Season 3 Episode 4 Feb 4, 2021

Jeff Bezos Stepping Down, Independent COVID Vaccine Systems, Ford’s Android Integration, and Tech Pricing Models


Amazon gets a new CEO


In this episode, we talk about Jeff Bezos transitioning from CEO of Amazon to executive chair of the Amazon board, Ford’s future Android integration, and what kinds of pricing models work for different tech products. Then we chat with Cameron Yick, software engineer at Datadog, and one of the creators of the NYC Vaccine List, a website with the aim of helping people find a COVID-19 vaccine in and around NYC.


Saron Yitbarek

Disco - Founder

Saron Yitbarek is the founder of Disco, host of the CodeNewbie podcast, and co-host of the base.cs podcast.

Josh Puetz

Forem - Principal Engineer

Josh Puetz is Principal Software Engineer at Forem.


Cameron Yick

Datadog - Software Engineer (Data Visualization)

Cameron is an engineer at Datadog and co-organizer of the NYC Data Visualization Meetup. He is also one of the creators of the NYC Vaccine List He writes and speaks about making public data useful, visual tools for thought, and creating conditions for Serendipity.

Show Notes

Audio file size





[00:00:10] SY: Welcome to DevNews, the news show for developers by developers, where we cover the latest in the world of tech. I’m Saron Yitbarek, Founder of Disco.


[00:00:19] JP: And I’m Josh Puetz, Principal Engineer at Forem.


[00:00:21] SY: This week, we’re talking about Jeff Bezos transitioning from CEO of Amazon to Executive Chair of the Amazon Board, Ford’s future Android integration, and what kinds of pricing models work for different tech products.


[00:00:33] JP: Then we’ll chat with Cameron Yick, Software Engineer at Datadog and one of the creators of NYC Vaccine List, the website with the aim of helping people find a COVID-19 vaccine in and around the New York City area.


[00:00:45] CY: It got us to thinking about how long it would take someone who was not very comfortable with computers to navigate this and realizing that there was an opportunity where the technical skills that we had could be applied to save a whole lot of people a lot of time and a lot of stress.


[00:01:00] SY: So on Tuesday, Jeff Bezos penned a letter in the Amazon Blog announcing that he is transitioning into the role of executive chair of the Amazon Board in the third quarter of this year, and that the new CEO of the company will be Andy Jassy, who is the current CEO of Amazon Web Services. In the letter, Bezos says, “Being the CEO of Amazon is a deep responsibility and it’s consuming. When you have a responsibility like that, it’s hard to put attention on anything else. As executive chair, I will stay engaged in important Amazon initiatives, but also have the time and energy I need to focus on the Day One Fund, the Bezos Earth Fund, Blue Origin, the Washington Post, and my other passions.” And I totally forgot he owned the Washington Post. Completely forgot about that. I was like, “Wait, you’re going to do what?” He also owns some newspapers.


[00:01:48] JP: It’s a long list of stuff he’s involved with. It’s like your Amazon Prime Membership. There’s just a laundry list of stuff that’s included when you’re Bezos, I guess. So I don’t blame you for losing track of that.


[00:01:58] SY: I can’t believe how many things. I mean, it’s not just him. Right? I feel like there’s so many tech executives that are juggling multiple companies, literally multiple companies like Elon Musk, like Jack Dorsey. How are you CEOs in multiple places? I don’t understand. It’s absolutely incredible to me. So I do not fault Bezos for being like, “Look, man, I’m spread too thin. I got like five potentially world-changing organizations that I need to run and I’m going to take a step back. I’m not mad.”


[00:02:27] JP: I mean, I also want to choose, like Amazon. It has been around for quite a long time.


[00:02:31] SY: Twenty-seven years! That is a full-grown adult. That is an adult who’s transitioning to a manager position at a tech company.


[00:02:38] JP: There’s a little photo, Amazon’s just walking around the headquarters. Amazon is at a fairly stable point. They’re doing interesting things, of course, but these other endeavors, Day One and the Washington Post and their climate pledge and all these other companies are possibly more exciting to him.


[00:02:57] SY: Probably. Yeah. I think so. I don’t know how old these other companies are. I don’t know at what point they were started, but they’re definitely newer than Amazon. So it’s probably a little bit more exciting. What did you think of the AWS CEO becoming the CEO of Amazon?


[00:03:11] JP: I think that part is actually fascinating. So I looked up Andy Jassy a little bit. He’s been with Amazon since I believe ’97. Let’s just say the late ’90s. So he’s been with Amazon for quite a long time. He’s been in charge of the AWS division for a while. So it makes sense I think both from a standpoint of you have someone that has been with Amazon and obviously has worked closely with Bezos for quite a long time. Also, you have someone that, I mean, frankly, AWS is, well, they’re not the largest share of revenue coming into the company. I think the area that they are dominating in, in terms of the cloud provider market and a very important market for the rest of the internet, we all feel the pain when AWS goes down or has outages. And I think that speaks to the importance of AWS among Amazon’s business portfolio that they would pick this person to lead the company.


[00:04:04] SY: I find it so interesting because I think that as developers, we know both sides of Amazon, right? We know Amazon is potentially prime customers and just as users of buying things, but then we also know them as AWS and just such an integral part of a lot of our infrastructure, but I don’t think most people know what AWS is.


[00:04:22] JP: Right. Right.


[00:04:23] SY: I don’t think most people will have any idea what that is, what that world is. I think it was just totally behind the scenes for them. And so it kind of makes me wonder, like, what future do they see for Amazon such that they would take something that’s relatively unknown, unheard of? Even if they knew what AWS was, I don’t think they would get it. I don’t think they’d really appreciate what it is. What does it mean? What is it signaling to us by bringing that person to lead the entire organization? I’m very curious about that.


[00:04:53] JP: I mean, it could also signal that Amazon is not taking it for granted. I mean, if you look at their retail operations, their mail order operations, they kind of own the show, right?


[00:05:03] SY: That’s true, in terms of growth. Yeah, that’s a good point.


[00:05:05] JP: Right. But when you look at the cloud market, sure, Amazon is a leader, but Google and Microsoft and all these other companies are very, very close behind.


[00:05:13] SY: They’re coming.


[00:05:14] JP: They’re coming and they’re not giving up anytime soon and they’re continuing to improve their offerings and actively try to challenge Amazon. And I think this might be part of a realization that they may want to put more resources into AWS.


[00:05:30] SY: We’ll see what happens. Third quarter is going to be an exciting time for Amazon, an exciting time to watch and see where they go.


[00:05:37] JP: Do you think they announced this or early to try to write out any kind of stock, any ups and downs?


[00:05:42] SY: Oh, that’s interesting. Maybe. I don’t know. I didn’t even think about that angle. There probably is. I mean, I don’t know if the third quarter is like, it feels really far away to me, but maybe that is normal. If a huge CEO is stepping down, maybe that much lead time is the normal amount for such a huge decision. But I’m sure that the transition is going to take forever. I can’t imagine going from a department, I don’t know if AWS counts as like a department, but a part of the business to being in charge of the whole thing. I’m sure there’s so much transitioning and onboarding that the new CEO will have to do. So maybe that just takes orders to do. I don’t know.


[00:06:19] JP: So another reason Amazon and Bezos may have had such a lead time at announcing this is that there’ve been other, I don’t know, not announcements, we’ll say revelations about Jeff Bezos and his personal life that have come out suddenly in the past, things like his marriage and some of his business dealings. This might be an attempt to try to get ahead of those and not make it seem like he’s resigning or stepping into the executive chair position because of some crisis or some kind of gossip that hasn’t been made public yet.


[00:06:53] SY: I don’t know. I feel like, no longer the richest man, as the second richest man in the world, I don’t really feel like he cares.


[00:07:00] JP: Sure.


[00:07:00] SY: I mean, if you think about the debacle of the photo of his precious parts and the fact that he was basically like, “Yeah, this photo exists.” All right, cool and it didn’t faze him at all. I just can’t imagine it being like a gossip protection situation, especially because, I mean, the reality of Amazon is a lot of people hate Amazon for very valid reasons, right? There’s been so much controversy on how they treat their warehouse workers. There’s been a huge controversy in how they treat their merchants and AmazonBasics is their line of products. So they basically take products that are working and kind of knock them off and do like their own version, which is not necessarily the nicest thing to do. So there’s just been so much controversy around how they work and how they operate. But still, they are the most-powerful-everything store. You know what I mean? Despite all these issues over the years and all this hate over the years and all this complaining over the years, I don’t know the answer to this, but has it actually impacted their business in any significant way? Because I don’t think they’ve changed. I don’t think they’ve really done anything. And given the fact that I don’t believe this stepping down comes in the face of any new controversy or any new revelation about how awful Amazon is to other companies, I don’t know, I kind of I feel like he’s doing it because he’s ready to move on.


[00:08:26] JP: Yeah.


[00:08:27] SY: But I am curious to see what it means for the brand, right? Because Bezos, I mean he started when Amazon first launched like this beloved tech internet rock star and then his brand definitely, definitely went down over the years. And I’m wondering if there is going to be a positive branding impact to Andy Jassy being the CEO. Is there going to be a new positive phase? Not so much that they do it because they were afraid, but more of will they use the new CEO as an opportunity to rebrand themselves? Will they take this and say, “Well, now that we have a new phase, let’s milk this, let’s try to change the way people see us”? And wants to be a more positive supportive business partner. Maybe they’ll take advantage of this change.


[00:09:14] JP: Maybe they’ll simply just stand in front of Congress and say, “Look, we changed the CEO. We’re good now.”


[00:09:17] SY: We’re done.


[00:09:20] JP: Well, this week, Ford joined other automakers like General Motors and Volvo in striking a deal with Google to use Android OS to power its vehicles’ infotainment systems. The company said that by 2023, millions of its vehicles will have this integration. Currently, Ford vehicles run an infotainment system called “SYNC”, which is built on top of Blackberry’s QNX Unix. Yeah, that Blackberry, the one from the late ’90s with the messenger phones. So Ford didn’t specify which models would be switching from Blackberry QNX to Android, but it did confirm it’d be building a new version of its SYNC software on top of Android. And they also left open the possibility that their vehicles might natively run an embedded Android auto stack instead of their SYNC software in the future. In any case, Google seems to be making some huge strides in this area, procuring a lot of the automotive market. It’s also terrible news for Blackberry, which claims to be the market leader in automotive software with over 175 million vehicles currently using their OS. Again, that Blackberry.


[00:10:26] SY: I mean, good for them. I had no idea.


[00:10:29] JP: I checked the store three times. Our producer Levi checked it several times. Blackberry, the people that made your stock market CrackBerry messenger from the late ’90s, early 2000s. Who knew? They make embedded Unix software now. It’s crazy, right?


[00:10:44] SY: Good for them. I’m so proud of them for being the market leader. And I mean, not for long apparently, but as of today, before 2023, they will continue to be the market leader in automotive software. So that’s awesome. By and through, they’re doing their thing, 175 million vehicles, a lot of vehicles.


[00:11:04] JP: So I have to point out Google’s terrible naming. I was poking around the Google website and Google has two products. One is called Android Automotive and the other is called Android Auto. And they are different.


[00:11:18] SY: Oh!


[00:11:18] JP: Google has an FAQ about this that points out, “This name was confusing.”


[00:11:22] SY: What?


[00:11:22] JP: Yes.


[00:11:23] SY: Then why do you name it that way?


[00:11:25] JP: Android Automotive is what Ford is going to be using. And it’s basically the low-level Android operating system and it gets embedded in a car.


[00:11:33] SY: Okay.


[00:11:34] JP: Android auto is software that would run on your Android phone, if you had a Samsung Galaxy or a Google Pixel. And Android Auto is the software that kind of projects or casts your phone’s interface to your in-dash display on your car.


[00:11:50] SY: Okay. Okay. Okay.


[00:11:51] JP: Ford is going to be basically using the Android Automotive as kind of the kernel of their infotainment system going forward. It’s a really interesting field because the lead time on these products is pretty long. Think about when a car gets made, other than Tesla, most car manufacturers don’t do over-the-air updates of their software. They have to build it into the car when it rolls off the assembly line. I don’t know if you’ve ever had this situation where you’re like, “Oh, there’s a great new feature in this car’s operating system,” and the car manufacturer’s answer to get it is, “Well, buy a new car.” No.


[00:12:28] SY: Such a frustrating answer. Oh my goodness. Yeah. This is a really interesting story. And it makes me wonder kind of like what is the future of Android in cars. And what is the future of just cars at alphabet, just in general? Where do they see themselves? And do they see themselves as a potential competitor in terms of their self-driving car initiative or do they see themselves as a partner to companies like Ford? On the partnership side of things, it makes me wonder, like, “Is there a world where we can build apps for cars?” How cool would that be? If you could build little apps and little things for the Google Play Store and that was available specifically for cars, that could be really interesting.


[00:13:13] JP: You could kind of do it now, but it has to be an app on your phone that also has a car mode, and that’s not the same.


[00:13:20] SY: Exactly.


[00:13:21] JP: This story made me curious about what’s happening on the other side of the fence over at Apple. There’s been so many rumors and so much speculation about, is Apple building a car? Will they someday build their own car? Will they build software? And I could see a future in which maybe Apple builds something like Android Automotive, where it’s an iOS variant kind of like that runs on the Watch or that runs on Apple TV, but could be embedded in a car. And the same thing. You could have your car apps there.


[00:13:50] SY: I think we’ve been waiting for a long time to see what Apple does in the car industry. And you know that Elon Musk approached Tim Cook to acquire Tesla back when they were doing really, really poorly and were basically like running out of money. And apparently, Tim Cook refused to take a meeting with Elon Musk. I wonder how they feel about that now.


[00:14:07] JP: These companies are sniffing around so much with cars. I don’t know about you. I haven’t really driven anywhere in weeks.


[00:14:16] SY: Yeah. No.


[00:14:17] JP: It seems quaint at this point that I’ve spent all this money on insurance on my car and it’s very well covered for sitting in the parking garage.


[00:14:24] SY: Yeah. It has not been getting any use. Yeah. But I’m sure that’ll change as soon as everyone gets vaccinated and hopefully this whole situation passes at some point in the near future. I think that people will be a lot more interested in these car conversations at that point.






[00:14:57] SY: RudderStack is the Smart Customer Data Pipeline. Easily build pipelines connecting your whole customer data stack, then make them smarter by ingesting and activating enriched data from your warehouse, enabling identity stitching and advanced use cases like lead scoring and in-app personalization. Start building a smarter customer data pipeline today. Sign up free at


[00:15:19] JP: Are you looking to build a chat for your next big project? Save time building in-chat, voice, and video for mobile and web applications with Sendbird. Get to market faster with Sendbird’s UIKit, pre-built UI components, best-in-class documentation, and support for developers. Join companies like Reddit, Delivery Hero, Yahoo Sports, and Hinge. Start your free trial today at




[00:15:44] SY: So the next thing I want to talk about isn’t exactly a piece of news, but it’s something that I’ve been thinking about a lot personally as an entrepreneur and was even more on my mind this week after I read an article posted in TechCrunch and it was titled “Subscription-based pricing is dead”. Now I love these dead stories.


[00:16:02] JP: Dead. Dead.


[00:16:03] SY: And then it comes with the subtitle, “Smart SaaS companies are shifting to usage-based models”. And I’ll put the link to that in the show notes. So let me give you a few highlights from this article. Essentially, it’s saying that a lot of SaaS companies like Datadog, AWS, and Stripe are doing well using a usage-based pricing model instead of a regular flat subscription-based model. Then the piece goes into a few reasons why this is so and how to do it well. It says that starting with a low cost allows you to draw users in with less friction, similar to the freemium model, and then investors can see what kind of value the user is getting out of the product, even if they aren’t locked into a subscription. So it’s a great way to kind of just warm them up. Right? And also a great way for you to get data on like what your customer is like and all that. And of course you have to pay attention and pick the right metric to base your usage-based pricing model on. That’s super important. And it’s interesting to think about app trends and how different pricing models have become more or less popular over time. I remember when a lot of apps were just a flat fee.


[00:17:04] JP: Right.


[00:17:04] SY: Adobe was paid like a couple hundred bucks. Right? I don’t remember how many.


[00:17:08] JP: Oh, it’s hundreds. Yes.


[00:17:10] SY: It was hundreds, right? Yeah. And then now they're exclusively subscription-based and then it basically it feels like everyone switched over to the subscription-based model as of, I don’t know, when the timeline was like five years ago, it feels like everyone just kind of switched.


[00:17:22] JP: Yeah. I feel like it’s been hitting consumer apps like just the last year or two.


[00:17:27] SY: And our producer, Levi, was actually telling us about an app he paid for that did soundwave visualization for audio files. So you get some nice little graphics when you listen to audio. There was an update on that app that stopped you from being able to use the app unless you paid for a subscription. So he’d already paid just like a flat amount for the app and then was basically like, “Look, man, can’t use it until you pay us every month,” which is pretty crappy. So I’ve launched two separate companies and I think a lot about what’s the right pricing model and how do you price and how do you build the right business model for different apps and services. And for CodeNewbie, it was all advertising, right? It was all sponsorships. And now with Disco, my new company, we are probably going to do the subscription route monthly and annual. And so I think it’d be interesting just to talk about models and what they might be good fits for and kind of the reasons why you go behind those decisions. So I want to start with like, just as a consumer, Josh, how do you feel? What are your thoughts on one price subscription, usage-based? How do you experience these models?


[00:18:33] JP: I feel like a lot of the consumer apps that I use are still in that transition from you’re buying software or you’re paying a one-time fee to moving to a subscription-based model. Whereas for business purposes, I definitely am seeing the usage model kick in now. I’m really curious if we’re going to see consumer apps go to usage models now. But the examples you gave like Datadog and AWS and Stripe, those are all real business-to-business type of services.


[00:19:02] SY: Yes.


[00:19:03] JP: And things like Netflix and Spotify, they’re still very much in the subscription realm. I don’t know if we’ll end up seeing things for how many movies I watch. I’ll have to pay a usage fee to Netflix.


[00:19:17] SY: I would be so mad if that happened. I’d be so upset. If Netflix started charging me based on how many hours of TV and movies I watch, I’d be so pissed. I feel like the usage model, I just don’t feel like it’ll work for consumers. I really don’t.


[00:19:32] JP: It’s complicated. It’s complicated to explain. For anyone that’s ever gone to the AWS pricing page and tried to answer the simple question like, “How will I pay?”


[00:19:42] SY: What is my bill going to be?


[00:19:43] JP: Yeah. What is my bill going to be? And then they bring up a rocket launcher dashboard of like complicated algorithms. And then you go to Netflix, they’re just like, “At 70 bucks a month, you’re good.”


[00:19:54] SY: I feel like consumers need more predictability that I think businesses need. I think with a business I think that because you’re used to having variables, because even with subscription, I don’t know if this counts as usage or a flat fee, but based on how many employees you have, you have a different amount that you’re paying. You know what I mean? I think this idea of kind of variable pricing, the whole contact sales for your custom solution. I feel businesses are very used to that world of not having frankly just a straight flat one-time number they can rely on for years. And I think that when we think about consumers and how we do budgeting and how, frankly, a lot of people are living paycheck to paycheck, I don’t think they can stomach the unpredictability of their bills. If all their subscriptions were a hundred percent usage-based, I think that’d be tough. To be fair, there is already usage-based when I think about electricity and water, right?


[00:20:47] JP: Right.


[00:20:48] SY: With utilities, that does happen, but I don’t know. I feel like there’s a limit of how much of that consumers would tolerate.


[00:20:54] JP: Something I also think about as well is that businesses with the usage-based models, a lot of times those track to the revenue they’re bringing in or the traffic they’re bringing in. So if I’ve got a site on AWS or I’m processing payments through Stripe and it’s all usage-based and my usage on Stripe goes up for the month, I mean, that’s probably a good thing.


[00:21:13] SY: I’m doing better.


[00:21:13] JP: I’m processing more payments. I’m doing better, right? With consumer-based stuff, I’m not making more income because I watched more, in fact it’s probably the opposite. I think it’s a harder sell for consumers because a lot of these services and subscriptions for consumers, they’re entertainment. They’re discretionary income. They’re not part of what brings in your income or what you do day to day in your job. Whereas for companies, the balance is flipped a lot. Companies are mostly spending money to make money and they’re not doing a lot of discretionary funds spending.


[00:21:46] SY: Absolutely. And whenever I look at someone else’s business model, even when I think about my own business model, I try to think, one, frankly, how much can I possibly charge that the consumer would tolerate? Right? That’s like the number one business question. It’s like, “What is the willingness to pay? What does that limit? How high can I go and still have it be a good price?” But I think the other part of it is what resources am I consuming to produce this product, to give this service. And so for Disco, the subscription model makes sense because every month we’re producing new content. So every month, there’s a cost to us in producing new content and you’re getting the value of additional stuff that’s always changing, that’s always updating and always getting bigger. So that relationship to me makes sense. And when I look at other models, that’s the question I ask as a consumer. I go, “Okay, am I getting like monthly value out of this? Is this costing them something where it makes sense?” How does it kind of fit into their business? And if I can see a situation where I’m like, “Okay, this is probably costing them money to maintain, to update, to upgrade,” it makes sense. I don’t get as upset about subscription models as I think a lot of other people do.


[00:22:56] JP: It really depends on how I use the app and when it’s like getting to my face. And I have definitely been on the end of, I paid once for some software or I am paying a set fee for a subscription and it either goes up or it transitions from being a one-time fee to subscription software. And that really just makes me analyze my relationship with the service. There have been services I’ve dropped where I’ve been like, “I paid for it once. I don’t really use it anymore. It’s not worth $10 a month for me.” But there’s other services where I’m like, “Oh, I’m actually kind of glad you’re charging me now,” because case in point, there’s a Twitter client that I use that I think I paid $5 for it three years ago. I have definitely gotten my value out of it. They just went to a subscription model. I think it’s like $5 per year now. And I’m like, “Yeah, great. I’m glad I can support you and give you money.” When we were preparing this topic, I thought about that Twitter client because I was kind of curious why they’d go from being a one-time fee to a subscription-based other than it’s hard on the app store to make a consistent revenue. And I took a look at Twitter’s API, and Twitter, I don’t know what exactly it was, but recently within the last year, they rolled out a new API for developers and that API is usage-based. So if you’re writing a Twitter client, now you have to pay based on how many tweets you’re pumping through your backend to Twitter. And that to me is a really interesting intersection of a business with a usage-based model that they have to pay into. And then how do they translate that to customers? I mean, obviously a Twitter client where I pay like 10 cents a tweet is not going to go anywhere.


[00:24:41] SY: No. That’s a good point. Yeah. Because what is the cost to the service that you are paying for them to provide that service? And what subscription models are they having to pay and how are they able to provide those services? Yeah. And oftentimes, I can imagine mirroring that, right? If it costs you money every month and you’re subscribing to a bunch of things that are usage-based and maybe you kind of transfer that usage-based cost onto your customers, that might be the way to do it.


[00:25:11] JP: It’s very tricky. I think we’re just learning how much consumers really want to be party to the nitty-gritty. I think one of the ways you see businesses doing this as with bundles. You see, Amazon has their Prime Bundle. Apple has a bundle. Disney has a bundle. All these services have bundles. And instead of getting into the nitty-gritty of like, “Okay, if you pay for Hulu, you pay this amount. If you pay for Disney, you pay this amount.” They’re just kind of like piling everything into just one fee.


[00:25:43] SY: So many bundles. Yeah.


[00:25:44] JP: So many bundles and that’s one way to auspicate the cost obviously and hide different packages every which way. But I think it’s really attractive to customers to say, “Here’s just one fee. It’s the same per month.” And that gives the business a lot of cover. If you’re not watching as much Hulu or are hosting prices for Disney go up or down, et cetera, we can bake that into the one cost and kind of hide it from the consumer.


[00:26:11] SY: Yeah. That is such a good point. And that was one thing actually, I can’t remember what article it was. There was an article talking about the huge issue of churn at the beginning of the year. Once these holiday sales have kind of passed and the promos have gone and the real subscription price kicks in, you have an opportunity to look at your bill and go, “Am I really going to use this for 12 bucks a month? Not really.” And then there’s just tons of cancellations. There’s a lot of churn happening in that industry as well, which is a huge thing to pay attention to.


[00:26:40] JP: So going back to Disco for a quick second, could you envision a model where someone would pay per view or pay per individual course? What are you thinking?


[00:26:51] SY: So I mean, I have thought about doing like if you only want four courses, because part of the thing we need to figure out as a company is just how much do people consume. If we see for example that there’s two customer segments, one customer segment who listens to one course per week. Right? And that’s kind of it. And there’s another one that’s just like consuming a crap ton of lessons and maybe it does make sense to have two different prices, to have one where it’s like, “If you consume just four courses a month, it’s X dollars per month. And if you are going all in and doing unlimited, it’s a little bit more.” So I think that might be an option. But definitely not on a per lesson basis. And that’s the thing about subscription that is kind of nice is like meaning non usage-based subscription is you don’t have to think about it. You can just be like, “I’m going to subscribe to it, put it in the back burner and then I’m kind of done.” Whereas usage-based, it’s like, “Okay, this is going to be another $5. Do I really want to listen to this?” You know what I mean? You’re constantly having to make a decision.


[00:27:52] JP: Which is anxiety.


[00:27:53] SY: Yeah, exactly. Okay, if no one has coined that term, you just did. But yeah, I don’t think that that anxiety really makes sense. And I don’t think it would work for most consumers.


[00:28:04] JP: That’s a good point because I think about the times like the services that do have usage and my cell phone plan used to be like this. I had a bucket of how many gigs of data per month. My cable plan used to be like that before. A lot of these plans went to unlimited, “unlimited usage”. I had definite usage anxiety. I had like all sorts of trackers installed to see how many gigs of data I’d used on my phone, what we’re using.


[00:28:24] SY: Wow! Yeah.


[00:28:25] JP: Because I was freaked out about it.


[00:28:27] SY: Absolutely. I mean, that’s just a crappy experience for users. I think you would like to avoid that if possible.


[00:28:35] JP: Right. Maybe it says something. The usage model for consumers is just more prevalent in utilities, things like, “I’ve been thinking about subscribing to water at my house.” No, you have to have water. You have to have electricity.


[00:28:51] SY: That is a great distinction, kind of the necessities versus the discretionary. I don’t think usage for discretionary would work for most businesses. But yeah, you have one provider for water and you will pay for that water.


[00:29:04] JP: Right. And the internet’s definitely like that as well. Yeah.


[00:29:07] SY: So we’d love to hear what you think about subscription models and usage-based models. So if you want to weigh in on the conversation, you can call our Google Voice, +1 (929) 500-1513 or email us at [email protected] So the COVID-19 vaccine rollout has been a pretty big disappointment, with states having a ton of logistical issues with trying to administer the shots. A part of the problem is the really jenky $44 million website that the US Centers for Disease Control and Prevention paid consulting firm Deloitte to develop called “The Vaccine Administration Management System”. This service was so buggy and ill thought out that it would randomly cancel appointments, lock staff out of the dashboard and ultimately not fulfill its function, really it’s one job of signing people up, scheduling, tracking inventory, and reporting. You can read more about this in a great piece by the MIT Technology review. We’ll put a link to that in our show notes. Because this system and the vaccine rollout as a whole has been so botched by the federal government, it has led to a lot of third parties having to pick up the slack and come up with better solutions. Coming up next, we speak with Cameron Yick, Software Engineer at Datadog, and one of the creators of the New York City Vaccine List, which is one sub-service trying to help people find and get a COVID-19 vaccine. We’ll put a link in our show notes to a couple of resources where you can find similar services in your area after this.






[00:30:55] SY: Are you looking to build a chat for your next big project? Save time building in-app chat, voice, and video for mobile and web applications with Sendbird. Get to market faster with Sendbird’s UIKit, pre-built UI components, best-in-class documentation, and support for developers. Join companies like Reddit, Delivery Hero, Yahoo Sports, and Hinge. Start your free trial today at


[00:31:21] RudderStack Smart Customer Data Pipeline is warehouse-first. It builds your customer data warehouse and your identity graph on your data warehouse with support for Snowflake, Google BigQuery, Amazon Redshift and more. Their SDKs and plug-ins make events streaming easy and their integrations with cloud applications like Salesforce and Zendesk help you go beyond event streaming. With RudderStack, you can use all of your customer data to answer more difficult questions and then send those insights to your whole customer data stack. Sign up free at




[00:31:56] SY: Here with us is Cameron Yick, Software Engineer at Datadog, and one of the creators of the New York City Vaccine List. Thank you so much for being here.


[00:32:04] CY: Thank you for having me.


[00:32:06] SY: Tell us a bit about your developer background.


[00:32:08] CY: So this project actually goes back to what got me started into writing programs to begin with. In college, I became interested in data visualization and I needed raw datasets to work with us. So I became involved with environmental policy research, collecting similar problems to what we actually have here, many different government sites reporting very similar information, but in slightly different shapes, and we wanted a way to have all that information in one place. So as a result, I continued to explore data visualization, learn about Python and various data analysis tools. My first role coming out of school was as a data engineer at a data science company. From there, I moved towards working in a more full stack role, working on the outputs of data pipelines. And then I came to Datadog where I work entirely on data visualization tools for monitoring cloud infrastructure.


[00:33:03] JP: So let’s get into NYC Vaccine List a little. Can you explain what it is and how it works?


[00:33:09] CY: So the motivation for this problem is that right now to find a vaccine, you have to check some tens of different pages and you have to go through a signup form. You have to click through all these different things only to find out that there’s no availability. So New York City Vaccine List has, I would say, two main parts. One is a set of crawlers that go and visit these different government sites. Some of them are government, some of them are from healthcare providers, and they find open vaccines every minute and we put it into a database. And then the other part of this is the front end that surfaces the contents of this database for people to browse. After we started the project, we realized that some appointments aren’t posted online, but it’s still important to find out about them. So we’re working on ways to get data through phone banking or from tips that are submitted through the feedback form and to combine that with the automatically collected data.


[00:34:02] SY: So tell me about the moment you realized that the vaccine rollout just wasn’t good enough and that you needed to do something about it.


[00:34:09] CY: So one of the co-creators, Dan, he went through the process of signing up his grandparents for this process. And it took a very long time. And it got us to thinking about, “How long would it take someone who was not very comfortable with computers to navigate this?” And realizing that there was an opportunity where the technical skills that we had could be applied to save a whole lot of people a lot of time and a lot of stress working on this problem. I think the fact that we experienced those pain points so early made it very straightforward to commit, to spending a lot of time to get an early version up as quickly as possible. One of the early challenges was thinking about scope so that we wouldn’t be overwhelmed, like if we should be New York State Vaccine List, for example, but an example of we’re glad that we didn’t do that was this past week, there was this snow storm and there were select closings in different parts of the state and sometimes conflicting information from healthcare site versus what people were hearing on the phone. But because we were focused on New York City, we were able to get volunteers to contact public health officials, make phone calls to the healthcare providers directly, and be confident in the information that we’re presenting, and we might not have been able to do that if we tried to go straight to state level or a larger geography upfront. So that’s why I feel it’s been good that we’ve focused on New York City to begin with.


[00:35:32] SY: What are some of the other challenges that you face building out this list?


[00:35:36] CY: The constant struggle is figuring out how much functionality to build, how much architecture to be including. There was discussion earlier about what schema to be using, what choice of technology to be using and a guiding principle throughout the project has been to choose the thing that allows us to get something out into the world so we can get feedback on it as quickly as possible, but leave space to grow as time passes. So an example I can give you for this, the first version of the site had no CSS, no styling, no JavaScript. It was just HTML being written by a flask app. And we only brought in React and component libraries and Next.js and other technologies when we felt that they would save us time. And it was actually appropriate to be introducing that complexity. A big area that was pointed out as a blind spot early on was paying attention to accessibility issues. So screen reader support supporting different languages, thinking about using semantic tags, all these sorts of concerns that might not be part of an MVP, but things that we realized were more important as we, one, got in contact with actual users and heard from public health officials, and two, getting feedback from a very large number of people who are senior citizens and expressing that they were having some challenges using the computer. And so we realized how important it was to make sure that we were designing with a broad range of ability levels in mind.


[00:37:05] JP: What were some of the things you worked on with your site to make it more accessible for differently abled people?


[00:37:12] CY: So one of the early things that we did was we used Lighthouse and various accessibility checker tools to do a pass on screen reader support, labeling our forms, making sure sufficient contrast was present for text. We used Tailwind CSS typography plugin to get some good baselines as far as spacing and fonts. The big project that I worked on this weekend was exploring different translation options or localization options. We wanted to do hand tuned internationalization but to get something out into the world to be able to help people sooner. And because the New York City government site also uses Google Translate, we implemented the Google Translate Widget. I would say those were some of the main areas that we worked on recently.


[00:38:01] SY: So can you talk a little bit about specifically what kind of stack the site is working on?


[00:38:07] CY: The app started its life as a flask app with our crawlers written in a mixture of Python and Node.js. The node apps were driving puppeteer because some of the sites required user interactions to get to the data and it goes into a relational database. On the front end, once it started to enter the single page app world, originally, it was built with Parcel because it was zero-config. But as time passed, it became apparent that using Next.js would let us do things like pre-render the site for the benefit of Google and offer a whole lot of optimizations that would allow us to deliver small bundles to people. Something that we felt strongly about to make sure that, going back to the accessibility thing, was that the site shouldn’t feel heavy and be slow to refresh. These people will be refreshing this page quite often, like multiple times per hour. So we went with Preact instead of React and have been watching our bundle size like every time we add new dependencies to make sure that things don’t get too crazy. So the front end is Next.js, React. We’re using Tailwind for utility classes and a mixture of different component libraries for different things. We brought in Semantic UI when we needed to put in a banner to let people know about us, no cancellations.


[00:39:24] JP: Very cool.


[00:39:24] CY: But yeah, we’re iterating and adding libraries only if they’re really necessary.


[00:39:29] JP: What was the biggest technical hurdle or challenge you faced putting the site together?


[00:39:33] CY: I think something that was surprising was involved working with machine translation. So the surprising thing here is that when you use the Google Translate Widget, it wants to take control of the DOM and it competes with Preact for that. So we had this bug where when you translated the site and you have any part of the page re-rendered, the text will show up twice because Preact and Google Translate were competing for the same thing. So I did some rearranging of the sites. So now the site only ever renders once, so that there’s no interference with the Google Translate App. And the part of me that does interactive data visualization was at first worried because like, “Oh, this site can’t be interactive.” But my friend pulled me about the details, HTML element details summary, which allows you to implement collapse and expand behavior in pure HTML without using JavaScript. It made me realize that it is possible to build interactive experiences with pure HTML without using JavaScript and doing that doesn’t just save you bundle size, but also makes it easier to do things like make things translatable.


[00:40:43] JP: The translation feature is very, very cool.


[00:40:47] SY: So one of things we’ve talked about on this show is how people in tech have kind of picked up the slack for the government quite a few times. The most recent example is the COVID Tracking Project. We interviewed the folks behind that and that project has been hugely useful to a lot of people. So I’m wondering, how do you feel about the need to develop things that in a sense kind of should have already been taken care of by the government? Do you feel some type of way about that?


[00:41:14] CY: I think while it would have been convenient if this tool existed already, I’m also not very surprised that it didn’t and it would be nice if there were a clear path to say someone independently develops a tool like this, that there’s a clear path to bringing this into production, like bringing it into government, but I also realize that because we don’t operate in a government environment, there are a whole lot of process and procurement or contracts that we don’t have to go through and that allows us to iterate at probably a speed that that might be difficult in a different environment.


[00:41:52] JP: Do you think developers and other tech folks have a responsibility to pick up these pieces when the government doesn’t come through on them?


[00:42:01] CY: I would say that for me personally, something that I feel strongly about is it’s valuable to apply technical skills to spaces that are important to you, especially because those spaces may not necessarily be the ones that are receiving the most venture capital or funding. I’m a believer in building bridges or transplanting tech from places wherever it may be. It may be tech, it may be design, it may be physics or mathematics, and trying to bring that to spaces that could benefit from them. And when that space that could benefit happens to be a public health issue, I’m very keen to try and bring some of those tools and skills over. I’m not sure if I would compare that to a moral responsibility, but it’s something that has been a driver for me as I pick up new skills and choose projects throughout my career.


[00:42:58] SY: Thank you so much for joining us.


[00:42:59] CY: Thank you.


[00:43:11] SY: Thank you for listening to DevNews. This show is produced and mixed by Levi Sharpe. Editorial oversight is provided by Peter Frank, Ben Halpern, and Jess Lee. Our theme music is by Dan Powell. If you have any questions or comments, dial into our Google Voice at +1 (929) 500-1513 or email us at [email protected] Please rate and subscribe to this show on Apple Podcasts.