do that, measure it for six months, and then see did our average go up to 12? Did it go down to eight, then measure that and figure it out.
And really in this kind of gets back to your original point of you don’t need to do a redesign. If you can identify things with that level of specificity, then you can probably just do an A B test with what you already have, right? And just make that one change. And then actually see how much you lifted we got when we did this, when we put the make the button blue and put it on the right side of the forum. You know, how much what’s the percentage of increase and conversions? You don’t need to do a full site redesign for that. You could do that in a weekend. Yeah. Or Excuse me, can do that in a week? Yeah, we’re not going to work on the weekend.
I say that that. That’s the the other side of that argument, though. The KPI thing is to also make sure you know what is working. Yeah, because you don’t want to throw out the stuff that’s working that already. Because there’s no possible way at all that your website is a complete and utter failure across the board. Maybe if you’re coming into a site that was designed 15 years ago, and was never updated. Sure. But generally speaking, if you’ve got a site that is relatively modern, in any aspect that has had any attention paid to it, and you’re at a company that makes money, the site is doing something well,
fair. And if you don’t know what those things are, you risk doing more damage than less, you’re going to change things that you shouldn’t have changed, you know, maybe the reality is you’re getting a great number of leads. And one of the goals of the redesign is that they want to change the branding of the company. And this is one of those reasons that you’ll hear from time to time, the branding of the company is changing. So the site has to change, new color palette, new logos, we’re trying to we’re trying to appeal to 17 to 34 year old chimps, whatever that list looks like. And so if you’re making that change, but you know, your lead generation system is rock solid, you want to be careful about how you change stuff around it. Because while they want to increase their appeal to that demographic, you don’t want to throw out the old demographic necessarily.
Yeah, if you and this and this gets back to the points you made earlier about asking why. If you know, if you can get that problem set really specified, then you can be sure that, you know, you may maybe you think so you’re new persons coming in, and they’re saying like, we need to increase our patriots. Okay, and like, let’s say that you didn’t ask why. And you’re just like, Okay, and so you do all the right things to increase your paychecks. But in the process of doing the redesign, you eliminated parts of the page that were actually crucial to converting into sales and your site. Then now, like your sales are going down, but your paychecks are going up, right. And so that’s why it’s really important to make sure like you were saying don’t do single dimension metrics. Find out how these things Correct, correct. Connect, find out what you’re doing well, and what you want to change. Want to do better with? I guess, in short, like, don’t, don’t treat the problem superficially. Do your homework, ask the questions.
Yeah. And the, this KPI kind of argument also goes all the way back to the very first thing we said, you know, this idea that, you know, somebody came in and said, well, they don’t like it. Yeah, not liking it isn’t a KPI. And if somebody comes in and says, Well, I don’t like it, I don’t care. Because generally speaking, the people, you know, if it’s your management, or another group within your company, or the sales division, or whoever, if they’re coming to you and saying, well, we don’t like this, you know, it doesn’t work for us, we’re going to read, we want to redesign the site, because all the stuff is bad. It takes a firm hand to sit them down and say, You know what, but you’re not the user. I need a, I need a KPI to anchor this too. And we’ll work on that. But if you aren’t the user, you’re not going to affect that.
I’m not saying I endorse this strategy, however, at a previous employer, a former coworker, and I’m not saying who was involved in this but a former coworker. The boss had said, basically, they don’t like this one thing, even though it was, it was the right thing to do. But the boss just didn’t didn’t like it because I didn’t approve their sensibilities, whatever. So what my former coworker did was, the site was part of an application where you would have been signed in already with a single sign on. And so they made it so that if you were signed on, and your username was this one username, which lets the boxes, then the site would render this section of the of the page one particular way. And for everyone else that would render it the normal way. It actually worked. I’m not
entirely sure how to take that. I don’t know if that’s a genius, or that should get you fired.
If the whole reason that I even know about this was because I was looking through the code. And I was like, why is so and so’s username in the code here? And then they were like, oh, OK, so
I mean, it does, it goes to the superficial reality of some of those requests. And yeah, and how important or not as the case may be to the business value of those decisions. KPIs are all about business value, it’s all about identifying the things that are worth doing, why they’re worth doing it, and how you make sure that you’re getting that out of it. And you can’t do that on a whim. You can’t do it on emotion. Yeah, you know, this idea of somebody who just instinctual II can sniff the skies and know a tornado is coming. That’s just not how it works. in this field, it’s there is you know, I say there’s a lot of art to good design all of this. And that’s true. But it’s also a shit ton of science, that tells us the way humans behave and and the way we can measure things and the way we can figure stuff out,
put it this way, you’re effectively making, let’s say, $100,000 wager. And so maybe you can’t identify a sure thing. But don’t you want to know as much as you possibly can to make the right decision before you drop? 100? k? Yeah. Like, if, if you’re working from subjective things like, Oh, well just don’t like it doesn’t pop enough, or whatever, you know, and that’s, that’s the level of detail you have, like, do you really want to be throwing $100,000 I mean, even if it’s them paying for it, like, really.
So the next part of this is to make sure that you’ve got a redesign coming, make sure because you know, the word redesign has that key word in it design. And like we said, The redesigns rarely our scope to that small, what you’ll end up realizing is that you need to make sure that somebody is taking the time to align strategy, get your technology stacks in order, get your designers on the same page, and get content people involved, because generally speaking, all of those things are very likely going to get touched at some point. Get stakeholders involved early. Yes, yo, yes. Anybody you can talk to that is going to have an opinion. They need to have their moment in the sun.
Yeah, and it should be early before you make. Before you move a lot of mountains. I more than one of the redesigns I worked on there was stakeholders and and really like I mean, they were shown the mocks beforehand. I don’t know if they just ignore them or Yes, like division or what. But they later on, they came back and said, Oh, we don’t like this color. It’s like it was Mark, what do you mean, you don’t like it. So it’s like, you know, five hours of my day,
you have to be prepared for some of that too, though, because the reality is, no matter how many times I say that in the course of an episode, the reality is, the reality is that people do ignore the cops, they do ignore the mockups, they don’t pay attention to the way content is laying out, they don’t pay attention to the warm up some or even the real content. In some cases, like you have to be prepared for that. And because most of the time, those folks are sitting in a meeting that was scheduled for them or showed up on their calendar, so they’re just there. And there, that’s where that’s where the PM comes in, you gotta have either you have to have a formal, who’s overseeing the project, or you have to have someone on that side of the table, who will advocate or follow up with all the stakeholders and stay on them, and make sure that they’re kept up to speed on things. Because really, I’m serious like this, this will bite you in the ass every time. wrangle your stakeholders are which PM are you referring to a project manager, I didn’t know if you were going towards product or project. Because for some places, it is the product managers job to deal with that.
It could see either one, I’m in Project Manager,
if you outsource your your project manager, maybe the guy at the other company who doesn’t have a lot of direct contact,
it needs to be someone who can contact your stakeholders.
That’s that’s the thing is just making sure the right people are empowered there. And it’s also important because you a you need that person to align all of these different silos. So they’re on the same drumbeat. Because many times they’re all on their own cycle, your IT department, if you show up a week before, you’re ready to start putting stuff in and say, Hey, we’re launching over they will see a mass, you have a laugh it Yeah. You don’t surprise them with requests like that, because they may be developing 17 other things right now, all of which have at least the appearance of more importance because of the CEO wanting them or they’re on a maintenance cycle that has to be met, or they’ve got a license that they have to deal with.
They might be doing a server migration, they might be doing upgrades, they might be doing infrastructure improvements, they might have big projects that they’re doing on their own. And this just doesn’t fit in with that, or it’s possible. And this is something that I’ve definitely seen, where the people who approach it will say, Oh, we just need this to be done. But it’s like, this thing that they want to be done is perhaps impossible, or huge, it is huge. Or sometimes it requires some significant infrastructural changes.
Well imagine showing up and saying, hey, by the way, we’ve got this brand new CMS, don’t worry, you don’t have to do anything. Oh, my vendor hosts it. All we need is you just need to integrate Shibboleth with it.
I’ve seen that too.
And you’re on an old app campus.
I think anyone who says that is no big deal, the vendors handling it, I always want to say like okay, well, then you can handle the support tickets to come in.
But it’s those kind of, you know, if you’re not involving these people, and you’re not making sure they’re all on the same page with you, your vendor is going to tell you the whole time. And just to explain Shibboleth is a single sign on protocol. So large organizations will use it to make it so that you use the same login for all of the different things fairly widely used fairly widely known. If a vendor says, Oh, yeah, we can absolutely support your existing logins, you’ll just have your IT staff, you know, set the Shibboleth connector, and you’ll be good to go. And your marketing team or whoever is in charge of the purchases great. And then you show up and nobody stopped to think maybe we should have made sure that is the one we use. Or maybe it’s just a straight up matter of Yeah, we do have Shibboleth but we don’t just connect it to random third parties. You know, there may be secure concerns that have to come into play and all this and they may say, yeah, we can do it. But you’re going to have to undergo a security audit, you have to have you know this, you’re gonna have to have this kind of test, you know, we’re going to bring in some folks and they’re going to do a pen tests are going to do these things we have to have, you know, such and such a verification from them. And they’re telling you that’s going to take six months just by itself to get through and you’re supposed to launch in two weeks. Because you were told that oh, well, you’ll just have them turn on Shibboleth and it’ll it takes them an hour. Well, yeah, it does. If there’s nothing in the way, and sometimes organizations have stuff in the way,
don’t ever let a vendor tell you how easy something will be to do. They should not without your own IT people being in the room every it. I don’t care what where you work. Your situation is a special, unique snowflake and has its own weird quirks and everything that have developed through the process of your institution or business, you know, coming to life there all different in different ways.
I would worry less about you know, the listeners being the ones who would fall prey to that. And they would be on the receiving end of the folks who did. Yeah. And that’s where it’s like, you just have to be ready to ask those questions. And be ready to at the outset. Say, whatever timeline you’re coming to me with, if this is the first time hearing about it, we need to adjust expectations. And make sure we sit down and go over this. And here’s why. Most, in most cases, explaining why something has to be slowed down or won’t make a deadline is all that is required to make that be true. And we sometimes feel like we don’t want to have that conversation or it’s not going to be well received or anything like that. But lying and saying oh yeah, we’ll get it and then miss a deadline or whatever. That’s way worse. Oh, we don’t do that. And I’m a firm believer in also not necessarily taking the blame other people’s, you know, what is the, your lack of planning does not constitute an emergency on life or emergency. Yeah, I’m a firm believer in that. And I will do everything I can to help you be successful. But I’m also not going to be unrealistic. And if I think something that you think is going to take a day, but I think it’s going to take a month, I’m not going to compromise and say it’s going to take a week. But I’ll tell you, it’s gonna take a month and next time, we’re going to have to sit down and talk about this. Here’s why I can’t do it faster. And here’s how you can help me do it faster next time.
If if you have a situation where you’re getting the surprise web work with short deadlines, and not every place does this, I have worked for some that do but I did work for one it was in higher ed long time ago. And I spoke to I think it was the vice chancellor or it was it was some one of the higher ups and I just sent an email said, hey, look, the moment anyone says the word way site, you must immediately get someone from my team in the room, they need to just be in the room just listening. And just like so they’re aware of what’s going on. This needs to happen. Because what was happening was that they would go through months of meetings about this web stuff, and it’d be stuff that wasn’t even possible. And then we they would come to us and say like, Hey, we got, you know, you got two months to implement this thing. And then it’s like, we can’t even do this, you should have been bringing us in six months.
And the challenge to That is right. To the stakeholders viewpoint. That is now a project that has taken eight months. But they don’t understand you know, the from a lot of the stakeholders standpoints like they don’t care. That’s and they don’t, they shouldn’t have to, quite frankly, and in many cases, but that’s sort of that outward vision. And that’s where like having that PM involved, we do the same thing for what it’s worth, like it’s there. I’ve sat through a lot of meetings that are necessarily a valuable use of my time, like in that moment. But it’s always good to have that ear. Because it’s gonna pick up time when I know, oh, hey, I remember them talking about this, I can let other folks know, hey, here’s what I heard. You know, this is probably gonna be coming in the next couple of weeks. Let’s kind of make sure we ask those some questions leading up to that. And those kinds of opportunities come up all the time. And it just requires somebody to make sure they’ve got their Google Hangout left open, and they’re listening while they’re writing some code or something.
To go back to your comments earlier about the Eldad versus Shibboleth thing, if you were in that meeting, you would have heard them say like, Oh, yeah, like the the vendor says that we can integrate this Shibboleth, that should be super easy, then you can immediately just, you know, raise your hand, say like, a KK, so we don’t use show FVSLDF. So that is going to be a blocker here. This is a things that would have to happen first, for this to happen. And you can nail that right in the writing.
And maybe it’s as simple as them shipping you a different connector, you know, it can be a very simple fix at that point.
But just having that presence in the room, even if you do nothing other than like, just listen and take notes. It’s just for those one single moments, you can stop bad ideas from germinating.
So the other thing you can do to always help a website redesign and I’m a big fan of this. If you go back to and listen to our 10 commandments of UX, it’s one of the commandments of UX that was listed, find reference models, steal stuff, look for ideas that match what people want. Especially if somebody says, Well, I don’t like something. Okay, well help me understand what you do, like, give me you know, three websites that do the thing that you want us to do and how they do it. So I can look at what they’re doing as a reference model and pick out exactly what that common thread is. So yeah, this works really well for users that aren’t good at articulating, you know, UI lingo and, or like experience type stuff by them saying, Okay, yeah, if you go to, you know, such and such calm and, and this place.org, and all this and look at the way they do it. If you’re, you know, experienced at web development, you built stuff you can, you can look at those things and say, Oh, I see what he’s doing, that getting reference models for that stuff is incredibly useful. Because it just helps center ideas and lets you attach, because the reality is, and, again, lots of UX. Most people are doing stuff on other sites, not yours. We’ve said this before, and it’s worth reiterating, make sure you are communicating all the time. Like, literally like I’m in the middle of a project right now, it’s not a redesign. It’s a redesign of tool, but not have a site. But as we are doing this redesign of the tool, we have daily meetings with the stakeholder five minutes. It’s not an hour, it’s not a half hour, five minutes, we pull them into a meeting, just real quick to say, here’s what we’ve done today. Here’s what we can show you right now and get feedback on. does everything look like it should? Great, we’ll see you tomorrow,
don’t when you’re communicating with your stakeholders, and management and leadership don’t fall into the trap of project making linear projections. My experience with three designs is that the power law and Pareto principles both apply here, you’re going to have, we had a see one of the redesigns we did, I think we had, I want to say there was 10,000 pages, something like that. But I would say easily 80% of them, I finished in one, it was just a matter of just changing the template that the pages used. And the remaining 20% took five months. If you’re telling them, like oh, we have this many pages done, and this is our progress, it’s not going to be like a linear crunch of time, they’re going to be like a very small number of areas that are going to require a whole lot of work. Because the new HTML template you’re doing, either it can’t be done with a find replace, or a reg ex replace or anything like that, or it just requires a lot of reworking or something, there’s always going to be something
I like the idea of the five minute meetings, in part because it it gives you the opportunity to manage those expectations and say, yeah, we made progress on another thousand pages today. Just remember, once we get through the next 5000, it’s going to slow down, because now we get to the heart, like yeah, you can reinforce that message frequently. And yes, with only five or 10 minutes to sit down and just go over what’s been done for the day, there’s not a lot of time to get caught in the weeds as a result, either.
I would even go so far as do like, like, if you’re going to do if you want to do any kind of chart or like progress graph at all, like do we’ll do a logarithmic one, you know, like the, you know, the the first the first block is all the easy files, and that takes, you know, the smallest chunk, and then the next order of magnitude is like the next block and so on. really communicate that like the density of challenge is going to escalate as you go on. Because otherwise it’s gonna, you’re going to get like 95%. And it’s like it’s been three weeks, you know, at 96% is like,
well be clear about the things that you don’t know to like, you will always run into certain amounts of work that are either going to require you to learn a new tool or require a change in implementation to something you’re not used to. And those unknown unknowns, or, you know, sometimes the known unknowns can still have a clock associated with them that isn’t fully registered yet. We aren’t omniscient on some of that.
Part of that means taking a really good inventory of what you have. That could be as simple as you know, writing out making compiling a full list of every file you have that has to be changed. That’s that’s the typical one I would start with I think.
inventories are tough though, because they’re so big, whether it’s you know, file inventories, content inventories, you know, if you’ve got 100,000 pages on your website, which may sound like a lot to a lot of folks, but isn’t for many sites, that’s very hard to take an audit of. And even you know, when you start getting into hundreds of pages, if they are legitimate pages, you know, and they have to be there for a number of reasons. you’re logging metadata, you’re logging page titles, descriptions, meta images, all this kind of stuff, there gets to be point where you’re working at a scale where you cease to be as productive. And that that’s where that challenges. Again, going back to this idea of aligning your content strategy. Make sure that’s all part of this, because in theory, if you if your content is going to be involved in a redesign, then you need content people also dedicated who will be doing the work to, you know, maybe you have to change the way certain phrases are used, we’ve had that happen where like, a product was rebranded. And so as a consequence, every usage of the old phrase had to be swapped out for the new one. And there are things like that, that content people for the exact exact same way as it folks, they need time and lead up to be able to prepare that if they have to scan all of these pages for those references. They need to know that
if you have application code, like whether it’s in line form processing in PHP, or if you have Rails apps that use your site template, or if you have, you know, cold fusion code, or I don’t know how far back you want to get. I mean, if you have stuff that’s not just basic HTML, that has like, you know, some dynamism to it, that’s kind of, I guess, take a note of how many of those kinds of pages you have, because those are going to be some potential hazards. Yeah. Or time sucks.
It’s also a great opportunity, though, to tackle a lot of things. So for instance, you know, if you’re doing a redesign, it’s the perfect opportunity to take that time and make sure you’re including good accessibility techniques into the mix. Oh, yeah, that’s, you know, one of the big things that always comes out of these conversations is they want to redesign because they need to improve their SEO. Yeah, I’m gonna let that slide. That’s a, that’s a relatively okay. Tactical goal that you want to drive more organic traffic, everybody wants more organic traffic, but by using realizing good accessibility patterns, what you’re going to do is drive up your SEO, because you’re ensuring your content reads well to anything. And that’s the best possible opportunity to do that. And also make sure that you’re not going backwards on it.
Right? Plus, it’s, you know, it’s implementing better accessibility 11 why stuff five or eight stuff, you know, you’re going to have to win. If you wanted to do that. Anyways, you need to touch every file in the site. Anyways, and so the overhead is already there, you’re already touching every file on the site. So adding a tiny little bit of doing like an accessibility audit, and identifying so low hanging fruit.
You know, it’s that’s like easy work, if you’re going to do it all the do it all.
Yeah, why not? I mean, you’re really not going to add that much time on to the thing at the end.
One thing that my current employer does that I think is pretty awesome with that with incremental redesign, is we run a lot of experiments like a beach house, where we’ll, we’ll try, you know, on a form will try like, what happens if we use like this plugin, when they’re filling out this data like this, this results in more form conversions. And then we can look at the data, we can run it on a subset of our users and look at two weeks of data on it and see, does it make a difference or not? And then we can make decisions and kind of improve things based on the app. And I think that’s really cool.
I’ll throw a link in the show notes to an article from monster insights that goes over how to get started using Google Optimize turn a B tests, because the AB test is the Keystone to doing good incremental changes.
Yeah, it really is.
So here’s the thing, right? This idea that if you don’t redesign, if you don’t do anything to a site, it spoils. And we’ve used phrases before to describe this, you know, phrases like content rot code rot, you put a bowl of fruit on the table, it just it goes bad over time, and websites do the same thing. The designs get dated the techniques, you know, think about the way we used to use drop shadows and gradients and color palettes, the way graphics were approached, all of these things change over time. And you go back and look at an old site that hasn’t been maintained well, and you can see this kind of just break down of it. This incremental approach, what that’s because that’s the thing, that’s what triggers somebody to say, well, we got to throw the baby out with the bathwater. They look at the site. And that it is it has gotten so bad that you have no choice but to redesign. And the sad thing that is not acknowledged is we failed to produce at plan out of our last redesign. That kept us in motion,
you should never be filling up a bowl of fruit and thinking, this is the only fruit I’m ever going to need. Yeah, like, you know, like you take you, you eat the apple, you replace the apple later, like you do parts of it. Over time, I think it’s a T shirt idea.
The reality is where full redesigns are wildly inefficient. incremental redesigns are probably the most efficient way you can possibly approach design,
most cost effective, certainly, it’s tough, because
from month one to month three, your site’s not going to really look any different. From month one the month 12, it probably will. But it’s that slow burn kind of change. And it doesn’t feel super effective, you know, to marketing or to your, you know, management team or whomever. But it’s designed to do that it’s designed to be able to go tweak things, adjust, measure, run your AV test, pick something that you think isn’t working, well figure out what the best way is to do it, and then make that change that one individual change. Because here’s the thing, when you’ve got an incremental change that is on a single element or a single type of pattern, you can measure that when you change your whole site in a redesign. It all the KPIs in the world don’t really help you measure everything.
The important part of that is that when you’re doing a B testing, it’s it’s not just that you’re asking questions, it’s that you’re asking questions like you’re, you’re actually approaching the problem in kind of a kind of a sort of scientific or experimental, experimental way, where you’re looking at the actual like enumerated issues. And so like, it’s good, it’s good that you are testing something. But the important part there is that you’re approaching it in that way.
Instead of just I don’t like this. The reason this is true, and the reason it works is for anytime you’ve ever done any kind of science, you know, science experiment in school, or something. And when they teach you about, you know, multivariate testing versus you know, single variable. When you change your whole website, you’ve changed a bunch of variables all at once. And even if your leads go up, or your sales go up, or whatever, by changing everything at once, it makes it measurably harder to actually know why that happened. You can get right and you can figure out a way to make the analytics, explain it. But one thing you learn over time is that analytics can say a lot of things if you just push them the right way. When you’re changing one one thing at a time, it ensures that the one thing you’re changing is producing the right outcome. Yeah, that’s the bottom line to that. And it means you’re making the smallest, most efficient moves that you can make. Because you pick and choose, right, you may have a list of things you want to change, but you’re going to start with the ones you think will make the biggest changes, and you’re going to work down that list. And over time, what happens is you’re going to find out, well, something failed, we thought making a change to our lead form, and adding a certain field would help us get better leads. And it didn’t, it ended up getting worse leads because they were less well defined for some reason, or people abandon you added a form field, they abandon the field the form more, that’s fine, because that means you can abandon that change, and figure out a better one with that information you made. If you go through a full website redesign and your sales go down, then what do you do? incremental redesigns, let you fail in a much more elegant way. And in a way that lets you learn and get better. which is I think one of the most important pieces of web design that we really aren’t good at yet, as an industry,
I was just thinking that as you’re doing these experiments, and as you’re making these incremental changes a while ago, and I forget which episode it was, and I can’t find the number for it here. But we talked about architectural design records, this is a great place to kind of log that history. If you have if you are keeping an ADR for your site, that’s a great place to log these experiments and the outcomes of them. And just kind of having that as part of a sort of your institutional knowledge base.
The speaking of knowledge base is the final big piece I think of this puzzle is when you’re doing incremental work. And I said earlier that, you know, I think the one reason that this approach has changed in general is because things are stabilizing a little bit. I think it’s because we’re maturing as an industry. And you can see it by us learning to develop design systems, building out pattern libraries, because that gives us sort of the that source of truth. For our websites. Yeah. And so the pattern library may define all the headers on the website. And maybe we are at a point where somebody does say, Hey, we want to change kind of the look and feel of the site, what can you do for us, you can go in and say, You know what, yeah, here’s the three style of headers that we have. Why don’t we update that pattern will use will use a different font for this, you know, we’re only using the spot here It feels out of place. Let’s make it match the font down below. But we’ll do you know such and such to it,
I think the thing we talked about in Episode 40 with the status SS This, this, if you want to help make your redesigns more efficient, have really clean HTML, it will make it so much easier. So much.
And anytime you’ve got that record. And I think you know, the pattern library is where it really starts. Because at the end of the day, it’s all about efficiency. And it’s about identifying patterns, not pages, you’ll find yourself in a redesign, figuring out layouts, well, we need a two column layout, we need the three column layout, we’re going to need this weird four column layout, but it’s only going to be used on this one page. That’s an inefficient way to go about design. Because column layouts are easy. What matters are the cards that go on. And what matters are the individual elements that are embedded there. And giving the attention to those because those are the thing where the rubber meets the road in terms of our forums, here’s our forum pattern. And you can target all of these things. And that pattern library becomes this growing, evolving utility, that whenever somebody says, Hey, can we do x? You’re able to pull out the book, so to speak, and say, Yep, we got three ways to do it. Which one do you want? And if they say, oh, none of those are real good, then you can stand and say, can we? Maybe we have an opportunity, then? And we’ll, we’ll use that as that growth point to add in the new feature of the what, what’s your goal? What are you trying to accomplish? Let’s work that in right, and we’ll we’ll make a tweak to that element. Instead of you coming and say, You know what, we’re going to build a whole landing page for this, well, we don’t need to do. Let’s do the small thing. And we have that, that sort of flower bed to work with there. Eventually, you’re going to come up and somebody’s going to say you know what, you we need a whole new CMS, the CMS we’re using is no good. And that’s going to necessitate a lot of background work and some other stuffs going to get thrown in there. So redesign gonna happen. But until you have that really good, gotta rebrand, you gotta do X, Y, or Z that is truly necessary. This gets you from point A to point z by going through every other letter in the alphabet first. And it lets you look ahead of you and look behind you and know sort of what that path looks like. Yeah. Alright, folks, kick back. We’re gonna take a quick break and come back and we will round things out and Unknown Speaker talk to you later.
The drunken UX podcast is brought to you by our friends at New cloud. New cloud is an industry leading interactive map provider who has been building location based solutions for organizations for a decade. Are you trying to find a simple solution to provide your users with an interactive map of your school city or business? Will new clouds interactive map platform gives you the power to make and edit a custom interactive map in just two minutes. They have a team of professional cartographers who specialize in map illustrations of many different styles, and are ready to design and artistic rendering to fit your exact needs. One map serves all of your users devices with responsive maps that are designed to scale and blend in seamlessly with your existing website. To request a demonstration, or to view their portfolio, visit them online at New cloud.com slash drunken UX that’s in you cloud.com slash drunken UX. I hope you found this helpful. I know this was a weird rambling episode for a change. But I just felt like kind of talking about our experience here. And hopefully, did you know you got something out of it. You ever have questions for us? Let us know. We’re happy to kind of explain anything we went into. And, you know, give you some advice. If you feel like you’re in a tough spot with the up theory design.
Come, come reach out with us. And I’m not just cycling into social here. But seriously, connect with us on on Twitter, facebook.com. Slash broken UX. If you’re doing a redesign, and you just want to like, share the pain, or like talk to people who’ve been through it a bunch. Please vent to come chat with us on slack. com slash slack. I don’t think that Instagram is really the correct form for this. But you’re welcome to come check us there anyways, instagram.com slash drunken UX
podcast. And if you are feeling so inclined, run into your app of choice wherever you’re listening to the show today. Get your thumbs up, give us a rating review, share the episode, anything, just let us know that you’re enjoying things and let other people know that you’re enjoying things. We appreciate it. We hope you appreciate all the effort that we put into it. And if you don’t then let us know that will change something maybe I don’t know. We’ll do it. We can outside of that. Let me see it’s it’s late at night. I’ve got some empty bottles sitting here I need to throw away. I’ve got to probably go use the restroom at some point. I’m going to go to bed. I know I know. TMI. But it gets worse than that. Because I have really TMI because the big thing that I have to end the show with is just let you know that you got to keep your personas
close and your users cursor. Well played that. That might be my favorite one yet. Well done.
This episode of The Drunken UX Podcast brought to you by nuCloud.