Unbounce co-founder Oli Gardner has seen more landing pages than anyone on the planet. He’s obsessed with identifying and reversing bad marketing practices, and his disdain for marketers who send campaign traffic to their homepage is legendary, resulting in landing page rants that can peel paint off an unpainted wall. A prolific international speaker, Oli is on a mission to rid the world of marketing mediocrity by using data-informed copywriting, design, interaction, and psychology to create a more delightful experience for marketers and customers alike.

Video

Slides

Oli's coming back to Edinburgh! Catch him at Turing Fest on 3 August 2017...

Transcript

Today I’m gonna tell you…we’ll take a little trip into the future, the future of marketing, specifically the future of conversion. Now this is something we’re really obsessed about at Unbounce. We’ve created a new team; a machine learning team, so we can develop an algorithm that predicts conversion. Now and this is all that conversion automation. The word automation tends to scare people because you think you know, the machines are gonna take over, they’re gonna make decisions for us. That’s not really true. At least that’s not where we’re taking it. I mean, the machines are coming and they are gonna do certain things, but one of the great things they’re gonna do is they’re gonna start fixing some of the things that we are doing wrong. Right? Because as marketers, we make mistakes. Quite a lot of mistakes, often. They are gonna help us fix these.

All right. So, what is it we’re doing wrong, that these machines are gonna fix? Quite a lot of things. I mean we test opinions instead of insights. You know, people will just say, “Hey can we test that? I think it would be better.” It’s not the right approach. Testing the wrong things. Some things will have no impact on conversion, other things will. We need to know what they are.

Stop and test too early. Everybody does this, I did this for years, you know. Every single A/B test you run will have a spike in the beginning. You’ll get, “Oh, this is so good.” And then you push your life and then your business goes like this, or the opposite you get scared. So that will remove that problem. And Andy talked about some of this. We focus on the surface conversion metrics. How many of these did we get? How many new customers? That’s the wrong approach. We need to look right down, like he was saying, cohort analysis to get lifetime value. And the machine learning part of this in connected systems will do this for us. Because it’s kinda complicated to do it yourself. So all right, so we need to make a change for this and a little bit about what could be happening in the future.

Now, this is the Unbounce dashboard. What you normally do is you make a page and then you click “publish”. You won’t need to do that. This algorithm will have looked in millions of pages and it will be able to predict pretty closely what the conversion rate of your page will be without having any traffic. Like no traffic sent to your page. Low traffic sites like any small business, they have massive problems because they wanna test, they want to optimize but they don’t have enough traffic. We won’t need any traffic. I mean, it will also analyze traffic behavior on site, all of these things, but as a first step you’ll be able to make your page perform better without anyone even seeing it. You know, because it’ll be all of these annoying pages but also, it’ll look like say 20,000 of people that are in your industry, the same use-case of your page and it will give you recommendations, sentence analysis, word analysis, things you can do to make it perform better, or analysis. Exactly.

So you can make sure you gain your ideal customer. You know, maybe if you were to run a test, A versus B versus C, B converts better. What you do you push it live? But what if C got you more of your ideal customer? You can go through that. It’s really difficult and most of time you just celebrate B. With an automated system, it will correct that and turn on the right version. Every page will have scroll and click map analysis done automatically on your behalf and these insights will be pushed to you and again not just taken control of by evil machines. So instead of turning it on, modern interactions you get stack message saying that nobody’s getting to your CTA in your blog, if you bring it up to below the fourth paragraph which is the hotspot of the scroll map and it’s semantically relevant to the text there, you’ll get 118% lift in click-through rate on that. This story is based on us doing this…actually doing this manually and that’s the result we got. With these kinda insights, once you push to us, if you want to do this, just type yes and I’ll turn it on. Right? This is the human side of this. The machines give us the insights and we decide whether we want to implement them or not.

I will never ever have to ask this question again which every single person asks. You know, if you ask the CRO expert, everyone just says, “What should I test?” We’ll never have to ask that again. And it will get rid of the need for all of these blog posts, many of which I’m to blame for. Okay, I sent out a survey last week and tons of you replied. Like 150. Which helped shape the talk. So, I’m gonna focus on some of the things you’re interested in but also some of you submitted your pages and a few of them are gonna be included in the talk. I’ll try not to be mean. But we’ll see how that goes. So, that’s the future. Right, we’re not quite there yet but what’s going to happen. But for the time being, I’ve come up with these conversion equations. This is the human part. This is what we can actually do right now and so I’m gonna walk through some of them.

What is a conversion equation? Well, it’s got many parts. Heuristic analysis, rapid experimentation — I’m gonna show you how to do that. Conversion research, they’re closely related. Conversion data from what we have, and also Wistia — they share a lot of amazing data with me. It’s so good. Academic studies, ConversionXL Institute; we have Peep Laja from ConversionXL. We’ve built a lab in Austin, they’re doing a lot of great work.

All the tools that we hear about all the time and some of the frameworks from other thought leaders. Because the equation gets stronger, the more inputs there are so it’s not just one person’s perspective. These are the seven. Based on your responses I’ll go through three of them. And the first one, all right, what they do is they tell you what’s wrong and what order you should optimize based on the results of these equations. What you do basically there’s lots of exercises to plug in…actually I’ll show you one in a second. Good. Most important one is clarity. If you can’t communicate, you can’t convert. You have to get that right at the start.

Audio: ‘Fuck off’ doesn’t mean ‘go away’.’Fuck off’ means ‘fuck off’.

Can’t get more clear than that. And that’s kinda you know, you wanna feel that way. We want our customers to feel that way when we communicate what it is we do. So here’s the clarity equation. Okay it says so, you can do tasks to fill in all of the gaps in the early acronyms to give you a score. And to make it simpler, I’ve broken it down into seven parts. Distraction. What are all the things that are distracting someone from understanding your value? Proposition… and there’s lots of tools in here. So actually on the next slide there’s a link to a landing page which has all the tools I mentioned, a whole bunch of links. So don’t worry about taking these down now. Everything is on there.

Expectation. Does your page meet the expectations set prior to the click and does your CTA communicate what’s gonna happen next? Readability. I know a lot of readability problems. This is a longer sub-equation. I’ll go into some of that. Visual identification. If you look at the main hero image on your page in isolation, can people understand what you do? Immediacy. Can you do all of this in five seconds? Specificity. Are the details of your offer present? Right? The things that differentiate you from someone else could be price, could be location, whatever it is, you need to get the details in there. And hyperbole, which is a lot of business just yap on about themselves. They don’t really talk about the actual value. Breaks down like that. It’s called dervish which moves into a chart like this. This would be a perfect score which is never gonna happen.

This is more likely what you’ll have there. Now you know you have a readability and an immediacy problem. You can stop trying to run tests or optimizing the rest of them because you know where the problems lie. Now, I’ll go through some examples and I’ll use different sites to go through them, but you would just use your own site or your own landing page for all of these. Can people understand in five seconds? Okay, someone shared this in the call. My girlfriend shared this with me. She was looking at the software, and I thought, “Well, what does it actually do? Like you’re introducing it, thanks, that’s great, but I don’t know what it does.” So I ran a five-second test. Usability hub five-second test, if you don’t know you show someone a screenshot for five seconds, it goes away and you ask a question. It’s great for uncovering clarity problems, it’s qualitative analysis.

I said, “What do you think this page is about?” Lots of generic answers. Only 6% knew it was business intelligence software. Because they don’t really say that. Maybe they got that from the dashboard. So that’s a terrible score. How do you fix it? Well, I looked around the site and it took me a while. I went to the products page to figure out what they actually do, which is that, so I just replaced it and ran it again. Now you can see what’s happening, right? Immediately people are getting what it is you actually do, 42%. And this is really really easy and this took me no time to do this. So now you can see you’re starting to optimize this. Really simply. Because people don’t tend to tell you what they do. I see site after site where they congratulate themselves or you know, talk about awards. But they don’t actually explain what it is they do.

Readability. Does anyone remember Page Fights? A couple of people. It was an online show we did, me and ConversionXL where we would you know, shit on landing pages and it got quite violent and at some point we thought, “This isn’t really working. You know, we’re not getting the ROI from this that we want.” It wasn’t growing. So Pat got on growthhackers.com and asked the question, “How would you grow this business?” This is the first response, “I’m glad I saw this on a lazy Saturday when I had time to immerse myself in this lukewarm shower of digital hipster parochialism seamlessly integrated with slightly veiled program invective.” He’s not a fan, clearly. So I was, “This is a really complicated sentence so I’m gonna…” I put in readabilityscore.com which analyzes the text.

Based on Flesch-Kincaid reading ease score, a hundred is very, very good, which is the best, and zero is you can’t read it. They scored 18.4 which is really, really low. This is a great tool, put your copy in there and you’ll get a sense. It also gives you a grade level. You have to have a master’s degree to understand this guy, according to this. You wanna get, you know, a really high score there. And it’s not too difficult. But is a great way of getting a gut-check whether you’re being a little too complicated. And data to back it up. So this is brand-new. No one’s seen this before. This is from our…we’re putting it into our algorithm right now and you can see it’s not a big impact but as the reading ease goes up, which you want the conversion also goes up. You can see a slight incline there. The more of these, these equation parts that we actually can connect to into the algorithm, the stronger it will become.

To the ROI part, Peep made a nice observation because Page Fights is kinda in the beginning of how I met Nicole. So that’s the ROI of Page Fights. I’m quite happy with that. All right. Here is a carousel, you know, these things we get on websites at the top, everyone love them, everyone loves to hate carousels, well, because they suck and they’re stupid but the other time you get them is for testimonials. Super common. I’m gonna read this one, okay, so, “Carousels are effective at being able to tell people in mar…”, “Carousels pose accessibility issues for keyboard and screen…” Now I’m really frustrated. This happens all the time. And it’s so silly. People just turn it on from widgets with things they buy, and they think it’s cool. It’s not cool at all. Fucking hate it. This is the motion to reading facility ratio. So, how many of the words can you actually read before it goes away? You have on these? Just take it to your boss or client and say, “Point out how much of it you can actually read.” Because it’s really silly. Worst-case, lowest barrier: slow it down — better yet just extract them all out and make them better.

So, point four eight, there was some more elements that I didn’t have time to go through. But on that page as well, and I’m writing an instructional guide that will walk you through how to do all of this. So put your email address in there and I will send it to you when it’s ready; hopefully, next week — and I will never send you another email. I’m not building a list. You’ll get one email only because I just wanna help. So that’s about in the middle, that’s okay. So turn off the autoplay, simplify the copy… there’s something I took out, but you can quite easily make changes and make it better.

Okay: visual ID. I saw this landing page and I thought, “Why is Larry King on there? And more importantly, why is he shooting laser beams out of his thumb?” It’s just…I don’t know. I’m like, “Well, if I think it’s weird maybe other people think it’s weird.” So again, I went and did a five-second test. No, I’m not affiliated with Usability Hub, which is a great tool for rapid experimentation.

So, what do you think this company sells? Money, insurance, business loans — hardly anyone got it. Specificity: the word ‘business’ is critical to their success. It’s not a loan for anyone, it’s just for businesses. So I thought I wanna see if it’s a Larry problem. So I blurred everything else out asked, isolating the visual, “What do people think it is now?” Meeting a guy, a talk show, something about old men, doctor consultation, Viagra — and all of you just didn’t get it. So we definitely have a Larry problem. So I just deleted him. Got him on Photoshop and I just kicked him out of there. Ran it again, “Look at that.” That’s exactly what they do. I mean, he’s a very high priced endorsement. So, you know, they have to put him somewhere but don’t let him distract from what you do at the top. You know, visuals need to mean something. That’s a 500% increase just by deleting him.

Okay, this is a magazine. Two different dentists with the same stock photo side by side. How lame is that? And this one’s cheaper, so I’ll just go with them but I stop trusting them because these guys aren’t professional enough to actually have real photography. Denise. This is “Sandstone Castles”, this is Denise’s page here. I looked at this image and I thought, “Now it looks like a stock image.” It’s almost as bad as the overhead view of a laptop and a coffee cup that most SaaS businesses tend to use when they first start out. So I thought, “Let’s have a look at this image.” So I went to TinEye, a reverse image lookup and there are 231 sites using this image, some of which you may have come across — and, like, maybe you didn’t like them, so then by association, you’re like, “I don’t really like you either.” And it was 229 yesterday. So more and more companies are using this, so just take it out. Do the Larry thing with him, with that. And you know, get your process further up the page. How do you work? And you, right? You are there as well but just bring that stuff up. Because that’s what people care about for small business marketing. Or find a visual represents something that you do and put that in there. But don’t put an image in just because people say you need to have, you know, a hero shot. Right, that’s what people thought you do. Computers and keyboards, that’s not all you do. Nobody got it right.

What do you think these guys do? Mattresses. Alright. Energy. Like, electricity and shit. Okay, don’t do that.

This is a lovely site. I love this. But the biggest problem I have with travel sites is: you don’t tell me where it is. I love this, but there’s nothing there. Like, I wanna go there. I wish someone would do Shazam for landscape photos. You could just like, “Oh, that’s where it is.” That would be amazing — if someone builds that, I wanna invest. Just put a caption, or maybe a CTA like, “Book a trip to this place right now!” That’s kinda what I’d like to do.

This is how you do it right, where’s Fishbox? Over here. This is how you do it. I look at this, it’s a perfect context of use. I know it’s gonna be fish with, one of them, what’s it called… instructions, recipes. You know, white box with blue bands around it. I know what it is, I know what to expect, it’s perfect. That’s what you need to be doing. Okay, use that as an example.

So if you run these things and if you figure out you have a visual clarity problem, how do you fix it? Well, so I took this one, because this is just about as generic as you can get for a dashboard. Test them with or without a caption. None; 60% got it. Just because people read captions. Now you have a design problem so, that’s not solving it’s just making it better. So you still wanna go and try and make, you know, a better graphic there, potentially with callouts that point to the features and benefits that are actually important and make you different. Okay, so that brought that right up. You know, like I said, on that page you can get the full guide once I finish writing it because I had to skip some of the parts there.

Now, where are the guys from Dublin? So, I was in Dublin speaking and I went to… I checked into my hotel and they gave me a card, a key card for room 324. So I go down the corridor and there’s a lady waiting at the elevator, so I await beside her, and we chitchat, elevator opens, I let her in first. She goes in, puts her card in and presses the third floor. See, now I don’t have to do anything, so I kinda awkwardly stand beside her, comes up and I let her out again first, now I’m following her out of the elevator, I see the sign, I go, “Yeah, I go this way.” I’m still following her like some creepy pervert, going down and it’s not until I see two doors in a row where I get like, “Shit, I’m going the wrong way,” but now I’ve walked her right to her door. She goes in, you know, and it’s so dodgy.

So I was kinda upset about that so I went back to my room and now I’m gonna run an experiment on this. So I describe the context, and I said, “You’ve got room 324, would you turn right or left?” 33% of people went the wrong way, right? So we have a problem here, so I mocked it up graphically so I can change it, manipulate it, ran it again, exact same problem. Now I’m gonna apply… So I know I have a design problem, right? So when you have a design problem you need a way to solve it. ATD, “Attention to Design,” is the 23 principles I have, graphic design principles that can help you focus people on what they are supposed to be doing or what you want them to be doing. Now, I used four of them to fix this design.

So first of all, contrast, I made that stand out because they’re semantically different. This is where I am, this is where I’m going. This makes me relax, I’m on the floor and then I put these numbers with the correct arrow and a little bit of white space in the middle. Only 8% when the wrong way. And then I put some interruption. I broke it apart so it’s really obvious which way to go. Everyone got it right, way fewer perverts in the hotel, good for everyone. Okay, so if you can identify a design problem, use these principles to fix it, because nobody has really ever been taught how to design for conversion. These principles will help you do that. There’s an e-book of all of that. Again, it’s at that link on that landing page.

Design matters. It matters a lot, much more than people give it credit, you know, it creates joy, we all know that with like, say, Apple products or whatever it is we like, we love, but it gets overlooked in many ways. It also reduces confusion. Now, if you have a car or if you’ve ever driven in your life… right? Obviously. You’re at the petrol station. I got like UK with you there, and you have that little arrow and it’s really important, because if you don’t have that… this is really sped up. It takes two and a half minutes and someone else comes, fills up, pays and leaves before he actually solved his problem.

It should reduce germs. You’ve already seen these things, so guys are coming, the guy was going and you lift it up with your foot. Try and do that. Do your business. But it’s kinda gross. If a sensor is blocked, automatic flush, manual activate, lower the seat and it’ll use a manual button. But you’re not gonna do that again because it’s really loud and you don’t wanna touch the thing. Just move it two inches to the right and it will be perfect. Right, design can so easily solve problems. It can speed up repeatable tasks.

Video: I find another design problem, over here. So this happens every day, all the time, many times and that’s the problem. Repeatable problems. So I need to get in, here’s the card reader. Here’s my card in here. Okay, sure I can do that, but it’s annoying, right? I’d rather…see, annoying. I’d rather do this, but it looks like I’m humping a wall, and it looks, but, I mean, c’mon, it’s just move it, from here to here and I could just walk past it. Right? Or, put it there. I still look like I’m humping something. But, see what I mean?

And then when you merge design, I’m thinking, with the data that we’re gonna have access to, that is the fastest path to delightful experiences. But you might be thinking, “Well, I’m not a designer or a developer, like how do I create the mockups for these tests that I’m doing?” Right? Because I’m a designer and a developer, I can do these things. You know, “I don’t have the Photoshop files, I can’t delete Larry; I’m not a coder, I can’t do some CSS.” Well, it’s actually it’s pretty easy. Without permission, you can actually start doing this stuff.

Video: All right, Edinburgh, this is the hack I’m talking about. You take that line of JavaScript and you paste it in the address bar. Now, for some reason, Chrome strips it out, so you put JavaScript back in and what it’s done, it’s made this page editable, live. Not live on the internet but live in my browser so I can make changes without asking for permission, without any tech skills. So what I’m gonna do I’m gonna change the value proposition of the conference website to see if I can sell more tickets. So let’s say, “The only marketing event in Scotland with a free deep-fried Mars Bar,” because clearly that’s gonna come back way better than the original. So what you do is, you take a screenshot of that, you go to Usability Hub and you create a five second test. You upload your image, then you can set up some context and then you can ask questions. What do you think?

It’s really simple. And you get all these answers back and sometimes they’re hilarious or scary and you can take them to your stakeholders and go, “Everything’s kinda broken. Can you give me some budget or some time to try and fix this?” People are much more willing to do that when you can show them evidence. So there’s another kind of test you can do. It’s called a preference test. Where they have as long as they want to look at it, and you’re basically saying, “Which one of these do you prefer?” So, I teed it up by saying, “Your boss has let you go, has given you the okay to go to a marketing conference in Edinburgh, which of these would you rather attend?” Deep fried Mars Bar won by 27%. So then that got me really curious. What else could we do to, you know, to calm brains, nerves, and sell more tickets for next year? How about a “How to headbutt a mugger” class? Right? Classic Scottish self-defence. No, that didn’t work, I mean it’s a little intimidating, Braveheart on VHS? That’s gonna be a collector’s item for sure. No. Scotch tasting, that’s bound to work. Yeah, okay we’re going the right direction again.

Let’s get serious now, free Hearts tickets. Hibs tickets? Whew, I don’t know how many Hearts fans are out here. Sorry. There’ll be none of you here next year. And then, how about Haggis Hunting? And then you can get the international audience, because they’ll think it’s real. Guys, here’s some food for thought. Or, here’s just a little tip you can talk to me about if you want. Upwork is like Elance. For 25 bucks, they’ll build you any page from anywhere, and build it inside Unbounce so then you can just change everything. You can move stuff around, change the copy, change the graphics, whatever you want, really easy, and then you can do all this experimentation without any permission. Just get a free account whatever, you know, just like, it won’t cost you anything. So yeah, the full instructions for all the clarity stuff will be in there, hopefully next week.

I’m from Vancouver, but I lived in Montréal for two years and it’s really cold there. So I was looking for a sunny vacation for Nicole and myself. I saw this one, “Oh, it’s interesting.” Weirdly it sends me off to airtransat.com, I mean, that’s kinda weird. Expedia’s usually pretty good with that stuff, but I’ll look around, it looks kinda good, I’m like, “Book early and save $400 per couple per week.” Now, I know it’s not a button but it’s kind of suggesting I can do — I can book early and save. Nothing happens, I’m like, “Oh, what can I do?” With my 20 years of experience, I know if I click that little tray there, it’ll open up and there will be a CTA, right, hiding inside it. Boink. Exactly what I was looking for. I’m feeling so proud of myself right now. So I click on, “Save up to $400 per couple per week”, so same thing. And this closes again. So I’m like, “Alright, what’s the other one? What’s that? Well, it’s for a single parent. I’m not a single parent but I just want something to work.” Closing again.

I do this for like, a minute and a half. Just like I’m clicking like the little arrow, maybe just $400 and like nothing happened. I’m like, “What can I do on this page? It’s so annoying!” Then I see these social share buttons, uh, because I’m gonna go to Google+ and talk about this because that’s dead. Maybe I’ll email it to myself, to remind myself never to come back to this shitty website. Printing it might be a good idea. You take it to a travel agent like, “Could you book this for me? Well, I can’t do it myself.” And then I see this. I’m like, “Why is there a QR code on here?” You know, I have a QR code reader on my phone. They don’t come by default, I’m just morbidly curious about shitty experiences, so I’m like, “Alright, I’ll do this.” Sure enough, up it pops. So I pull out my phone.

Video: I have seen QR codes on websites, and I have scanned them. Do you know where they take me? Back to the website.

It’s exactly what it did except now it’s a call-to-action. Like, I can spend three grand on my phone but not on desktop. I’m not happy about that.

Audio: Bullshit!

It’s all bullshit. There’s so much bullshit marketing out there. The first rule of CTAs. Now, you probably think, well that’s really obvious, everyone has to call to action — well, they didn’t. And I bet if you look at the top 20 highest traffic pages on your website, half of them will have old content and half won’t have a CTA. We’ve done it too. For seven years, we’ve owned ‘what is a landing page’ on Google. It takes you here. This is like an information page, lots of internal links to other kinda generic pages, side navs. So what we did, we removed the internal links, took away the nav, added a button. We know on our highest conversion path if people go to the templates page, they’re more likely to convert. So we just say, “Hey, look at the templates.” 171% increase in people going to where they want to go, just by following a few simple rules. Every landing page, every page on your website has a conversion opportunity and every conversion opportunity needs a CTA of some form.

Whose site is this one? Oh, there, okay. Now, this video, I mean, I don’t know how much you look at your video analytics but, there’s a good chance most people don’t watch it. Maybe as low as 10%. You need to know that. And this is a great video. I think the content in this is better than the content below it, and you need the supporting content in case people don’t watch the video. What I would suggest is, a contractor is your target audience. Sit down with them, make them watch the video and then simple instruction, whenever they see something that is…they like or is beneficial to them, tell them to stop and have a conversation about it. The copy you’ll get from that, it will fuel your website, your campaigns, your ads…everything. It will be…that’s the voice of users you really need. And you know, they may also not like any of it, but you know, I think that’s a good technique, because video is complicated. Right?

This is because of all the data I got from Wistia. And you know, about size of the video player, where you should put it — but more importantly, talking about CTAs, there are three types of CTA if you use Wistia for your video hosting. There’s a turnstile, that’s an email collection a portion of the way through; there’s an annotation, like, little thing you see on YouTube that you can click up there; or a full-screen CTA. And I want to see what the impact of certain words, I’d love to do some of Nathalie’s stuff with this, on this — so, again, this is really long, there’s the acronym because we all know if you’re running in a TALC you need to be really quiet about it… because why do you have talc?

Anyway, so looking at annotations vs. CTAs, is there a difference when using certain words in the impact of conversion? So in annotations, up here, “Click here” destroys not saying “Click here,” and it kinda makes sense, because it’s over there, like “Click over here”. For a full-screen one, it’s ‘click’, it’s a little softer but both of them would be not saying it. It’s a nice action word, people kinda need to be told what to do sometimes.

“My” versus “your”. Lots of peoplel debate this. ‘Start my free trial’, ‘download my e-book’, ‘your e-book’. Well, for the annotation, “your” wins; fullscreen, “my” wins. There’s a pattern happening here, that the bigger “right in your face” thing needs to be a bit calmer, a bit nicer. How about, “now”, some urgency? “Now” works for annotations, but not for the full screen CTA. Again, these things are complete opposites. How about some exclamation points? With annotations, you just go crazy like… actually, that’s low sample size, this is massive amounts of data. So this is very, very good data, apart from that one. So you have to go into… one exclamation mark actually works; full screen, don’t do it, right? Every single piece of data here is completely opposite. Which is why it’s dangerous, like, say you’ve got a blog post about a certain thing and you implement it everywhere like, “Oh, this in a CTA works.” No, every one is different.

Here’s like a summary page — you’ll get the slides after. And this includes our Unbounce data about lead gen forms of the same kinds of things. Every conversion opportunity is unique. In the future of marketing, at least what I wanna see in the future, when you’re typing in your CTA, it will give you an auto-suggest, kinda like Google, but with a predicted increase in conversion, based on all of this data — which would just be mind-blowing, right? Every single conversion opportunity is unique. You can’t apply what you know on one thing to others. We have to… and that’s what machine learning will do. It will stop us making these mistakes I was talking about in the beginning. Because that’s what makes us human when we can forget the human part of this.

Here is a photography course I was critiquing. And I got to the testimonials, and I started reading them, “Megan has been a life saver.” No, she hasn’t, this isn’t Baywatch. This is a photography course. “Reaching out to Megan’s one of the best things I’ve ever done.” You must have a shitty life, if that’s the best thing you’ve ever done. I mean, don’t get me wrong. This is actually a really good course, with some amazing photographers that come out of it, but this is getting to realm of hyperbole, where you’re like, “Uh, it’s a bit over the top.” You stop believing in it so much. You’ve gotta be really careful of what you put inside quotation marks.

I don’t want him outside my house. How mean is that? “Yeah, we’ll let you marry her but you’re not getting any of my money.” It’s terrible. This is from someone at some HR thing, I’m gonna read this testimonial out. “I must honestly thank you for all your assistance. You have personalized the whole experience that I have had with you, and kept in touch by communicating with me on a regular basis which other agencies have to failed to do following initial registration with them and I appreciate this very much. I would have no hesitation in recommending you to others, you may be looking to secure…” God, this is so boring. No one’s gonna read that, ever. It’s by Barbara from Glasgow, I mean, it’s so anonymous.

The secret to a great testimonial is demonstrating the transformational impact of someone being your customer and going through this. That is why the sleazy before-and-after photos work. You don’t have to be sleazy, but it’s a really important thing. Here is the testimonial equation. Again, there’s a lot, but we’re in a church, so Reverend PHAT CRIB is actually quite appropriate. So I’ll apply it quickly to this one, one of these generic ones. Readable, that’s for like, does it have the swoop? It doesn’t, so it gets one there. Enumerated? Innumerable, there’s no numbers in there by the transformation. Verifiable? If you have a link to the source, that is great — just make sure it’s in a lightbox so it doesn’t leave. Proximal, is it near what it’s talking about? It’s not. With a group of other testimonials. Non-hyperbolic? Yes, it’s not you know, blabbing on. Has a photo? Doesn’t count. Transformational images, video text. None of those. Contextual, relatable, isolatable — do you know what it is in isolation again? No, you can never tell what they do and it’s not branded. Right? 18%. So this is a terrible testimonial. This is on our home page. So going through it again: readable? Doesn’t swipe. Renewable: you can see it goes from five to twenty. Verifiable: you watch the video and you see him and he’s kinda trustworthy. I’ll give that you know…and it’s a little bit of that. It’s not kind of sourced, but it is linked and it’s in a lightbox, so it scores well there.

Proximal: it’s right next to our main value prop, talking about conversion. It’s non-hyperbolic, has a photo which isn’t stock, authentic image gets a one. Transformational images: no, it doesn’t, but it does for the video and the text. These are weighted. Images are best because you immediately see the change. Video is second best because you might not watch it and text kinda comes in third. So, score’s great there. It’s contextually related to where it’s around. It’s relatable; I believe he’s a marketer. Like, so, the kind of tribal thing. Isolatable. On its own you know it’s Unbounce, you know it’s worth the trouble digitally. You understand the whole thing and it’s branded with our name, 85%. Right? You can see how much better that is. So what you can do, on that landing page again, there is an interview script. Take one of your customers or several customers, and ask, read this script out. The questions in there will give you every single piece you need from the equation to have an amazing testimonial. It’s really, really easy to do.

I bought something online, it was a health product and they sent me the wrong thing. So I emailed them and said, “Hey, you sent me the wrong thing.” And they’re like, “Yeah, we know. We sent yours to someone else and you got theirs. If you print out this thing, put it on mail it back to us, we’ll check it back into our inventory. When we make sure it’s in there, we will send yours out.” And I’m like, “No, you send it right now. You messed this up.” And they’re like, “That’s not our policy, sir.” I’m like, “I don’t care, you just lost me as a customer.” Eventually they agreed to send it to me right now. When it arrived, I had the right product in it, but it also had a whole bunch of samples because they were trying to suck up. And this is one of them. You probably can’t see, it’s a… Right? They didn’t care enough to actually make a proper apology, you know, the only relevant part is that. Or maybe ‘formula’, because we’re talking about equations.

So, we did an event, an online event for agencies. It was a really cool event. This was the landing page. Now if you left, you got an exit-intent pop up. You know, “Everybody hates pop-ups! Pop-ups are evil! Nobody likes them! They’re terrible, don’t use pop-ups!” It’s not true. Anything can be delightful. Right? You know, emails, banners… pop-ups can also be delightful. We acquired one of the companies that does exit-intent stuff. So we had access to all the data. On average, 2% to 3.5% they convert at. A really good one, five; an excellent one, 8%. This one converted 19%. And we did a similar event in Germany, almost 27%. Right? So, these are really helpful. We’re giving them recordings, videos, very high-value. We see this all the time, when you’re giving away a recording or video, these things work like gold. Because it was relevant, valuable, well-timed and designed to delight.

Technology is not the problem. We don’t need to be scared of machine learning or this or any kind of automation. We are the ones that cause the problems because we misuse it. We used psychology in manipulative ways, you know… a lot of affiliates do that, because they don’t give a shit about your customers. They care about the high-level conversion because that is what they get paid for and then they leave. We can’t behave that way. We have to be more respectful, and then we can put our value systems into these algorithms like, you know Moz, TAGFEE. If you put TAGFEE in there, then the suggestions it makes for you and for a headline change or something will be more in line with how you want to talk to your customers as a human brand. Right. So we shouldn’t be scared of it. It’s going to be amazing. Thank you very much.