Latest / Talk Commerce / ADA Compliance and Web Accessibility for Ecommerce with Michael Bervell of TestParty
Transcript
- Brent Peterson (00:03.086) Welcome to this episode of Talk Commerce. Today I have Michael Bervell He is the CEO, founder of Test Party. Michael, go ahead, do a much better introduction than I did. Correct me for all my mistakes and tell me your day-to-day role and maybe one of your passions in life. Michael Bervell (00:20.616) Yeah, looks good to be here, everyone. Thank you so much for having me, Brent. Yeah, I'm Michael Bervell, founder CEO of TestParty. I guess you got all that right. It's all good. One of my passions, as of lately, I've already started my New Year's resolution for next year. I always start mine on December 1st just to get the habit built by the time the new year actually happens. And my resolution for this year is to do the Bradbury Challenge. Brent Peterson (00:31.116) All right, good. Michael Bervell (00:50.184) I don't know if you've heard of this before, if your audience has heard of it, but it's from the author Ray Bradbury, who's like, he wrote all these amazing books, lots of like science fiction stuff. He was like a science fiction writer. So if you've ever read, you know, any of his stuff, you know, they have not seen the stars or something wicked that way, you know, any of those sorts of books, he wrote a lot of that stuff. Anyways, I'm doing the Bradbury challenge. So I read one poem a day, one short story a day. and one essay a day. And then I am trying to write also one poem a day. So that's kind of my fun. That's what's been bringing me joy. That's what I do for fun lately. Brent Peterson (01:29.014) Yeah, that's awesome. Good. Yeah, we did the hard 75 as a family running up to Thanksgiving. I was the biggest offender. We ended up adding eight days. it ended up being the hard 83. And that was hard. Anyways, yeah, it was hard. All right. Michael Bervell (01:45.578) no. Michael Bervell (01:49.547) Well, that one's, that's like physical, you know, I feel like for me it's the mental stuff. Okay, I'll just read a poem a day and a short story. It's like almost more fun versus like, yeah, you know, run 83 miles. Brent Peterson (01:58.092) Yeah. Brent Peterson (02:02.393) Well, we had to do like a diet and, you know, my kids are spreadsheet kids. we had to read nonfiction every day, 10 pages of nonfiction. All right. So before we get started, we're going to talk about accessibility and testing for that. I do have a joke for you. We do the free joke project. I'm going to tell you a joke. You give me a rating, eight through 13. So here we go. Michael Bervell (02:14.366) Yeah, I love it. I love it. Michael Bervell (02:29.674) There we go. Brent Peterson (02:30.24) I can't believe that viruses and bacteria would just invade my body without a permission. That makes me sick. Michael Bervell (02:39.85) You know, Brent was good. That was good because it was clever and I do like puns. So I'll give it an 11 and a half. Brent Peterson (02:49.998) Good, thank you. Yeah, lot of people just like get the joke later in the day. They're thinking jokes usually. Appreciate it. Yeah. Michael Bervell (02:57.139) it's interesting. You know, it's the Bradbury challenge is keeping me on my toes, you know. Brent Peterson (03:01.538) There you go, perfect. All right, so tell us, let's just do the 10,000 foot view of accessibility. I mean, there's a whole group of people in the world that can't access websites for whatever reason. So tell us kind of how you got into what you're doing and really the importance of it. Michael Bervell (03:22.868) Yeah, I accessibility is a subset of what I would call more broadly product inclusion. Like the idea of when you're building a digital product, what's the default usage pattern? And I think if we think of products today that have launched in the last two or three years, a lot of them are AI native, and they don't have, a guidebook or a rulebook. So think of the launch of ChatGPT. It's very different than the launch of Blu-ray because there was no instructions on how to use ChatGPT. It's emergent. That's technology. You use it and you start to discover what it can do. And so when I first got into product inclusion as a field, it was in the era before AI, right? Where everything was very defined. And so was using the Xbox Kinect, which was built to be this like device that you connect to your Xbox and you can use your body to control it. And I remember it was like sixth grade. I got it as a Christmas gift from my parents. I was like so excited with my brother to try it out at like 8 p.m. and we plugged it in and it didn't recognize our skin tone. In Seattle, 6.30 p.m., 7 p.m., it's later. It just wasn't working. And that's kind of what got me thinking about product inclusion. Eventually I went to college, I went to Harvard to study philosophy and CS, and my focus was another segment of product inclusion, which was privacy and GDPR compliance. So was writing my senior thesis in 2018, 2019 about Cambridge Analytica and the kind of election. I guess you would call it election plus social media plus advertising scandals and how that kind of led into a lot of GDPR compliance. What cookies can you track? How can you track a person? And finally, the third leg of the stool that led me to kind of create test party and accessibility was meeting my co-founder who, like me, had a background in product inclusion and product design and software engineering. And he went to go work at Amazon when Amazon was hit with an ADA lawsuit. And he was like... wow, there's no automation here, there's no technology to fix automation for accessibility, maybe we should build that. So those kind of three legs to the stool, that's kind of how I first got involved and engaged from very academic lens in college with my senior thesis to like the very personal lens of using a product that didn't actually work, to then thinking of the actual like, what's the business opportunity here in talking to a friend? And I will say like, I didn't expect to be doing accessibility. My, before this, I was in business school before that. Michael Bervell (05:46.876) I was a venture capitalist at Microsoft. We were writing a billion dollar checks to people all over the world to go build their own startups. So it really was a matter of opportunity and also passion that drove me into this. Brent Peterson (05:59.468) Yeah, and so I mean, I think so you're focused on more than just websites. It's just any product that is digital or even physical products. Michael Bervell (06:10.195) Yeah, it's interesting. Definitely there's a need for it, like physical products, AR, VR, it's another interesting space that we think about a little bit. But we're primarily like websites and we've seen a lot of interesting traction, specifically in e-commerce. Like all these brands built on Shopify, WooCommerce, WordPress are getting hit with ADA lawsuits, which in the physical world, you know, it makes sense. You go into a bathroom, the sink is too high. For a wheelchair user, you understand that, right? You try to get into a building and there's... only stairs and no elevators or ramps, you understand that. think a lot of online business owners don't even know that there are online standards for accessibility in the same way that there are in the physical world. And so that's kind of how, when I think about kind of the space and I think about accessibility, like we focus on websites, but certainly there's a need beyond that. Just not our focus. Brent Peterson (07:02.39) And from the website standpoint, probably the, and I'll make it, I'll just guess here, but the primary focus is just maybe site or visually impaired, or is there other things that you would look at as well for websites? Michael Bervell (07:17.865) I mean, definitely vision, physical impairment, can even think of cognitive impairment. I mean, one of the best ways that I kind of conceptualized this wasn't through examples of how a site might change, but it was through how a person might change over the course of their life. And this was big when I was at Microsoft, because I did help design some of the inclusive design toolkits as a consultant at Microsoft and at Google. But one of the things that they used in the example was Let's suppose that you're a pregnant woman, right? Eventually you have your baby and you're carrying your baby around with one arm, right? That's a temporary quote unquote disability where you only have one arm versus two, right? And this idea that one's physical ability is fluid throughout one's life. Sometimes it's situational. So you might have to hold a baby, maybe you're pushing a shopping cart so you can't use both of your hands, right? Maybe for whatever reason. You broke a toe, you broke a leg in your crutches. These are all temporary disabilities that affect one's interaction with the world. You can think of it in the kind of tech world. One of my friends got Lasik two months ago, and he was like, can no longer look at my phone. So you just couldn't do anything for three to four days. Is there a world where technology was built more inclusively or more globally that it's very intuitive how one would use? voice navigation without having to look at a phone after eye surgery to actually use the device. And so that's kind of what we think about a lot more is like, obviously the use case is people with disabilities who might be blind, physically impaired, mentally impaired, whatever it is, whether permanent or temporary. But I think where the industry is going and where I see not just e-commerce, but every business going is trying to think through how can this sort of innovation, which traditionally is at the edge of the bell curves, how can that benefit the vast majority? in the center of the bell curve. So that's a big part of where I focus my energy and attention. Brent Peterson (09:20.27) Yeah, and I think the, especially the visually impaired, it is more than just the bell curve. It could be 10 or 20 % of the population that has some sort of visual impairment. Maybe it's nighttime vision or just, yeah, any type of near-sightedness. It is a larger segment of our population than a lot of people think, right? Michael Bervell (09:46.858) Yeah, and what's so interesting, and I guess this is kind of more philosophical, bear with me. It's kind of the thought experiment of what I studied in school. But a big part of it is like, how does one know what average is? Right? Like, you know, might say I have average vision, and somewhere at some point, average vision was defined as 2020. And if you're better than 2020, you're beyond average. And if you are at a certain threshold, you can call yourself low vision. And at a certain threshold, you might need glasses. But it's interesting because visions is one of those areas where we've defined what the standard is. But there's so many other abilities or just things that affect people that there is no barometer of what's typical. Oftentimes people don't even know if they're typical or non-typical until after the fact. After the fact being maybe some sort of diagnosis or some sort of trigger or whatever it might be. So yeah, definitely it's super interesting. And for visual impairments like... what you just mentioned, that's a big example and then some of these other questions as well. Brent Peterson (10:50.606) Yeah, and I think, so on your website, you have some compliance issues and there's some government compliance and there's certainly compliance from the EU and things like that. How important is it to, and are they a good baseline to make sure we're doing what we need to do? Michael Bervell (11:10.535) Yeah, it's interesting. All these standards, like there's the Americans with Disabilities Act in the US. And if you don't comply, you're at risk of like legal lawsuits. It's fun side note, fun or unfun depending on how you see it. But a lot of these lawsuits are done by like kind of drive by lawyers, which is kind of trolls. And that term drive by comes from the idea that, you know, in the 90s, someone would drive by a building and if they didn't have a ramp, they would know to sue the business. Right. And so the equivalent is happening online where people Sometimes with impairments who don't really want to even use the product will drive by the website and see that there's issues and file lawsuits. So that's the ADA in the US, the European Accessibility Act, there's section 508, which is mostly government. All of this really rolls up to a set of standards called the Web Content Accessibility Guidelines, or the WCAG. So in theory, if you really know the WCAG and how to implement it, it should be really easy to make your site accessible. at least on a case-by-case basis for whatever impairments someone has. But the standards are there and defined, and it's just a matter of really following it trying to follow it with ease. Brent Peterson (12:18.188) Is it, so as a developer, it easy to, if you're, let's just say you're a coding developer and you're not necessarily thinking of how the website's gonna look and, you know, I'm a big cloud code user, so cloud code doesn't see what it's making typically. You have to kind of describe it or give it a screenshot. Is there a good way as a developer agency to go about making sure that you're compliant as you're building? Michael Bervell (12:48.39) Yeah, I mean, it's interesting because I use AI tools all the time. I use cloud code like no one use business. And what's funny is there's a school of accessibility practitioners who are doing essentially like LLM challenges, trying to see if you just described an LLM, build a web page does XYZ. How good at it is building, is it at building an accessible web page or an accessible form where you can submit information. And typically we find that these LLMs, because you're trained largely on inaccessible website data. like 97 % of the website, what the web is not accessible, they spit out inaccessible code, right? That said, like it's definitely possible to prompt them to be accessible. But I think the best solution is really to use a more like deterministic tool. Like one is called Wave by the University of Utah, released that one. Another one is called Axe Core, which is like the biggest open source engine, right? It's what like Google Lighthouse is built on Google's accessibility engine. Microsoft has one called Accessibility Insights that's also built on Axe Core. There are tools out there like Power Mapper. Actually, interestingly enough, these Power Mappers actually used by a lot of these troll lawyers when they're scanning a website to then file lawsuits. So you could argue for developers, if you're scanning with all the open source tools and making sure that your website at least complies or shows zero issues on those open source issues, that's a great baseline to protect yourself legally. And then on top of that, these automated tools scan for maybe 60 % of issues. So that last 40%, you do have to do manual testing or user testing. And that's where kind of the product inclusion more broadly comes in. Have you designed your product to be inclusive beyond just automated tests, but also through your own manual testing? Brent Peterson (14:36.194) Yeah, so let's talk a little bit about test party. So you've brought this in and made it and added more automation to the accessibility testing or explain kind of your solution. Michael Bervell (14:46.524) You know, it's so funny is that we call the company test party, but we probably should have called it fixed party. It's like our, biggest differentiation in the industry is that we don't just test. We also fix things and, and we now have like close to 55 e-commerce clients and they're doing collectively close to 2 billion in revenue online and offline. And when we chat with most of these clients, they have audits and they've done audits, you know, once a year, once every two years. paid 20, 30, $40,000 for these audits, but then they don't know what to do with the results. And so the true problem that we're trying to solve is, you've done an audit with Wave, PowerMapper, Axe, you've used Level Access, Deque, these are all the major players in the space, but what do you do with those results when you've never been trained on accessibility? And so we're building kind of a cloud code-like or a Tachypt codex-like engine that can take in these results. and automatically fix an e-commerce site or fix a website to be compliant at the code level. That's our big innovation in this space and that's what investors have been backing now for two and a half years. Brent Peterson (15:58.348) Yeah, that's awesome. So I kind of think of there's a product called Sentry that discovers errors in your site. And then there's actually an option within Sentry to fix the error. you're going down that same path where somebody could find an accessibility issue and have the code sort of fix itself, or it would be like an MCP, or tell us a little bit about that solution and what you're working on. Michael Bervell (16:25.414) Yeah, it's exactly. Actually, I remember seeing the Sentry AI launch video probably a year ago, and I put that in our team's all hands deck. And I was like, this company somehow took all of our marketing and did it for security, because Sentry is primarily security testing, not accessibility testing. We're like the first and only accessibility version of a Sentry. Or Sneak is another big one that's really popular. Or Sonar Cube, Sonar Source. These are all security companies that do security testing and security. Now they're starting to do AI enabled security remediations and fixing. And so they have interesting approaches. I think the approach that I think is going to win out, like while the MCP server is interesting, we're focusing more on like human in the loop, which MCP servers sometimes are, they kind of run on their own, honestly, and they're quite new. And so we're focusing more on like giving developers tools within the IDE. So like within VS code or within CI CD. So in like GitHub. or just scanning an end website and trying to give an audit that you can feed into an MCP server versus building our own server. It's an interesting, you we might change in a year or in two years to kind of double down on MCP once we see where the market's going, but it's kind of the wild west out there in terms of that. So we're focusing in on more precise tooling in the software development process. Brent Peterson (17:46.414) I want to change, just change topics a little bit. I want to talk about your book, Unlocking Unicorns. Tell us a little bit about that, what led to it and... Michael Bervell (17:57.88) man, I was locked indoors for three years during COVID. And actually during the first week of COVID, I think I saw a quote on Twitter and they're like, you know, during the Black Plague, like Newton was indoors and he wrote like the Principica on gravity. Like, what are you going to do during COVID? And I like, that's a great point. So I decided to do research for and write a book. and that was Unlocking Unicorns. And that's a book where I reached out to startup founders in Asia, Africa, and the Middle East. So startup founders who I traditionally didn't hear stories about growing up in America, like super Western, and just wanted to learn, does someone in China build a billion dollar company in 1980, 1990? And that was the story of Alibaba. Or how does BTS, who five years ago or 10 years ago, no one knew, become bigger than the Beatles? Right? That's to me a fascinating business case. And that's the story of like Bong Si-hyuk and his company Big Hit Entertainment, right, in South Korea. And so I was just studying these stories and trying to understand like what can I as a primarily American business owner learn from international emerging economies about how they do business and how is it different? And so there's so many stories that I learned from like Kareem, which is like Uber for the Middle East. You how do you launch Uber? in Saudi Arabia when at the time part of the population couldn't drive? How do you lobby government to help people drive so that you can drive more drivers and also more riders in a car system? How do you match gender for rides versus not? Just really interesting questions that I hadn't even considered growing up here. So that was the impetus for the book and kind of what we wrote. And yeah, when I published it, it became a bestseller for the first month that it was there on Amazon, which was nice. Brent Peterson (19:54.412) Yeah, did you find that different, like maybe the US misses out on some opportunities because the US is so US centric to everything it does. I'm not trying to be political, but maybe the US, we do think we're doing it right all the time. And we've sometimes overlook other options that are out there in the world that other countries are doing and are successful at. Michael Bervell (20:24.645) Yeah, mean, some of it is a mindset and some of it is based off just like, just pure infrastructure, right? And so like the mindset side of the house is, like, I do think that if you look towards say like even the continent of Africa, like half the population is under the age of 25, right? Whereas in the U S that's not, that's not the same, right? Everyone knows the example of Japan, right? Where Japan has an aging population. where if they don't have more kids, there's a chance that there may not be a Japan in 200 years that's the same as the Japan of today. So that's kind of just like the situational and kind of mindset part of the house. And then the second kind of interesting thing is just the infrastructure. Like the continent of Africa has no fiber in the same way that America laid down a bunch of fiber in the early 20th century and railroads. And the positive of that for America is like, hey, we can repurpose this sort of stuff, but it has high capital costs. Whereas if you're in Africa and you want to build a tech startup, like, you know, now you just use the, you know, some of the satellites that maybe Amazon is launching or Blue Origin is launching, or you use like even, you know, Elon Musk's satellites and you can now have a company anywhere, right? And so that concept is called leapfrogging, right? Where you don't need to build certain technological inner points. You can just leapfrog and build starting today. So if I were to go launch a company in Africa today, it actually would be... in some respects easier than launching it in the US depending on which industry I'm choosing. So those are just interesting. And it's just the nature doing business in some of these countries where they don't have the same history, culture, or infrastructure as in the US. Brent Peterson (22:09.314) Yeah, super interesting. Michael, as we're going into, we're recording right around Christmas time, but this will get released probably in Q1. What do you think we're going to be looking at in terms of new innovations or something like the, maybe even e-commerce? Like, is there anything you see that's happening in the industry that's going to disrupt other than what's already happening with AI and chat GPT? Michael Bervell (22:38.695) Yeah, I mean, think the biggest thing is just the unbundling of use cases. Like I kind of said at the very beginning of the call that like, um, like AI is an emergent technology and how people use it is defining like a lot of how people are building products and making businesses and so on and so forth. And I do imagine that in Q1, Q2, we're going to see even more unbundling, meaning like, you know, you used to go to chat CPT for everything. Now you go to something like, you know, for Plexity, for Search, and ChatGPT, for, you know, maybe writing your emails, but then you go to Cloud for your code reviews, and like that unbundling is really gonna lead to a rise of like really niche players. Like already there are probably niche players for SEO versus GEO versus tracking conversion rates and optimizing conversion rates. And some platforms will try to do them all. But I bet if you asked your audience today, like, you have a favorite tool for X niche thing within e-commerce? There may not be a solution today, but in 12 months there certainly will be. And some solutions that I see that are really cool as it relates to like e-commerce or just commerce online and AI is things like automating A-B testing. Like there are really cool companies like Moonshot AI. I think I've met the team at conferences a couple of times, but they claim to be able to look at your webpage, run an A-B test. find the results and automatically stimulate all of that. So you don't even have to dream up what A-B test to run. You can just say, hey, moonshot, look at my website and my data and do it for me. That's pretty cool. Or another tool that just saw that was really interesting was one where you could simulate different users on your website and try to track user flows and paths that may be broken. So it's almost like a next generation of clarity by Microsoft or any of those tools that would track what someone's doing on your website. It's not only tracking what they're doing, but also then mimicking what they might do and using that to help build better products and experiences. So yeah, just really interesting, super niche use cases of like, chat GPT, right? Where you could have uploaded it yourself. Now there's a whole product and platform doing it. Brent Peterson (24:48.982) Yeah, and I think that people are... Well, think businesses are tired of generative AI and the real magic for AI is going to be what you're describing, some of the pattern recognitions and some of the automations that are happening in the back office that are helping you do better, even automating your testing. And I like that idea of finding the different ways that users use things, because I think typically what happens is a developer makes something, they do the same things over and over again, and they're like, yeah, it works great, let's release it. Michael Bervell (25:21.73) Yep, exactly. Brent Peterson (25:21.94) and your first user goes on and like, I have no idea what I'm doing. can't figure out how to go to the end of this, whatever the workflow is. Michael Bervell (25:32.454) Yeah, it's interesting. We're in this weird part of the technology cycle where a lot of these optimizations seem so hyper-niche. Or it's like, it's actually like, you we're not optimizing for 0.1 % improvement or 2 % improvement. And is there going to be another seismic shift, you know, in commerce or really any sort of technology? Probably there will be, and I'm just curious to see what it might be in 12 months. Brent Peterson (26:00.236) Yeah. Mike, we have a few minutes left. As I close out the podcast, I give everybody a chance to do a shameless plug about anything they'd like. What would you like to plug today? Michael Bervell (26:10.994) man, read more books. Read more books and if you need accessibility services on your e-commerce site or if you're an agency and you want to either white label accessibility or have a referral partner for accessibility, you can go to testparty.ai. T-E-S-T-P-A-R-T-Y dot A-I. And yeah, you just say like, know, Brent sent you and you'll get probably a nice discount as well. for the brand set at your discount. Brent Peterson (26:39.518) Awesome. And what book are you reading right now? Michael Bervell (26:46.138) I've actually, even though I'm doing this Bradbury challenge, I've never read Fahrenheit 451. That's like his magnum opus. That's the book he wrote that everyone was like, this guy is a genius. And I've never read it. I've only read his short stories and stuff like that. So that's my next book. I guess that's my recommendation. I never read it. So again, if it's a bad book, don't blame me, blame society. I think it's also banned in a lot of schools. I don't know why, but we'll see. Talk to me in like two months. Once I finish it, I'll give you. My full thoughts. Brent Peterson (27:17.132) Yeah, think is Animal Pharmacy is another one that's sort of a classic that a lot of people have missed nowadays. I just finished David Goggins Can't Hurt Me. So if you want nonfiction about things you shouldn't do that the author does, that's a good one to read. Don't do as I do or talk about. Listen to all the crazy stuff. Michael Bervell (27:31.251) yeah. Michael Bervell (27:35.47) I do. Michael Bervell (27:39.174) I remember David Goggins, what a guy, ran a double marathon and broke both of his feet, something like that. incredible. 100 miles, that's like five marathons, that's incredible. Fuck. Brent Peterson (27:46.05) Yeah, yeah, 100 miles. Brent Peterson (27:51.98) Yeah, exactly. So Michael, it's been, I'll make sure I get those on the show notes, ways to get in touch with you. How can people get in touch with you if they want to talk to you personally? Michael Bervell (28:03.248) Yeah, I just shoot me an email, find me on LinkedIn. I think it's LinkedIn slash in Michael Bervell, so just my name. There's only one of me. So even if you Google me, there is a small period of time where I had one of those little search panels on Google, made me so happy. I think they took it away. They're like, this guy is not that notable. At the time, for like six months after I published the book, they're like, this dude might be someone. So you can Google me too. I'll pop up and there's lots of ways to contact me. Brent Peterson (28:29.73) Perfect. Michael Bervell is the CEO of TestParty. Thank you so much for being here today. Michael Bervell (28:35.969) Thanks so much for having me.