INBOUND 2016: Peep Laja "How to Figure Out What Really Makes Your Website Work For You"

My name is a scientific experiment to figure out what people trust more, what they hear or what they see. So it goes like this. I meet somebody at a conference, hey, my name is Pep Playa. Nice to meet you. Here's my card. And they read the card and say, ah, be pledger. I just told you what my name is. So, yes, people trust their eyes more than ears. So I'm an optimizer. I run a lot of experiments. I'm testing ways, how I introduce myself or how I explain what I do for a living. So at a random cocktail party, when I say I do conversion optimization, they're like. And then I try different approaches. Like I sell money. What? Now they're interested. So I optimize stuff. I help companies grow their businesses. And how does conversion optimization really fit into the growth thing over here? So let's imagine that this is our company growth line. And it's flat. We're not really growing. And the upper line, the blue line, that will be our growth line. If we could improve our conversion rate just 1 1% per year. But what would our growth curve look like? If we could improve our current conversion rate, a modest 10% per year, no more traffic, no more ad spend. Everything else is the same. We just improve our conversion rate 10% per year, that would be our growth rate. Or 15% per year, that would be our growth rate. So in six years, we would double our revenue without doing any additional traffic, no acquisition. So that's conversion optimization. The way I see it, it's compound interest for growth. Okay guys, bad news. You're all fired. So let's imagine that you're fired. You just got a text from your boss. No more job. But there's some good news. Some idiot gave you a job and you're now optimizing this website. And the goal is you. You have to improve the conversion rate of this website 20% per year. Now the question is, how do you do it? Think about it. That's the website we need to improve the conversion rate. How do we do, how do we do it? How do you go about it? And that's a million dollar question. So how do you optimize websites for higher conversions? We could just go to Google and conversion optimization techniques and you might find a blog post like this complete list. Why would you need anything else? If you have this kind of a block, it says it has 100 plus techniques that you can sort through. Amazing. Okay, okay. So we have 100 conversion optimization tactics. What do we do with that list? How do we use that list. Maybe we'll just implement every tactic all at once. Great idea. Then your website will look like this. I was like, it has everything as call to action now, urgent, discarcity. You name it, it has it. But it will look like this. So that's not the way. Oh, what about testing? We can test each of these 100 ideas or tactics one by one. See which ones will work and which don't. Easy, right? So how long does an average a B test last? So an average AB test for an average website runs for like 28 days. So testing 100 ideas one by one takes only seven and a half years. Yeah, we don't have that kind of time. So a list, A random list of tactics we find on the blog post. Not very useful. What about checklists? This is from moz.com, they produce great content, right? 91 point checklist. You know, cross every dot and whatever and then you have maximum conversions. Well, unfortunately, life is way more complicated than that. We can't use checklists or lists of tactics. That's not the way to optimize a site. What about best practices? There are best practices for everything. Best homepage, best checkout, best cart. You name it, there's a best practice for it. So are best practices useful? Yes, they are. But as a starting point, best practice is where you start, not where you end up. So it's not optimization. This is just the beginning. So best practices, not optimization. What about design trends like ghost buttons? Everybody has to have ghost buttons now. It's a button we want people to click and we make it less visible. Great idea. Or what about video backgrounds? Every cool website has to have a video background now. Well, the thing is that human brain reacts to movement and we can't help but look at the movement. Right now you're having trouble listening to me because you have to look at that guy and shake his ass. So that's what's happening on your website. People don't understand what you do. They don't read your value proposition. They don't want to buy from you because it's a distraction. Don't do that. Well, what about market leaders like Amazon? They're doing pretty fine. Conversion rate for Amazon prime subscribers is 74%. 74% of prime subscribers who go to Amazon.com, buy something. What's your conversion rate? Maybe 5%. Let's imagine it's 5%. That's not bad. So that means it's 15 times lower than Amazon's conversion rate. So is the Design of Amazon 15 times better than your design or is their copy 15 times better than your copy? No, it's not. So you thinking that you're just gonna rip off a market leader and you're gonna get Amazon style results? It's just naive. World doesn't work like that. What about copying competitors? Well, they don't know what the fuck they're doing either. Don't copy your competitor. They looked at some other competitor who looked at some other competitor and it's just blind leading the blind. It's like, oh, they're doing this. They must have tested it. No, they didn't. They just copied it from somebody. So don't do that. Your competitors are probably not very smart. So how do we then optimize a website? Well, we need a process. Because if you can't describe what you're doing as a process, you actually don't know what you're doing. So if somebody asks you how would you go about improving the conversion rate of whatever website by 20%? And if your first instinct is to start saying, oh, maybe I'm in, make a better headline. A bigger button if you start with tactics means that you're actually pretty clueless about it. That's not how you optimize a website. So we need a process. How do we optimize our optimization process? And the most important thing to understand about optimization is that we need to figure out what actually matters. Well, think about your website homepage. It has a bunch of content, different blocks, images. Why are those content pieces in the order they are? Probably random. Probably just the designer thought it looks cool. So which parts of your website actually help to improve the conversion rate? Which parts of your websites lower the conversion rate? Do you know which part? Which words on every single page actually help and which actually hurt? Do you know? If you don't know, how the hell do you think you can optimize? We need to figure out what actually matters on our website. And then we also need to figure out what matters to our target users. How do they buy? How do they want to buy? Why aren't they buying more? How whatever we're selling fits into their life. We need to figure out what matters to them when they're buying for whatever we're selling. What is the number one decision making criteria? Second, third, and so on. Because if you don't know that, again, how can you optimize it? It's not about make the button bigger, it's about understanding what actually matters. So a good conversion optimization will do three things. It will tell you what the problems are where the problems are on your website and why these problems are problems. There are two ways to figure out what actually matters. One is testing, which is test something, see if it works and if it doesn't, or it does or doesn't, whatever we learn. But we can't start with testing because we have the same problem as we would have with this 100 ideas for testing. Because we wouldn't know how to prioritize. We have no idea what will work. So it's wasting time. So we need to start with research. And when we're doing research, what do we need? What do we need to gather? We need to gather data, right? But you don't want to be this douchebag because you don't need more data, you need better data. Often people say, oh yeah, let's log into Google Analytics and see what the data says. Well, data doesn't tell you shit. Data is just passive, it's just there. Being data driven is actually all about being people driven. It's up to us humans to look at the data, interpret the data, pull insights out of the data, and so on and so forth. And a lot of people have a lot of data, but they don't know what to do about it. It's like I log into the analytics and like, I don't know what to look for. Well, yeah, because if that's how you're doing analytics, you're doing it wrong. Because the way you handle any analytics cases is that you start with a list of questions, what is it that we need to know? And once we know what is it that we need to know? Then now we can figure out what data can we gather that helps us answer those questions. And then any data that we look at now makes sense. We know why the data is there. It's there to answer some of the business questions. And what these questions are, is highly contextual. Different websites are different and different audiences. It could be one of these questions, some questions also universal, but it's about understanding the user and it's about understanding where, what, why is a problem on a website. But be asked to be careful because if you torture data long enough, it will confess to anything. If you want data to show you something, if you like your, your idea and you want data to prove your idea, you can torture the data in a way, slice it and dice it, that it, it will support your point. But that's not what we're doing, right? We don't care about egos, we care about the truth. Right? So Research Excel is this framework that I've developed and now used over the last five years successfully on hundreds of websites building their businesses. So this is a framework that helps us figure out what really matters. So it's a method, it's a process for gathering data, analyzing data, and coming up with a list of what these problems are with our website. And step one is always technical analysis because we need to make sure that our website works with every single browser and device combination. It's ridiculous how many websites are losing money because it doesn't work in every single browser. I thought this process to this guy who told me later that, oh yeah, now I noticed that conversion rate for Safari Mobile was halved because of a bug and 10 minutes of investigation, I could fix that. So you go to your Google Analytics report and you compare conversion rate per browser version so Internet Explorer 9 to Internet Explorer 10 and 11 and so on. If 11 converts at 5%, but 10 at 2%, maybe there's a problem, maybe there's a bug. And that's the easiest thing to fix because you can have the most persuasive website in the world. But if it doesn't work on the specific browser device combination I'm on, it won't help. Step number two is heuristic analysis. So this is where, where we actually go through the page, through the website, page by page, like a regular user. And we observe different things. We observe relevancy. Do I understand where I am and is it for me? And who are you? And who am I? Do I understand, like the clarity of the value proposition? Do I understand what they're selling to me? So often it's completely unclear and what is being done on a given page to increase our motivation to take action and how can we make it easier to take action? So let's look at some examples here. So this website says, start accepting credit cards today. 2.75 per swipe card. And there's an image showing how to use the product. You can't have any more clarity here. It's the maximum clarity. And think about the way they're using this image here. Because without the image, how many words would you have to use to explain how to use this product? Like, I don't know, 10 paragraphs. Take it out, put the thing in. Swipe images help to clarify your value proposition. Which of these circles is the most important circle? The blue one. Right? The blue one is the most important circle because it's the biggest. So every single page on your website has to have one primary goal. And then if you Know what you want people to do on that page? Make it the biggest thing, make the page all about that one thing and then people will focus on it. So that's visual hierarchy. Oops. Yeah, technical issues here. So there's an obvious website here that sucks as. Yes, exactly. Good question, good question. So heuristic analysis is about us going through the website page by page and asking what's wrong with this picture and unfortunately can't show the images here. So let's move on. Next step, we go to digital analytics or Google Analytics or whatever. And the goal here for us is to figure out where is it leaking money. Every single page on your website is leaking money. But where is it leaking money the most? Maybe people get to you. Let's say you have an E commerce store. Maybe they get to the category page but they're not clicking, clicking through to a product or maybe they get to the product base but they're not clicking add to cart. Or maybe they get to the cart but they're not going to check where is the leak the biggest because that's where you want to start your work, your optimization. Second thing is what's the difference between high converting users and low converting users? What are they doing differently? And let's see if we can see some examples hereafter. But the key thing is that every single thing that the user can do on a website, they need to be able, you need to be able to measure it. Because if you have some widgets and filters and I don't know what on your website and you have, you're not measuring whether people use it or not, how do you, how can you ever improve it? So if there's like say people can sort by price or filter by review, like if they do that, is it helping conversions, lowering conversions, no difference. How would you know? How can you optimize if you, if you're not measuring stuff? Seems like the slide deck has given up. There's going to be an encore later, hopefully with working slides. If you guys want to see the full stuff, I think the inbound people will share the slides. Yeah, yeah. Awesome. So digital analytics figure out where is it leaking money and what correlates with higher conversions and in mouse tracking. So this is, you know, tools like hotjar or if you're more enterprise or like testable insight and tools like that and they can show you mouse click stream data or mouse hover data. Hover data, not useful at all. But clickstream data where people click how far down they scroll. Very useful. So for instance, if we know that people who read more content convert better, but how can we get them to read more content? So with scroll maps, we can see how far down they scroll on any given page and we can see where the drop off is. So when we see where is the drop off point, we can compare that to our design and figure out what makes people drop off at this point. And it's usually when on a page you have strong lines or background color changes. So different content blocks of different background colors or background images, people often think that that's the end of the page or these unrelated content blocks, even though it's a whole, it's the same narrative. So be very careful about switching that up because that will stop people from scrolling. And very important feature of these tools is session replay. So we can actually watch videos of people going through these pages. So let's say that you have a checkout page and, and you don't know why people are not buying more stuff now. Now we can go and watch videos of people filling out the checkout and see what the problems are. Once I was working with like a resume building service for truck drivers and there was a five step funnel and in one step three, a bunch of traffic was dropping out. I was like, what the hell is going on there? And then I started to watch videos and I noticed like people start filling out the resume and the name and you know, employers and blah, blah, blah. And they get to a question that asks now name three references mandatory. People didn't have three references, so they abandoned process. I told the client, hey, this is happening, so we don't need that. Took it off, conversion's up. So session replays helps you help you figure out those, that kind of stuff. And then we get to qualitative surveys. So this is us talking to our buyers in two main ways. We want to survey people who just bought from us or signed up for software or whatever, people who gave us money. And we want to send them an email survey within the seven days after they signed up or purchased something for the first time while they still freshly remember their purchasing experience. And then we want to ask them, hey, what kind of doubts and hesitations you had before completing the purchase. What was the one thing that nearly stopped you from buying? Did you have any questions you couldn't find answers to? We want to understand the friction they experienced in the purchasing process so we can fix it. Also, we want to understand their motivation. So we want to ask, what did you buy? What made you buy this thing? Like, what problem are you solving? For yourself with this purchase. So we want to understand the user intent and all these questions should be open ended. We're not doing multiple choice. This is not like NPS, like OH3. Okay, so what do you do with that information that you know it's three? Nothing. That's why you should never. So this is an example. So on this page, massive drop off. Massive drop off on this page. And it's like, why, what is the problem? When we look at the page, maybe it's the header, maybe it's like the whatever thing. And we have like all these speculations what it could be, how can we improve this page? So we put a poll on this page with hotjar and asking them, we were asking them what's holding you back from completing this purchase right now? People filled out the survey, the poll and then we created a word cloud out of it. And yes, it was pretty clear what the problem was. It was high shipping costs. High shipping cost. But just looking at that page, it was very difficult if not impossible for us to figure out. So don't speculate why people are not taking action. Ask them, ask them what's holding them back. They will tell you. And finally we do user testing. So this is, we recruit people who are our Target audience through trymyui.com or usertesting.com and we give them three types of tasks. One, find something specific. Find a pair of jeans in size 34, under 25 bucks. And then we observe how they go about it. Or we give them a broad task as a second task. It's Christmas coming up, find something you would like and then just, we want to see how they go about it. If they don't know what they want, if they just want to browse and in the end go ahead and make a purchase so we can see the final completion and, and never listen to what they say. Just observe what they do and what they struggle with. Obviously this is not working. I have a video here that the guy is just trying to check out. It's a horrible, horrible experience. Like seven minutes trying to complete the purchase and it's really traumatic. And in the end there's a post user testing survey. Oh, did you have any problems or what did you think about the website? Oh, it was so easy to use and people were just lying. So don't listen to what people say, just observe what they do. And if you go through this process, you have your own list of problems. You don't need a blog post with 100 tactics. You have your own list of 100 problems you have identified on your own website. And now you start fixing these problems and obviously you don't know what the optimal fix is for each of these problems. Even though even if we know what the problem is, we might not know what the solution is. That's why we need to test. So first of all we need to measure, fix all the measurement stuff. If we're not measuring everything that users can do, have to measure that some stuff is just no brainer. People can't read the small font size on a website, just make it bigger. And then the rest of the more complicated ideas we need to test because we don't know what will work. And usually I like to put all these problems in a Google spreadsheet like this. So what the issue is, what type of category issue it is, why is this issue an issue? Kind of a general idea how to go about it. And I put like a priority rating of five stars. It's like, oh my God, we're losing so much money because of it. And one star is like, oh, this is like a minor usability issue that should be fixed eventually. And there's a prioritization sheet that is obviously not working. Yeah, some case study. It worked. Great job, good stuff. So with Research Excel you have a systematic, repeatable, teachable process. So when you want to hire new people and teach them conversion optimization, you can just teach them this framework and they'll be ready to roll from day one. It's not about, you know, a list of random tactics, just teach them a process. And this is industry agnostic works across any, any type of website. And of course this stuff is difficult. It's like, like anything else, it requires practice, it requires you just putting, putting effort in. Ronnie Coleman like six time Mr. Olympia said that, you know, everybody wants to be a bodybuilder, but nobody wants to lift no heavy ass weights. Almost the same thing with conversion research that everybody wants to get lots of wins and gains and uplifts, but they don't want to do the heavy ass conversion research. So remember these things think in terms of processes, not tactics. And the discovery of what matters is the most important part of conversion optimization. And you need to do the heaviest lifting because if you're not putting in the effort in user research and conversion research, it's not gonna work. So thank you so much guys and yeah, I'm sorry for the technical stuff. If you want to learn, I'll give you a quick offer as well. If you guys want to learn more conversion optimization stuff, I have a training, certification, training program. If you want a free month, send me an email to this email address. I'll hook you up with a free month of training. Thank you.