The push for digital accessibility aims to ensure equal access and inclusion for all individuals. So why are healthcare companies failing to keep up?
In this week’s episode of Dev Interrupted, we sit down with Plum Ertz, Director of Engineering at Ro, to dissect healthcare’s digital accessibility problem. Following explosive growth in telemedicine due to consumer behavior changes brought on by the pandemic, the healthcare industry has struggled to provide accessible healthcare services on the internet, regardless of the technology or disability someone may have. Plum sheds light on the work being done at Ro to address these challenges, emphasizing the simple steps that teams can take to enhance their digital accessibility.
Plum also shares stories from her time at Buzzfeed, where she accidentally took down the entire site after forgetting a semicolon and once mistakenly calling a midterm election 5 hours too early.
Episode Highlights:
- (1:35) Defining accessibility
- (6:20) Why accessibility matters
- (9:50) An explosion in the number of telemedicine users
- (12:19) How Ro defines success
- (16:05) Screwup stories
- (18:15) Stories from Plum's time at Buzzfeed
- (22:50) Getting started with accessibility at your company
Episode Transcript:
(Disclaimer: may contain unintentionally confusing, inaccurate and/or amusing transcription errors)
Conor Bronsdon: Welcome back to Dev Interrupted. We are live in New York and we've got Plum Ertz with us. Plum, welcome to the show.
Plum Ertz: Howdy. Thank you so much for having me today.
Conor Bronsdon: Anyone who comes on the show with sparkly red heels is absolutely a hit in my book, this Casual Tuesday shoe. I like it. And you're the director of Software Entering a Ro, which is a direct telemedicine startup.
Would love to dive in on a topic I know you're very passionate about. Which is accessibility. And I know it's for the work you're doing here.
Plum Ertz: Yeah. Yeah, absolutely. There's so many different ways to define accessibility. When you think of healthcare and telemedicine, I'm based in New Jersey and New York, whatever you wanna say.
And I can find a doctor. There's a million doctors, a million specialists. For a lot of folks, you don't have that opportunity. Even the United States it's not a choice of, oh, do I go online or do I go to a doctor? It's do I get healthcare or not? So putting healthcare on the internet, How we make that happen.
And it's not just enough to put healthcare on the internet, and that's where digital accessibility comes in. Just because something is online, on a website doesn't mean you can access to it because you really have to build it in a way where folks, regardless of what technology they're using, what network they're using, what ability or disability they may have, they're able to use that.
And that's not the responsibility of the person to try to figure out the Konami code to use your website. That's the responsibility of the people building these sites, building these experiences to make it accessible and sometimes that really gets lost in the build, especially weirdly enough in healthcare.
I don't know you've ever gone on like your insurance website or any like digital websites? Exactly, yes. Probably had some of the less good experiences in those spaces because so much attention is just brought to the functions of getting the insurance and the form fields and everything, and not enough on just making it usable.
Conor Bronsdon: Does the website, where can users effectively navigate to find the care they need? just an example for my personal life. When I was trying to switch my primary care physician. I couldn't figure out how to do it online based. I like looked up the documentation like, oh, here's how you do.
It didn't work. I dunno if it was just cause I was on Google Chrome or what it was, I had to call in and say, Hey, I really want to switch my pcp. How do I, can I do that? And they switched it for me. Yeah. This is the process that. Should be easy online at least was presented as easy. Yeah. And I am fairly able-bodied despite if anyone watching, I currently have a cast on my arm, but normally able-bodied person.
the challenges for others who maybe have, further accessibility issues I know can go deeper. And it seems not just in healthcare, but broadly in software engineering, we often optimize for just oh, we're getting 80% of people, 90% of people, and the edge case is really good, ignored.
Plum Ertz: Yeah, cuz it's. And even you're the perfect example right now. Cause it's 80 or 90% of people because folks are thinking about permanent disability. They're thinking about, oh, we can't get the blind people or the deaf people or whatever have you. But the problem is disability is a spectrum.
It's the kind of thing where in a given day, in a given situation, some mine might have a limitation. They might have just had shoulder surgery and only been working with one hand. So yeah, there are now a set of interactions maybe on a mobile device or a touchscreen device you can't do right now.
Conor Bronsdon: My typing speed is about a third of normal.
Plum Ertz: Yeah, exactly. Exactly. And that's the sort of thing when you think about things like one-time passwords and security codes where they have that sort of time sensitivity element to it. That's something that if it isn't built in a way with the proper alternatives to accommodate different abilities at a given time, you're not gonna be able to do, and that's locking people out of whatever they're trying to access, which in my case, close to my heart is healthcare.
Conor Bronsdon: Where are most of the pain points as far as accessibility for health?
Plum Ertz: from, there's the digital, general digital accessibility, which also applies to healthcare, and then I think from regular healthcare, a lot of it is just being able to get in touch with someone who is a specialist, get in touch with someone who is a clinician.
You can scale doctors to a certain point. In a physical space, but then you have to move to a digital space, and that's where a lot of folks have issues finding that access. There's also an issue, and especially with Ro, we deal with a lot of sexual health. We deal with erectile dysfunction, we deal with semen analysis.
So these are words that I say every day at work, and it's not as weird as you think it is after a while, but. Those are things that are really uncomfortable for a lot of people, so a lot of access is creating that safe space and that comfortable space, and that's why digital healthcare can be helpful because someone can do that whole onboarding experience without having to say, I have to drive somewhere and go to a doctor and say that I have a really embarrassing personal problem and I don't know what to do.
Can we provide that place for people to be able to have those conversations? And it doesn't necessarily have to be a synchronous conversation where you have to say that to someone, but you can still get the information you need in a way that can help you move forward in achieving your healthcare goals.
Conor Bronsdon: The work you're doing, I know, is really important to just provide that. General accessibility across, all care functions worldwide. I'll say to bring it back to another personal example, during the Covid 19 pandemic, I, did therapy online. And that kind of access was really important to me in the time as I was going through some personal challenges in my life.
And I know that's true of a lot of other folks, but some folks, I'm sure listener are saying I have access to care. This isn't a big challenge for me. Why is this relevant to my software engineering team? Like why is accessibility like that important? What's your pitch to the folks who are listenings and having that question in their head?
Plum Ertz: The pitch, you can always throw a lot of the like, larger statistics of number of people who have a disability. It's something like one in four Americans of some form of disability. Yeah. So if you wanna play the pure business route, it's Take 25% of your potential customer base and just drop 'em off a cliff cuz that's what you're doing.
Yeah. But the way that I found resonates the most with engineering teams is having them see someone struggle with the experience. It's the kind of thing where, and if you've worked in any kind of startup, a bug could hit a bunch of people and no one really notices it. But then The CEO's second cousin, like texts him and says, there's a problem, and suddenly like the entire world is stops to fix this bug.
Yeah. It's that kind of personal connection. It's seeing one person, maybe someone they know, maybe someone they've just met, go through the experience in a way that they haven't seen before, using a screen reader. There was an example when I was working back atBuzzfeed prior to this in media where we brought in someone who was blind, who was part our partner in our accessibility.
And he just went through the website with a screen reader, like One Pass. And all of the news editors, the people who wrote this content were just floored because they were hearing their content in a new way for the first time, and they didn't like it. They didn't like the way it sound like, we're not presenting this, yeah. We're not presenting it. So it's about really having that personal connection of it. Here is a single person who is struggling because you didn't put in the effort. And when I say effort, a lot of digital accessibility is free.. It's this weird thing where accessibility is both hard and very easy.
Like you don't have to, if you just had straight up htm, forget frameworks, forget everything. You have an HTML page and you put a form on it. You use the correct markup, semantic markup, it's accessible. We spend 90% of our times as engineers thRoing all this JavaScript stuff on top to put accessibility at risk.
And so it's about going back and. Making the right choices, using the syman, using the things that have been in browsers for 20, 20 or 30 years now. Which doesn't take a lot of time unless you've spent all of this time building this product without it, and suddenly you're sitting the day before launch, you run it through a checker and it's oh, and now we have 700 accessibility bugs.
Conor Bronsdon: Like whatever technical debt I'll deal with later.
Plum Ertz: Yeah. Whatever technical debt, you'll deal with it until you know you're creating negative experiences or. I hate to say it, but a lot of times, it's like a lawsuit-based thing where that's when the impetus comes for companies, which I hate. I hate that people can't just go in and be like, Hey, we need to build the right things.
When you're building for folks for disabilities, you're really building a general user experience overall. You're reducing the risks of bugs overall. All of those things, but sometimes it literally takes the legal aspect to give people that kick in the pants.
Conor Bronsdon: It's really interesting cause I think a lot of us think. Disability and, accessibility as this thing that's, far away from us. If you're in that 75% of the population that doesn't have to deal with it.
But there are gonna be periods, to point to my personal example right now with my shoulder or just as you age, where you're gonna experience this and. Particularly with telemedicine and access to healthcare, this becomes very important as maybe you need to access more healthcare services while starting to have, maybe it's hearing disabilities as a, like a very common example or, yeah, simple eyesight issues.
So I love that you're zeroing in on this because particularly in your field, it's very important. Where are the other challenges that you're seeing as far as the healthcare vertical, today?
Plum Ertz: There are so many challenges just in terms of there's so many things that have yet to move online.
I think one of the things that I'm really excited about when it comes to the industry is how much we've been able to start moving online, moving to be an in-home thing. At Ro we're basically the only company in the US that does pharmacy physician. In-home lab testing, diagnostics, all of that is something that someone can get without being able to leave their home.
This is actually really, and you brought up the pandemic earlier, like Post pandemic or I guess during the pandemic. I don't actually know if we're post pandemic yet. That's topic challenging question. Yeah. Other podcasts, but the pandemic brought forward telemedicine so quickly.
Yeah. And so maybe that's actually the challenge is the fact that a lot of people got online really quickly and I was, look, I was definitely looking at Pew Stats yesterday. Pew Media Research, because I studied comms and economics, and so they were my go-to for things. And it was something like in 2021, 40% of people surveyed use the internet in a new way for the first time during the pandemic.
And so now you have this whole new population of folks who are coming onto the internet. Potentially becoming more comfortable with things like healthcare online, whereas before they weren't. Even though telemedicine has been around since the 1960s, like but it doesn't have this focus on it, to your point. Yeah. But it hasn't had this focus and now it's like we maybe in, three years brought ourselves 10 years into the future in terms of what folks are looking for. And so part of it is catching up. Part of it is actually bringing all those experiences online in a good way.
And part of it is helping people catch up. Cuz I'm, yeah. It's not just like everyone who's a teenager went online for the first time. It's folks who never grew up with technology or grew up with a different type of technology who are now trying, and in a lot of cases, struggling to use these applications to get healthcare in a way that should be available to them.
And so working to make that more available to them.
Conor Bronsdon: I think it's also one of those areas where once you experience the convenience in a successful way, it's hard to go back. let's say I had an eye infection. And normally I'd have to drive in or take the bus or whatever into a doctor spend all this time.
Maybe instead I'm like I have work meetings all day. Can I set up a telemedicine appointment this afternoon? Hop on. Doctor can look at my camera very quickly over, zoom and say, oh yeah yeah, here's the antibiotic. Go pick up for your pharmacy, or even we'll ship it to you. And so I think this is where your points around access and having it go beyond simply the folks who may have challenges reaching access, but also just the convenience you can reach on this population are huge.
And I really liked the way you described this when we were talking earlier, about what Ro does. How would you define, the success of Ro and what you really do for people?
Plum Ertz: It's goal-oriented healthcare and I think the important part, and I guess the secret sauce to our success, is the fact that Ro is backed by a lot of people who care.
I go into work, I didn't go into work today, but I go into work most days. I'm in standups with physicians. I'm in standups with pharmacists, I'm in standups with our legal team. We're. Building a technical product like fully in our react world, arguing about how we want to use TypeScript to defi. We do argue about TypeScript, but we're the problem, the problems that we're trying to solve are way less technical and more about how do we take the parts of the clinical process and put them online in a smart way.
And I'll say this up front, we're not using ai. There's AI and chat, G P T and like the alphabet soup of different algorithms. We're not using any of that. We're just putting clinicians online in a smart way. The technical problems we're solving are not these. Deep algorithms where we have to do all this, are difficult.
We're using things that have been in existence for a while, but just using them to automate healthcare.
Conor Bronsdon: And it's a focus on value to customers, which is what I love. Where, folks may hear this say, oh, I'm not in healthcare. Does this matter? Yes. This is like the, a key focus on providing value to customers.
And so many organizations lose this In the alphabet soup you mentioned. Yeah. Where they say, oh. Our CTO thinks we should, innovate this way, or our VP of sales said he heard this on a call and instead they need to be focusing on what's the actual outcome for customers instead of just using anecdote data format.
Because that kind of laser focus is what great companies succeed off of. And I find a lot of product and engineering organizations have challenge. With aligning the work they're doing, the features they're delivering back to the customers. So I love that's this focus for you. And, I'm curious, in your past work how have you had that same customer obsession?
And some folks will call it and brought that to the engineering and product orgs.
Plum Ertz: In my past work, it's definitely been very different cuz my past work was very focused in agency, in media and so a lot more of that is around just the sell and the users are more of eyes. When I worked in a purely agency, I did some contract work for Verizon like years ago, and.
There was never any connection to the Verizon user as a customer. It was just like, here is this thing. And so working in healthcare has been a totally different experience for a lot of reasons. One, just obviously the regulatory aspect of it and all the due diligence that you need to do. And the second is just really caring about people and.
From various media stories. There are certain health organizations who don't put that at the forefront, and that's why I love Ro because we put the patient at the forefront. Everything we do and everything that we're working on is built around. How is this solving a patient problem? If anyone said anything about the blockchain, I would straight up walk out of the building.
I've told every person at the company, like you say, blockchain, I'm all like, that's not what we're here for. We're not here for th eAI, the innovation, or the technical innovation for the sake of technical innovation. We're here at technical innovation for the sake of how can we make patients lives better?
How can we make it easier?
Conor Bronsdon: That's a really apt point that we sometimes forget. I have to admit I'm guilty. This sometimes I got all excited. I get to have these conversations where people talk me these like amazing technologies, blockchain and AI and all these things. I'm like, yeah, it's really cool.
But remembering that delivering value to customers is so important. So it's interesting to see how entering this vertical as the, as kinda your focus last couple years has really informed that viewpoint. Cause I, like I said, I think it's something where the best engineering leaders I talk to say, we are so focused on delivering value to customers, here's how we do it.
Yeah. So love to hear that and that kind of focus. I'd also love to dive into something that we talked about briefly before this show. Which is, we like to ask this question of leaders who come on sometimes and say, what's a screw-up you've had as an engineer, engineering leader that you share with us?
Plum Ertz: Oh, goodness. I've made screw-ups at every level of my career, which is excellent. I think it's like a rite of passage at a certain point. Yes, you have to. Break something. And there's a couple of reasons why you have to break something. One is because it shows that you're really digging in your learning, your application, you learn from it, and you also learn how to fix things and work together and save it the next time.
So specific example, my like favorite one is for my last job I took downBuzzfeed. I took down the entire, I took downBuzzfeed, I took that. Oh man. All of the articles super. We were doing a full relaunch, like fully accessible relaunch of the article pages, which are, the most traffic pages on the site.
And we had spent so much due diligence. It was the first time we were using next js, and it's Node base, backend, all the temps there. We got that all set up and we had scaled everything except that bold little reddish cash that we had been using in staging. And it was like whatever the smallest AWS reddish cash is.
And we flipped the switch to turn the site on and sent all ofBuzzfeeds traffic, which is I don't know how many, hundreds of thousands of views, like straight at this little Redis cash. Amazing. She just fell over. She immediately, she's what? Sign down? And I just, it was the kind of. And again, like at this point, I was a staff software engineer.
I was the engineering leader. I was managing four people. I was the one leaving this and I just completely knocked it all down. And the thing you don't panic. You very quietly type out a slack. Me, I don't panic. Panic. Yeah. Very quietly. Type out a Slack message to our DV admin. Shout out to Felicia.
And I'm like, Hey Felicia, can you just set up a new Redis cluster for me? 15, 20 minutes, we are back.
Conor Bronsdon: Oh man, that's a great story. And I love these kind of stories because it feels like every engineer has one and everyone who's worked at engineering for a while is oh yeah, I forgot a semicolon.
Took the whole site down something.
Plum Ertz: Oh, I did that too. I do that too. Both atBuzzfeed and at Ro.
Conor Bronsdon: Awesome. This is a great transition cause I'd love to dig into your career atBuzzfeed a bit. Because we actually haven't had a lot of folks from media on the show lately. What's different when you're working with, as an engineer in media?
What was interesting about it.
Plum Ertz: Ooh. Working for, and I worked forBuzzfeed I guess from like 2018 to 2020 to set the stage. Cuz the first thing that everyone asked me is oh, did you know the try guys? No, I did not know the try guys. Yeah. Gone past, had time. Sorry, I have no good spicy stories from there.
Working with media. I actually, when I started atBuzzfeed, I worked withBuzzfeed News, so I started right around the time they launchedBuzzfeed news.com. Cool. And that was a cool experience because you get to work, I sat directly in the news pit, so you get to see folks working day-to-day and that's where you start to, You don't build as much connection to, theBuzzfeed reader audience, but you build a lot of connection, or at least I did to the news team, and really understanding like they're putting a lot of effort into these stories.
There's some Pulitzer winning stories. How do you build the technology, the experience to support getting that information out there and getting that work into the world? So building that kind of connection was really fun. I covered the 2018 midterm elections, which is cool. Wow. Also broke something then accidentally profit election, like 5, 2, 5 five hours too early.
It was fine. All you gotta tell us about that. Okay, so we're using a third party. API that was sending back the results like as they were coming in? Yeah, and I guess at some point I forget, it was either they were putting test data in production and we had pointed a production too earlier. We were still pointing to staging.
So anyway, I called Texas for Ted Cruz five hours to I have never seen the editor-in-chief, or not the editor in chief, one of senior editors like materialized behind me because this is when we were sitting in the opposite, she saw the website and she. How do we make it go? I'm like, one moment please.
Conor Bronsdon: I'm sure someone screenshots that's on Twitter immediately people are like, oh,Buzzfeed's calling it.
Plum Ertz: I think it might have been early enough that we were okay. Okay. But cuz it was like we were pointing most traffic to a couple specific like article pages. This on the home page. So fingers crossed.
Conor Bronsdon: Oh man, that's a hectic day at the office for sure. And it's interesting to see the kind of switch you've had in your career. You've worked on this like very high pressure, front facing piece, and then you have a lot of like extreme customer service work that you're current doing today where it's like, how do we align directly to customers?
But to draw it back to your accessibility point earlier in the conversation, I think it's really crucial to understand the importance to have the accessibility for both c. Because you mentioned the stat 25% of Americans have some sort of accessibility challenge, and with media that means like a massive portion of your audience is gone.
So you brought up the screen reader example earlier. How didBuzzfeed bake into their user testing and their, exploratory plans around new features, et cetera, that approach accessibility to ensure they weren't cutting out a large portion of the audience.
Plum Ertz: A lot of it was just building in the rules, the tools, technology buzz.
I don't know how much I'm allowed to say, but both people don't work there anymore. They have their own internal cms. Yeah. And so a lot of it was actually that CMS powers a lot of the content that's being built out there. So there's two pieces. There's making sure that the render, which is the piece that I built a fair portion of is accessible and obviously two years down the line.
So if there are any accessibilities issues now, Look, you've been gone for a while. Yeah. But no, they still have a really great team there who's still really focused on accessibility and then it's building into the actual tools. We built things into the quiz editors. So if you had a color on a background, it would give you an approximation of Hey, this might be a contrast issue, like doing as much as possible to surface that the best way to build accessibility is to build it as Hurley into the.
As possible. So when you're literally talking about initial feature spec or a new quiz format or something, you wanna think about, how will this work with a keyboard? How will this work if someone has a screen reader? What are the alternative content options for it? And moving forward from there.
And honestly, the same thing we're doing at RO now when we're looking at boats, landing pages and ways to explain to folks. And as we're getting into more complex treatments, we're working on GLP one s now, Ozempic Gobi, which are in the news for a whole bunch of things. When we're building those experiences, we're teaching folks how to do something at home that they may have never done before.
So how do we give them as many means and ways as possible to learn how to take their medication?
Conor Bronsdon: So this brings up a really interesting question. You alluded to baking in accessibility, early whatever you're. For engineers who are listeners and saying, okay, I hear what you're saying.
How would you start that journey of saying, all right, maybe we have to go back, fill some pieces of our technical debt around accessibility or maybe we need to, build something new from the start the right way, so to speak.
How would you initiate that?
Plum Ertz: It's very organization specific. So my number one piece of advice is always start somewhere. There is always some piece of something you can do, no matter how small that can make a change, even if it's, Hey, I'm gonna go back to this form and I'm gonna add labels to all the form fields.
And congratulations. You've solved one of, I think the top 10 bugs that appear on websites to date. Then it's about building that partner. You can always be the loan accessibility advocate, but you really need to build that partnership. And so it takes a little bit of politicking in a way. So working with your team, getting them excited about accessibility and in whatever way that means for your organization.
For some folks, As much as, I hate to say it, accessibility has a benefit to seo and for some folks it's like the SEO slaps. All right, if that's the thing that's gonna get this funded, let's go for it. For some folks, it's seeing A user having struggles and seeing how we can make that better for some folks, literally just the empathy aspect of Hey, by the way, there's a whole set of people in the world who want to be able to use these sites, and we're preventing them from doing so for literally no good reason.
You find that selling point and you start to get folks on board and start to have accessibility conversations to see what can we do, what can we fund? And every company is in a different spot. If you're in a really early stage startup. It's gonna be really hard to get the buy-in for Hey, we actually need to redo our entire platform.
Is that cool with y'all's? That's not gonna fly. Yeah. But if you can say, Hey, there are these five things that we can do that will take two days, that will greatly enhance the accessibility of certain pieces. I think one of the most common issues actually is color contrast. Which is also one of the easiest issues to fix, but can be really problematic.
It can create a lot of problems for folks with low vision, different variants of color blindness.
Conor Bronsdon: I’ll be frank here and say our website currently, we realize we had an issue that we need to address around color contrast. So like it's absolutely a problem that's very pervasive, even for a media podcast like us.
Plum Ertz: Yeah, and those are the ones from the Dev estimate level of effort. Gimme five minutes from css. I got you covered.
Conor Bronsdon: So this is a great way to I think, close out this conversation because this is such a challenge for a lot of folks. And we've had Fable on the show before to talk about the approach to develop it first with accessibility.
And I really appreciate you expanding on your approach here with RO and how you approach it at bus speed as well. Do you have any clothing thoughts? As far as accessibility or anything we touched?
Plum Ertz: I think I'll reiterate what I said before, which is everyone can do accessibility and take the time every day to do at least one thing to make your experience better for someone in some way.
Conor Bronsdon: I love that. if you enjoyed this conversation with Plum and maybe you found a nugget of Atlas accessibility conversation helpful, you wanna apply it in your own organization, let us know on social media. we'd love to hear from you and we'd love to have that feedback from our listeners help guide the.
And Plum. Thank you so much for coming on.
Plum Ertz: Yeah, absolutely. And I think I'll do my closing of please visit us at ro.co.
Conor Bronsdon: I love it.
A 3-part Summer Workshop Series for Engineering Executives
Engineering executives, register now for LinearB's 3-part workshop series designed to improve your team's business outcomes. Learn the three essential steps used by elite software engineering organizations to decrease cycle time by 47% on average and deliver better results: Benchmark, Automate, and Improve.
Don't miss this opportunity to take your team to the next level - save your seat today.
