 How's everyone doing today? Good. I am not going to stand at the lectern. I talk. Has anyone ever seen one of my talks before? Couple people. So a couple things. This is interactive. Please, please, please, shout out to answers, raise your hand, tell stories. Oh, I will throw things at you. They laugh because they haven't seen me before. So this is not the talk you probably think it is, but it's talk that I've given literally to over 50,000 people over the years. I have changed it quite a bit and I added stuff from the presidential campaign. So what could go wrong? So, yeah. Who am I? What am I doing? I'm an evangelist for BoldGrid. That's the show. I want to talk about BoldGrid. We can talk out there. I'm also the Treasurer of Open Source Matters Inc. Anyone here know what Open Source Matters is? One person. Okay. Open Source Matters is the foundation of the Joomla project. Oh, How did he get in here? For everybody with pitchforks, Joe used to be on the board too. So You can do that. Here's my thing is I think Open Source is above all else. I love Open Source. I don't care what tool you use. I love WordPress. I built hundreds and hundreds of WordPress sites. I built hundreds and hundreds of Joomla sites. Open Source is first. I don't care what tools that you use because we all should pull the rope in the same direction and put the internet forward no matter what tool set you use. You know, I don't care if he has went to his Mac Linux. It doesn't really matter. So I started with Joomla and I thought I wanted to give back because it will gave me my start. So I should manage it for the Joomla World Conference when there was still a Joomla World Conference. Also the I'm a husband, video game lover, former Disney cast member and owner of a wonderful dog. And you can see her up there second photo on the top. That kind of talks about what I do. Why I do it. I go to events. I go to schools and I talk about WordPress. I talk about internet and I talk about why I love this community so much and why I love open source. And then I ask her a lot of hard rocks and then, you know, that was Paris last year during Matt's talk at WordCamp Europe. Would anyone go into Serbia in two weeks? WordCamp Europe? No, Serbia is beautiful this time of year. So you should definitely check it out. Besides the flights, everything's very affordable. So to get out there. I go to a lot of different types of events. This is Salt Lake City at a mommy blog conference. Those are our professional male dancers dressed as unicorns at a mommy blog conference. I was one of two men at the conference. The other guy was a Lego master builder. So him and I hung out as a while. But I think he enjoyed the dancing a little bit more than I did. That's the official drink out there though. And here's the thing I love about WordPress. On an open source in general, that is very inclusive. This is such an inclusive community. I am so happy that there was a code of conduct when we walked in here because WordPress is just people. And that's what is most powerful for anything I do. This is an open source talk. I have another one of those. But I just love how we all pull the rope in the same direction. We come from different races, religions, creeds, political beliefs, all this stuff. We all can come together for WordPress because it's our passion. We're all not going to agree on every topic and that's okay. But as long as we're all respectful of each other, you know, I think that's what you teach that in elementary school in Canada, right? We don't get that in the U.S. But please, just I just love, really love open source as a whole. And that's kind of why I do this. And that's why I do these things. So that's enough with me getting on the soap box about open source. So now let's talk about A-B testing. So this talk is not a technical talk. I'm not going to give you a lot of plug-ins. I'm not going to give you a lot of things to do. I am going to talk about how I did A-B testing for Fortune 500 companies for years. Some of them are brands you've probably heard of. Actually, all of them are brands you've probably heard of. And talk about the pitfalls and how it's so powerful, what you can do with A-B testing without even having to think about it. But before we get there, can somebody tell me, give me a definition of A-B testing? Anyone tell me what A-B testing and split testing is? Anybody? Yeah, yeah. So you try to do different things. Let's say a picture. You try one picture and try another picture. You see which picture performs the better and then you kind of go from there. Has anyone ever tried any sort of split testing or A-B testing on their projects? Okay. Did you say it went well? Good. Mix of both. Try a lot of things. You don't know why stuff happens. Can somebody give me an example of a time that it went well? No? Anybody? Okay. Yeah. Okay, cool. Awesome. How are you tracking that? We using a tool or did you just change the color and track before and after? Okay. Okay, cool. So the average card value, so yeah. So A-B testing is basically trying two variants of something, sometimes more than two variants, and seeing what performs the best, what performs the best. So that takes us to the thing is A-B testing needs to be controlled. There's not going to be a lot on these slides, so don't get over excited. Most of it is in my head and that's how I can keep the same slide deck for years and just change the content when I, well, as I read about cool stuff. So it has to be controlled, right? You have to make sure it's controlled. I want, you know, and by with that, does anyone here have a PhD in applied data science? No? And your web developers? You might want to go to the next room. I mean, actually I did that once and this guy was like, yep, MIT. And I was like, oh, okay, you let me know if I'm right. So he was very proud of that fact. This actually happened twice, so I've given that. So here's the thing is that there's margins of error with A-B testing. Just like any sort of survey or statistical thing you do, there's margins of error. And unless your test is controlled, you're not going to know what that margin of error is. We'll talk about how you control it in a little bit, but first of all, we have to talk about why your website exists. Has this ever happened to anybody? Oh, actually I should ask, who makes websites for other people? Most of them. Who makes websites just for yourself? Okay, good amount of people. And some crossover there too. Now, has this ever happened with a client? Or heck, if you're with your own site. The client, you launch a site, honeymoon phase. Site goes live. You're the best person in the world. You're a superman. Site's amazing. You call them six months later to do an update to WordPress or something else. And they're like, hey, WordPress 4.9's coming out. We should probably update whatever it is. Well, no, I don't know. The site's not doing enough. Not what we expected. The honeymoon phase is over. And it ends quick, right? And has that ever happened to anybody? That the client just turns over in a couple of months and you have no idea. Like why? You love the site a while ago, but now you hate it. Now it's not doing it. You just look at it as a waste of money. That happens all the time. And it happens for one very simple reason. One reason. We don't ask the clients why. We assume why their sites exist. We assume their KPI, their key performance indicators. By the way, if I say an acronym and you don't know what it is, throw something at me. I will say the full thing. We assume what it is. Great example. I once met someone and it was like, okay, hey, I have a utility. And they told me they care most about bounce rate and time on site. So I did all this stuff to lower the bounce rate and increase the time on site. They didn't ask the second question. Why do you care about that? You know what they cared about? Because their time on site went up and their bounce rate went down. The client wasn't happy. The client wanted the opposite. They wanted bounce rate as high as possible and the time on site to be as low as possible. They wanted their site to be so optimized it would find the FAQ support page for the question the person was looking for and not have to navigate to get the answer. Because their utility, that's what they cared about. I've talked to people that have done user, have done things for different provinces here in Canada and how to get information to relocate to that province. And the province actually wanted to add hurdles to get the information. Like I just said, an email and do all that because they're like, we want the people to prove they're worthy to come to our province. I won't say the province. And you have to ask the client, what is the goal? Does anyone here at every project ask what the goal of the project is? Okay. And for the five or seven people that raise their hands, how many times you write that down in a non-subjective phrase? Can you a year from now look back at a project and say, with data, the project was a success or not? Can anyone here do that? No one's, okay, a couple of people are raising their hands. That's great. Most people don't ask the question. The client says, well, if it's an e-commerce site, I want to sell more. I want to sell more. What does more mean? You have to put an uppers on it. I tell a client when I was doing freelance stuff, well, actually I was in freelancing when I was working in an agency in downtown Minneapolis, because I'm from Minnesota by the way. If we can do this by this date, will this be a success? And then you add stretch goals in there. If you don't lay out exactly why this project exists, you are going to lose clients. Period. You're going to lose clients. If you don't write down day one why the site exists, with quantifiable yes or no answers, you will be fired left and right. Because the clients have a different opinion of what they think they want to what they really want. Clients think they know the answer. Has anyone had a client that has self-prescribed something? This probably happens all the while. I want WordPress. They call you to say I want WordPress. That is not the best answer. Because maybe WordPress is a good option for them. Maybe it's not. It probably is. But it really depends on their use case. And you have to do that. So you have to be controlled. And what I mean by controlled is use a tool. And use a tool like Optimizely or VWO, or there's plugins out there that will tell you if the testing you're doing is within the margin of error or not. And then it will say yes, this test was a success. So this test was a success. Because just because sales go up doesn't mean that that AB test was a success. You've got to compare it to multiple factors. You've got to compare it to how many users were. Because you're only getting 100 visitors, two sales can skew your data, quite frankly. If you're getting millions of visitors, that's less of an issue. But you also have to look at year over year. I had a client. They're like, oh, I did this change and my sales went up 40% month over month. Okay, you're an e-commerce store to compare in October to November. What happens in November? Black Friday, exactly. Black Friday. And we compare it year over year and their sales went down. You have to make sure your data makes sense. And there's tools that will do it for you. And this talk is nothing more than very basic ideas on how to make money. Because AB testing is the number one way you can have residual income for forever. And not really have to fight it. Where did my, there it is. And not really have to fight it. The cool part about it, excuse me, is it's so easy and it's fun. You can have so much fun playing with someone else's brand. I'm going to go over some real examples in a little bit. This I just talk is to give you ideas and actionable things and how you can make more money. And with that, here's the secret. Here's the secret of web designers. Most people that make websites, they make a website, they leave it alone for two and a half years, three years, and they start over. As a client of a colleague said, oh, we did our site four years ago and now we need a new one. And then they don't touch it again because it takes seven months and it's such a tedious process. And clients hate dealing with web designers. I mean, they have to give content and write stuff. And they have their job to deal with. They don't want to deal with that. That's why they're hiring us. And then they say they're going to blog. They say they're going to do it, but they don't. And then the site sits for two to four years and then they get another call. Usually they call somebody else. Let's be honest. Sometimes they don't. And they want to do it all over again. It's a definition of insanity. Here's the secret that web designers don't tell their clients. Here's the secret. We're all guessing. Every site that goes on the internet is a guess. Now, here's where people are like, what a minute. I'm a professional, Mike. I don't know about you demo if that is your real name. It is, by the way. But I'm a professional. I know best practices and I know responsiveness. And I know all these other buzzwords that apparently are check boxes that I have to check off. Every set of clients are different. If you think you know the users of your client, most likely you don't have a full enough picture as you think you do. And I guarantee your clients don't know who's visiting their site. I've had so many times clients like, I had one client. It was a food client. They sell food product. It's like, you know who comes to us college students? And sororities and fraternities. That is number one who buys from our stuff. And we're like, okay. So we started to do advertising. Gear did that demographic, BMX, stuff like that, whatever. You know, whatever the kids are into. I feel old. And nothing was selling. It was a food product. You know who was visiting. We looked at the data. We started doing some user surveys. You know who was visiting in the site? Single family mothers, single family homes. Because it was an all natural product. So we started shifting to the marketing that way. We did AB test that way. And guess what? Seal Skyrocketed. You know why they thought it was college students? Because the marketing manager just graduated from college. You laugh, but this happens all the time. This happens all the time. People think they know who are visiting their site. They don't. You have to test it and use AB testing to get you there. The cool part about this is you can sell residual income from day one. I've had tons of clients. I've sat down and said, hey, you will never have to do another website again. This will be the last website you have to make in your entire life. Because if you sell a monthly residual income AB testing or whatever you want to call it, the site's going to evolve. It's going to change. You're going to use data to make it better. We will make best guesses based on our knowledge and expertise and what we know, but stuff's going to be wrong. It's going to be wrong. It's not going to be perfect day one. And only data and AB testing is going to get you to where you need to go and to make that site go from minimum performance all the way up to something really amazing. So with that, it needs to be statistically relevant. So I talked about the margin of everything. Use a tool like VWO, Optimize Lead as Google has one too. I'm not here to say one tool is better than the other. That's not this role. I'm here to talk high concept stuff. And it will actually tell you if the test is a winner and implemented or if you need to keep testing or if the test is not a winner and just to kill it. So you can do all that stuff. Micro testing. I recommend only do micro testing. What I mean by that is only testing one varied at a time. The reason I say that is a lot of times people are like, I'm going to do AB testing and I'm going to have variation one, have a new photo, a new headline, a new picture of the button. And then I'm going to have variation two, have two photos and two buttons. And variation three, I'm going to have no buttons and no photos and no content. And we'll see which one performs better. But really, you have to have a lot of traffic. To do multi variant testing, really, in my opinion, you have to have at least 100,000 visitors a month. And that's a lot for a lot of sites. So if you test one at a time, it might take you longer to get the results you want. You test individual elements and then you can test the winners of the individual elements grouped together. Against your original. And then you can see if that overall gives you the list. Because just an image that wins and a headline that wins and a button that wins individually might not win together. So do some micro testing. Oh, what to test? Can anyone give me any ideas of what you could test on a website? Because I get a lot of people saying, I don't know what to test on my website. Anyone have any ideas of what you could test on a website? AB test? Navigation. Navigation. Oh, first paint, like speed and things like that. Yep. Headlines. Copy. Call to actions. Value proposition. Yeah. Font. Text. Excellent. Color. Yep. So there's so many things you can test. We'll go over some of them. Buttons. There are so many things you can do with buttons. The color of the button, the size of the button. I once had a national insurance company. National insurance company got an 80% lift. We did usability testing. We did focus groups. We did all the research we could think of. But I always recommend trying a crazy test. And I'll give you some examples of crazy tests. Tests that you think will never work. Because sometimes those are the winners. We tried everything. And I couldn't get any lift, like 4%, 2%, whatever. Tried everything. I tried a two-pixel drop shadow. A two-pixel drop shadow on a button. Nothing else changed. 80% more sales. A two-pixel drop shadow we added. I didn't believe it. I ran 4 million people through that test. Every time. Different pages. Every time. Consistently. Now, before we all run out and add two-pixel drop shadows to everything, every setting of users is very different. So you've got to test your users and your sites. I mean so many people say, well, according to this case study, that original photography is better than stock photography. So I'm going to use only original photography. Because you use photographers in the room. And then it's like, no, this case study over here, it's a stock photography works better. There's no right answer. It all depends on the users that are visiting a site, because you're not building for you. You're not building for your client. You're building for the users. And we forget that all the time, because you're trying to service your client, right? So buttons. Text. What the text says is called action, the font, size, style, case. So many different things you can do with text. I mean, heck, I've run AB test with lorem ipsum on it, that I performed better than the original copy. It's true. If you ever want a fun test and your clients are letting you do it, I recommend a ipsum called corporate ipsum, because it looks so correct. And especially in block content, no one's ever going to read it. And it just feels at the page very nicely, and it has the right words like synergy and balance and things like that. So yeah, these are all real ideas that I've done. Images. So many different things with images. Here's what I ask. Who only uses original photography? Okay, one person. Are you a photographer? No. No, okay. Usually it's the photographer in the room who's like, yep, and I sell it. Okay, who uses stock images? Good amount. There were some people that stock images performed really well on. There were other ones where stock images performed really poorly on. There were some user groups that images at all don't perform well. I've seen websites that have grossed five million dollars in a week with stick figures drawn with crayon. It all depends on the users. And the only way you're going to get there, now they tested a bunch of stuff. I know the owner of that site, they tested hundreds of things to get to that point. So whatever you do at the first time isn't going to work. You need to make it better. Everything. You can test everything. I hear people all the time say there's nothing to test on this site. I don't know. You can test navigation. You can test, you know, if the sidebar's on the left or the right and all that stuff. And a lot of times people are like, well, how do we do this? I'm not a coder. But what you can do is you can use a tool like VWO, which is what I use. It's not, it's just what I use. There are other options out there. There are WordPress plugins. I like to do it on a third-party service for this simple reason. The reason I use VWO is it allows me to make the changes and have my marketing team make the changes without needing to know any development. And it doesn't do it on the source because the worst thing you can have happen is Google index a test. And unless you're very good with robot files and things like that, which I'm not saying you're not, but in my opinion, it's just too easy to break and have a test get indexed. Because you don't want production, a test to get indexed on production until it's ready to go to production, right? So VWO, what it does is it puts a JavaScript into your theme file and it will change it on the user browser side, the agent side, so that all the changes are being served on only the browser level. Now, there is some studies that say on slower connections, this gives you false positives. Simply because on slower connections, you see that CSS pop on slower. So then that element might change a color or something, then that person might notice it. It's less of an issue now with higher speed connections, but something to keep in mind. And if you're getting a lot of people that are on lower speed connections, you might want to try a different tool. It doesn't use that tactic. But you can test everything. And for those of you who think, well, that's fine for your Fortune 500s, but my clients are special. My clients, they care about their brand. They care about design. You know, there aren't majors at heart and all that stuff. Well, here's what I say to that. Yes, all these tests and crazy ideas I'm about to show you, please do run by your clients first. Don't put any of these on without asking. At least don't do it and tell me. You know, you have to make that. But you can get stuff. I've done things that you would think never would pass a Fortune 500. Done with Fortune 500s and they've approved it. Because I say it's only 5% of your audience and I test about 5% of the traffic. You might need to have that be higher for you guys based on your traffic. These tools we usually recommend one. I'm like, it's only 5%. And if we learn something, we learn something. Because at the end of the day, they care more about their goals, their bottom line. Again, we're going back to day one. What's the purpose of the site? So here's where I talk about the presidential campaign. So let's talk about Donald Trump first. It gets dangerous and I'm in Canada, so. So. Where do you draw the line? Like the two pixel drop? How do you know where to stop? It depends how much traffic you have. If you have a little high traffic site, you can get a lot more lenient on it. I always recommend doing four tests based on your experience, based on what you see in the analytics and what you think will be best for that user group. Do tests that you think will work best. But try one crazy one. I give a list of crazy ideas at the end. And then most of the time, the crazy idea will fail, like you expected to. But the two pixel was my crazy idea at that time. But the other ones are legitimate ones. Like you know what? I think us getting original photography would work really well. So we'll try that on this landing page. So I recommend if you're doing five tests a month, that's what I recommend for agencies to sell their clients. Because most clients can afford five tests a month. You know, that's probably about 40 billable hours, depending on your agency. And most clients will be willing to swallow 20 to 40 billable hours a month. If you sell it to them as, this will be, you'll be paying a residual income, but you'll never have to do a new site again in your life. Visual website optimizer. So let's talk about Donald Trump. One of his most successful, now it's no surprise that political campaigns, especially in the U.S., they do A.B. testing to maximize their donations, right? Everyone does it. Every party. I'm not saying it's bad or good, I'm just saying it's what it is, right? I know, because there's companies that specialize in that. I used to work for one. Donald Trump's campaign, a week before the election, one of his most successful fundraising efforts ever. He did a thing that said, okay, donate money, you'll get your name on our Facebook page live or our website live. If you were going to do that, assuming you took the job, just suspension of disbelief now, and how would you accomplish that from a technical perspective? People donate money, their name goes on a page. That's like a thank you to Josie from New York, whatever. How would you accomplish that? Anyone have any ideas how you would accomplish that? Yep. Like a third party service, there's ones out there that will like zoom in on the map and fly video and all this cool technical stuff. What they did is for seven hours, the week before the election, they had a live webcam of a laser printer. This isn't a political action committee. This isn't third party. This is the official Donald J. Trump campaign. Now, regardless of what you think about our current president, I think we can all agree he cares about his brand. Right? I think that's fair. No matter where you are in the spectrum, I think it's fair that Trump cares about his brand, right? And they did it. And it was one of the most successful fundraising campaigns. And for people that are sitting there thinking, like, oh, they don't know technology, haha. If you think this wasn't deliberate, you're far mistaken. They'll use a group that they were targeting this connected with them. They tried the Polish approach before, doing it before the convention, super high produced video type of thing. Didn't work as well based on what we know based on FTC filings. We can assume things and stuff like that. Hillary did A.B. testing too, right? She tried different giveaways. You know, when more people donate if I offer you a copy of my book or a sticker and things like that. And she was trying different scenarios with that. And my whole point to this is no brand is too holy to do these concepts on. Maybe not to this extreme, but you can make baby steps. That's why I recommend one crazy test to four real tests. So what are the crazy tests that I do? This is literally the checklist I use. I'm going to go through it real quick. Professional versus unprofessional. There are times unprofessional converts better. Anyone here get one of those five page long letters asking you to donate to save the unicorns or something as like writing in the margins and things like that. Those long form like solicitation mails. As a reason why those exist. They're less effective than they used to be, but they're still effective. The companies that send them out still send them out because they work. Sometimes unpolished stuff just converts better for certain types of people because you got different education and you have different backgrounds and you have different interests and you have different devices, right? You have, you know, so many people are so focused on the desktop experience, but the fastest growing market on the internet is mobile only. This is a developing world, but depending on what your client is, that might be very important. Imagine you're making a website for the human rights campaign. You might need to develop a website that is text only so that people in the developing nations can access that content. Maybe not important to you, but it's worth thinking about. Photo versus illustration. If you don't have illustrations, they have stock illustrations out there. They work. One color versus another color. Colors are so powerful, as we heard in our button example over here. And one of the things is test accessible colors. You can look up the color contrast that are accessible or not. If you're not sure what those are, you can look up the WCAG, WCAG, 2.0, AA and AAA. It'll give you the mathematical numbers right in there for the percentages of the ratios. And if you test an accessible versus a non-accessible color, most of the time, accessible colors will win out simply because one in five adults in North America suffer from a disability related to color. One in five. Low brightness contrast versus high brightness contrast, border versus no border. Clear image versus a blurry image. Static versus animated. Imagine the car was going. It's supposed to be animated. Ugly versus sex appeal. Some people really respond well to super produced model photos. Other people perform really badly with highly produced photos, and they're looking for people that look like them. It is not all the same. I guarantee you the only way you're going to know this is to test. And that's why I care most about data-driven decisions. I don't care what something looks like at the end of the day. I don't. I care about servicing the client because I guarantee you if the site's ugly but the client's making bank, they won't care. Now, that's an extreme. You can test these elements in a vacuum. I am not saying to all of a sudden get rid of your beautiful theme and to have only text and no pictures on your website, but test these elements individually and see if they perform. Even if the test doesn't succeed, you'll learn something. So I always recommend do four tests, five tests, four that you think will perform based on your experience and your wish list. When you meet with that client, you write down every feature you think they want or that you want, and then you do test with them to prioritize based on most impact, based on your experience as a professional. Static versus interactive, get 50% off, or you can choose if you want to get 50% off the pizza or the doughnut. There's not all of the above options, unfortunately, but... Professional stock photo versus amateur informal photo. Some of these perform way better than the others. And here's where we talk about the title of the talk. A.B. Testing, which way does the duck face? Anyone came here because of the duck title? No, nobody's not honest. Oh, I got one. There we go. I have a sequel to this talk called What the Duck, by the way. So there was a test, I think, by HubSpot. They had a picture of a right facing duck and a left facing duck. And the right facing duck converted five times more than a left facing duck. Can anyone tell me why that would be? Any ideas? Duck facing towards the text, possibly. Right hand versus left handed, like how you read. Okay, right brain, left brain, sometimes like the direction of the text, people also say. Any other ideas? Yep. Arrow to the point of action. Every fifth five times I do this talk, somebody raises their hand and say, because the duck is looking into the future. Here's the answer. Here's the true answer. It doesn't matter. It doesn't matter. Seriously, why? I'm sure you can spend all day wrapping your brain about why are two pixel drops. Sorry, I'm a little hang up on the two pixel drop shadow experience I had in my life. But you can spend all day focusing on that two pixel drop shadow or that duck, or you can test it and move on. If the test proves out and you can prove it with data and it's past the margin of error and you run it again, maybe on a different page and does it again. If it's making the needle move, that's all that matters. This is how I explain AB Test to a client. Your website is a magic money machine. These are the exact words I use. I'm not at all making this up. Your website is a magic money machine. If I could give you $2 for every dollar you gave me for that magic money machine, how many dollars would you give me? They laugh and I say, no, seriously. How many dollars would you give me? If I can prove to you, this machine makes money. And you can always cancel any month. We'll go month to month. And they say, well, unlimited dollars. I say, well, let's start with 2000. And every month, every year they increase because they see the data and it's based on the goals. And if you can prove to them that their goals are happening, looks fade, the honeymoon period on a nice looking site end like that. Thank you. Now, there are times when they care where there are different KPIs. Maybe they care about time on site or engagement or whatever their goals are. Your testing can help you get there. One layout versus another layout. One call to action versus another call of action. Images versus text only. Believe it or not, that does work for some people. Upright versus angled button versus blue underlined text. Red Laborta versus a false board and drop shadow. Custom ad design versus familial UI elements. I recommend when I say that, I don't say make your website look like Facebook or Twitter because that's, I don't know what the word is, fraud. But you can use some of the ideas, right? People started noticing that the Facebook app is now doing the bottom menu on some phones or testing that. And now some websites are starting to mimic the bottom menu because the top left hand figure is out of the reach of the thumb. So the elements like that is kind of what I mean by that. A visual image or some air image, there's our friend, the duck again. One font size style case versus another font size style case. A standard ad shape like a rectangle or a custom ad shape like a circle or in this day's day and age, a standard ad shape like a circle or a custom ad shape like a rectangle. Positive versus negative. Want to fall in love with your spouse again versus sick of fighting with your spouse? I once made a website for a product that stopped stow fires before elderly people. You know what worked? Your mom's going to leave the stove unattended and it's going to burn down. She's going to die. It's going to be your fault. We tested other things too, but if it works, it works. Come on, click. Click. Generic versus relevant to time period. Want a girlfriend? Or are you home alone on a Friday night? Or do you want a girlfriend for Christmas? It is Friday night and it's almost Christmas and yeah, I would like a girlfriend. Sorry, honey. Mentioned prices or discounts. Regular $20 or leave it a mystery. Add your card. Anyone here see you have to add to the card to see the price or something like that? You know, the prices, you have to add it to the card to see the price. That's the old version of, oh, Bob's Appliance Mart. Our deals are so crazy. That you have to come on down to the showroom. Same thing with online. Sometimes there's legal things that you can't show a price unless a customer has an intent to buy. Most of the time it's just marketing. No trust logos versus trust logo. It's like Better Business Bureau. I don't know if there's a Canadian version of the BBB up here, if it's also the BBB. You know, as seen on ABC, CNN, whatever. Anyone here watch America's Got Talent? Okay, I don't know. Yeah, so the most common people that use as seen on America's Got Talent are the losers in the first round, because when you're at your local comedy club and you see like, oh, Joe Smoh, as seen on America's Got Talent, and the season lasts like 12 years, so you're not going to remember what happens in the beginning. You're like, well, they must be good. But if actually you type in Joe Smoh, America's Got Talent, they were laughed off by the judges in the first round. But trust logos work for certain people because it adds validity. Weapons of influence go through these really quickly. Consistency, scarcity, liking, authority, social proof, you know, free trial or risk-free trial. I used to sell timeshare. This is risk-free. It's not free, it's risk-free. You know, things like that. Appeal to the eight universal desires. You know, companionship, enjoyment of food and beverages, comfortable living conditions, care and protection of loved ones, social approval, survival and enjoyment of life. Features, benefits, not features. Nobody buys on logic. Nobody buys on logic. I don't care if you're buying anything from a widget. I used to do A.B. testing with people that sold rotary parts for NASA all the way up to, you know, fine art. People buy on emotion. They justify it with logic. Even if it's a mechanical part, the emotion is, okay, this company has it taken care of. I'm going to look good to my boss, things like that. So focus on emotions when you do any A.B. testing, especially if it calls to actions and for the pictures. And playing it have fun. You can have so much fun with A.B. testing with somebody else's brand. You have no idea how satisfying it is to throw an ugly version of your client's website up on the internet. It's kind of fun. Of course, when these all buy your clients and all that, I only test like a small subset, like 5%. I got that from a book called Russell Brunson. He owns a product called ClickFunnels. He started from the supplement industry and he has some good training out there. I'll show you how to get a copy of that book, but I warn you, his stuff is focused on the direct response world. Supplements, coaching and things like that. So start with the free book before you buy any of the other stuff because a lot of people are like, well, I sell digital marketing. This is talking about how to sell weight loss pills. Concepts are concepts. You can test it. So here's a cool test called 108 Split Test Winners. So you go to .comsecretslabs.com slash free dash book. My slides are on slide change so you can get this later. You don't have to write it down right now, but I'll walk you quickly how to buy it without spending money. You go to this page, you fill out the form, fine. You need to pay for shipping. You get an offer. You say no thanks to the offer. You hit another page with another offer at the bottom, you say no thanks. You hit another page with another offer at the bottom, you say no thanks. You hit another page with some more offers. You close the tab and then you get an email and you can unsubscribe from that email if you'd like. I want to be clear. I have bought hundreds and hundreds of dollars of his products. I think they're really good, but too many people can't mentally get past the examples of direct response like supplements and online coaching. So try the free thing. Pay the $8 shipping and see if his coaching style works for you because I've got in so many ideas for Fortune 500s from his training materials that I'm not an affiliate or anything. It's just I really like his stuff, but I recommend start with the free and I'd walk you through that because his sales cycle is so good you're not realizing you're buying something for $800 even though that's what those pages were offering you. Upwards of $2,500 sometimes. And his stuff is I think worth it, but start with the free one just like anything. And you can unsubscribe and if you don't want his emails, but I keep his emails because I find him interesting and see what he's doing and then I try to mimic it with that. So with that, we have five minutes. So I have time for questions, comments or death threats, depending on how many designers are in the room. So yeah. I use a tool called VWO and I use the JavaScript that they offer to change the test to the end user in the browser side. So the production site never changes until we have a winner. And then in that tool, I can say I want 5% of the traffic to see this because I don't do side by sides because unless you're very careful with URLs and robots and stuff, Google can index the test. So that's why I let a third-party service. It's not a free tool, but I think it's worth $99 or whatever they're charging now for 20,000 tests or something. So there's other ones out there. The Google has a free one. It's like Google testing, Google it and probably find it. Optimize Lee's another one. And there's some native WordPress plugins, but again, it's superstition on my part. I prefer not to touch production until I'm ready to apply the winner. So any other comments or anything? You can test with any traffic. It's just going to take you a lot longer to get past the margin of error. Ideally, five to 10,000 users a month is what you're looking for, but there's shortcuts you can do, right? If you're looking to test a headline, a way to get around it is to test the headline in a Facebook ad. It might not apply to your website, but if you test two headlines or two photos in a Facebook ad and you run 1,000 people through each one, cost you 25 bucks, you can get the ideas and at least it gets started. So that might not apply to your website, but a lot of times it does. It's kind of a hacky way around it. So the tools will tell you based on its algorithm because there's a whole algorithm. I found ones online that you can punch into your own numbers, but that's algebra and no. It's not what I'm into. So I literally will just tell you there's three settings. I want fast learnings, accurate learnings are balanced. Depending on the traffic, I'll choose one of those and it'll tell you test as a winner, kill the test, apply it, test as a loser, kill the test. No, or we're not sure, keep testing. So that's why I like VWO because it's just simple, red, yellow, green. Anyone else? No? Cool. Well, thank you so much. My name is Mike Demer.