Going from a kitchen table to 100,000 customers, an interview with Dave Hayes, VP of Product at Sentry

Going from a kitchen table to 100,000 customers, an interview with Dave Hayes, VP of Product at Sentry

Richard Huffaker
Director of Creative Marketing
October 27, 2020

This is the first edition of Welcome to the Shortcut, a new series where we interview Product and Engineering leadership from our customer and partner companies. The order in which we do these interviews is random, but that randomization is based on a process in which our Director of Creative Marketing travels to a Yosemite National Park, hikes to the top of Half Dome, yells every potential interviewee’s name as loud as he can, and then hikes down and interviews the first person to yell back, “yeah!?!?”

These interviews are meant to provide a glimpse into what it’s like to lead a software team during all of, well, all of this (gesturing to the world around us), while also exploring the topics of going from a kitchen table to a 1000 person company, putting yourself in your customers' (and potential customers' shoes), training for a marathon by doing sprints (not ideal), and playing every video game available on the Switch.

For this week’s edition, we spoke with Dave Hayes, VP of Product at Sentry and a very early employee at PagerDuty.

David Hayes, VP of Engineering at Sentry
Dave Hayes, VP of Engineering at Sentry

Let’s start this off by talking a bit about your career. How did you get into Product?  

I came into Product kicking and screaming. At my last company, PagerDuty, where I worked for 7 years, I was originally hired as an engineer. Everyone in the company could fit around a kitchen table back then, and basically everyone was an engineer. But at some point you make the transition from building a product for yourself, where you’re scratching your own itch, to building something for a hundred thousand customers.

I helped in all sorts of areas and eventually landed in product management for probably the last three quarters of my time there. We took PagerDuty from that kitchen table to being a public company.

And how did you go from PagerDuty to Sentry?

One thing I did at PagerDuty is manage partnerships. We were data-driven, so, naturally, we had a leader board of sorts for what software our customers were using, and Sentry was one of the rising stars.

When I first met them Sentry was probably four or five people back then (Note: they have over 120 employees now). I liked everybody. They bought me a lot of drinks and, more importantly, continued to climb the leaderboard.

So when a recruiter pinged me about being their Head of Product, all the pieces came together. I really enjoyed the early days of PagerDuty, and relished the opportunity to go back and essentially be a coach for myself using what I’d learned from the last time.

What’s a lesson that you’ve taken from PagerDuty and are now applying at Sentry?

PagerDuty and Sentry had a near identical genesis in that they were both created to scratch their founders’ itches. PagerDuty’s founders had spent a lot of time on call and didn't like any tool they used. Cramer and Chris (Sentry’s co-founders) needed a good error tracking / monitoring tool, so they built one that they and their friends could use.

The biggest lessons for me are about making the transition from being your own target customer. Our CEO sometimes jokingly calls me the Head of Product for the most boring company from the S&P 500 he can think of. I'm not going to name any of them because by the time this goes to print, who knows, they could be our customer. But it’s important to champion people who may not have the same way of solving problems as you do.

An obvious example: Sentry is an open source company. Hugely popular among people who prefer to work with an open source product or who do a lot of research, identify the best product, and swipe their credit card. But there are certain steps that go into you buying software at large companies and I'm also championing those people. Our engineering organization might be very excited about languages like Rust and Go, and I'm excited about those languages too, but I’m equally excited about Java or PHP or .NET. Things that you might see more in the Fortune 500 than at TechCrunch.

To build software that applies these lessons, how have you handled planning with engineers spread out to their individual homes? How can you communicate and throw ideas around the room, even though no one can ever be in the same room together?

We've been getting much, much better at planning and a lot of that has happened during these COVID times. It's a little bit hard to disentangle the effects of us getting better from the WFH changes. We brought along some great managers. We've tried some things that haven't worked. Tried some things that have worked and obviously, we've evolved and we've gotten better.

One of the cliches of software engineering, you see it in Silicon Valley, you see it in any movie representation, is a couple of nerds around a whiteboard.Obviously, that's a very real loss. One on one like we’re talking right now is very easy. You ask a question, I reply, but in heated debates where we're actually, let's say there's four of us who all have ideas firing off quicker than our mouths can move. Suddenly this quarter of a second lag in the best case, we're talking over each other. I think, actually, you're frozen, which is--


There could not have been a more perfect time for my connection to slow down.

A productivity hit did not show up in the numbers that I track, but everyone felt in their guts that our communication obviously took a hit in March, in April, but one of the things that's worked really well is that we've gotten much better at clarifying what our asks are. It's no longer, "Hey Richard, I bumped into you while we were getting lunch and I want to talk about an idea."

You have to be both intentional about the question you're asking, but also intentional about who's in the conversation. Let's say you and I happen to have the same lunch hour. Maybe we're both early lunch eaters. You'd see that we have a lot more communication, even if we weren't the best people to talk about a particular issue. When I have a problem I have to think, "Okay, it's Richard and Tiffany and Mona that I want to talk to about this particular issue."

Other than Slack and Zoom, have you found that there have been any tools that you use in your daily work that have now become more valuable?

This is something that probably overlaps with what I was saying, that we've gotten better at our process over the last year or so. It feels like it's accelerated because, naturally, I can no longer come into a room and convince you by being louder than you. There's no way for me to be louder than you on Slack. One family of tools that I've seen more of is obviously ticketing tools.We actually use a mishmash of four different ticketing tools, because, again, you need to communicate what the ask is, and the ask needs to lead somewhere.

This has always been true, of course, but we've convinced ourselves that, "Oh okay. It looks like Richard took notes in this meeting. I'm sure he will do the thing that we agreed on in this meeting." You can no longer convince yourself of that, so it really needs to have some kind of artifact. Often, some of the communication happens outside of the artifact, but we generally do a good job of linking to it.

The other tool that we've seen a lot more uptake on are analytics tools. This was not distributed evenly at the beginning of the pandemic, but by now I would say the majority of people who are trying to convince people of a particular action are pointing to data to do it.

A lot of our stories are essentially, "Hey, the curve looks like this," or "We didn't know that new users used the product this way, whereas users that have been here with us for a year or more are using it this way." Over the last six months, the number of aha moments that are happening, period, have increased dramatically. The number of people who are capable of generating an aha moment has dramatically increased, and this is a good change, in almost every way. We've gone from getting to a whiteboard and coming to an emotional decision, to somebody has to intentionally say "I think we should do X because of Y." Clarify X. X goes into the ticketing system and Y is usually a graph in Amplitude or another analytics tool.

<!-- Code for CTA component starts here-->

<div class="post_body_cta"><div class="post_body_cta-heading">
Enjoy your work
</div><div class="post_body_cta-text">
Project management software should be helpful, not a hassle.
</div><div class="common_buttons_wrap"><a href="/signup" class="button primary text-color-white w-button">Get Started - free forever</a>
<a href="https://api.app.shortcut.com/oauth/signup" class="button is-google w-button">Sign up with Google</a></div></div>

<!-- Code for CTA component ends here-->

Sentry's a tool for engineers who tackle sometimes very large problems as they happen. How has Sentry seen it's own customers adapt to a world in which engineers dealing with bugs and other issues can't point at each other's screens anymore and talk to each other?

One thing that I would change about what you said there is Sentry's a tool for software development teams, right?

Obviously, the majority of a software development team is the engineers, but the personas that we try to solve problems for are those engineers. The engineering managers, the engineering leadership, and adjacent people, like either product managers or like a support liaison or even a really technically savvy sales engineer.

Like Neil and Vu (two of Sentry’s really savvy sales engineers).

I think that this is one of those cases where, fortunately, we get to be our own customers again in the sense of the same problems that we're seeing, our customers have seen. We're adapting to the shift in much the same way. If you look, we're spending more calories on integrations for exactly all the same reasons.

There has always been an oversimplified in person work pattern of: support notices a problem, convinces me it's important, and then I start working on it. Was that ever the right way to solve anything? It’s not the way that we would have ever recommended you solve it at Sentry. Sentry detects the problem and prioritizes and triages and helps you assign the problem internally. For the way it's impacted our customers is basically business as usual only more so, and everyone's a little bit more stressed.

One of the things that we've invested a lot of calories in is simplifying the new user experience. Because again, so many things that just companies or businesses were able to get away with when there were friendly people sitting beside you, become a little bit harder now. The idea of, "I'm a new user onto this team, I don't know what this means." When I say this, I'm instinctively moving my hand in front of the screen that you can't see, showing you exactly what I mean. This is not an easy time for anybody. We've had to take that into consideration. It hasn't dramatically changed where we're going, but it has prioritized or deprioritized a few things.

Sentry just moved to a new HQ in San Francisco, but has the pandemic changed Sentry's notion of work from home? Or is it going to be that once we can all go back to work at an office, everybody at Sentry is going to go back to working from the office?

I can't speak for Sentry as a whole but I can tell you, though, that it changes my thinking. I was a firm believer, and I have been for my whole career, that PM is an on-premise job. PM is connective tissue, and I want to put the connective tissue closest to where it's needed. The evidence has not backed up my assumption. My personal belief is that I still think there's a world where we need that high latency whiteboard discussion for complicated things, but I definittely need to reevaluate some of my priors.

We didn't mean to, but we are all training for how to work in this world right now. Every company has unwillingly invested in this long term training program on how to work from home and it would be crazy if that didn’t have a long term impact. It's not only our company, but everybody you hire, everybody you interview has gone through the same training program.

We have three offices: San Francisco, Toronto, and Vienna. We've always had a little bit of asynchronous work going on, and the people who are the connective tissue between those offices have often had to work outside of nine to five to make it work. But it was all predictable. Vienna and America have different daylight savings times, so they would shift by an hour twice a year, but that's not a lot of randomization.

It’s like you were training for a marathon by doing sprints.

This has been especially hard on parents. I've taken meetings at 7:00 AM. I've taken meetings at 1:00 AM. I've had meetings where somebody's child had a bathroom emergency and the meeting ended and we just had to resume it later. I'm definitely worried about burnout. This can’t be forever or, if it is, we need to understand that parents are going to be very negatively impacted by it. You can imagine a world where we figure out childcare first, and then we figure out our work spaces.

That's actually another thing, a little less so in Toronto and Vienna, but definitely in San Francisco, our median employee did not have a lot of extra space to set up a full-time home office.

I know I don’t. I feel very lucky my partner is a chef in a test kitchen (a kitchen that can be run from home), which means her office is naturally on the opposite side of the apartment.

How do you feel like this has impacted PMs in general?

A friend of mine was talking about a project that they were working on and it has a PM on it who gets their way by bullying people. While they're attempting to bully people on Zoom, the actual meeting is happening in Slack and they're not in it! That would not be possible if they were all in a room because they would notice that everyone had physically walked out. I don't know how the project's going, and I strongly suspect that the PM doesn't know either.

I think we're going to see little changes that add up to a lot of situations like that. Where we had weak processes or people who were not team players, not moving the ball forward or missing some vital skill for the role, we're able to cover that up through bullying or the sheer force of their position or personality. I have more questions than answers, but those are some of the trends that I'm looking out for.

What would you say to a PM like that to help them be better, if there's even anything you could say?

That's a doozy of a question. I mean, I won’t hire bullies. So let me answer an adjacent question instead, which is what has recruiting and hiring been like? It's been a shock for some people hiring people sight unseen. Something like 20% of Sentry's employees have never physically been in the Sentry office. I assume that that's similar at Shortcut and other fast-growing companies. There are a lot of interviewers, and I think we're getting better at this, but historically there've been an awful lot of interviewers who pick up on irrelevant signals based on something ridiculous like the strength of handshake. I don't care about that. I hope you don't either, it's not relevant.

From one angle the interview process is almost automatically fixed, but I think in general, it's a general management problem. Most people, definitely myself included, are a bundle of things that don’t map perfectly to responsibilities of the role. There are things that they excel at, things that they're a little bit behind where they or their boss would like them to be. I haven't even met this particular PM, only heard the complaints, but I don't think it's any different than a normal managerial challenge, I hope.

Alright, so what are you up to when you’re not leading product management at Sentry? How do you-

Oh my. You told me this question was coming and I don’t have an answer.

How do you keep yourself entertained?

I’ve been playing a lot of video games. Also I love my Peloton. I didn’t realize I’d love it, but I double extra love it now.

I miss rock climbing. Actually, let me bring you back. One of the reasons that I went into product management is because I like to be a part of everything. I genuinely enjoy everything from sales tax conversations to what a user conference might look like and everything in between. I’m a failed renaissance man in my hobbies and so the thing that I’m looking forward to is getting back to my to-do-list, which is getting both longer and more urgent. It’s like, "I want to learn to scuba dive," just because I can’t right now.

I want to learn to fly a plane, although that one I know I’m not going to do. I really do want to get back to all the dilettante things like travel and stuff like that.

I would also love to do those things. I also would be interested in learning to fly a plane, but I will never do that.

I took one lesson, it’s not fun. Have you ever driven a motorcycle?

Yes, I have.

It's loud, it’s bumpy. It feels unsafe, but it’s also complicated and you’re in a tiny little can and you’re nauseous the whole time.

What a great combo! I love being in a plane. I love traveling around, but I think I’ll probably stick with being able to drink cocktails while I’m doing it. What are a couple of your favorite games you’ve been playing lately?

I’m super jazzed for Breath of the Wild 2, whenever that comes out. I might have literally every single Switch game, but the game that I’m spending the most time in is Beat Saber.

I have an Oculus Quest. Do you have that or are you a---

I have a Rift. Specifically because you can get custom songs on it.


Switching the topic away from VR, check out Sentry for Application Monitoring and Error Reporting. One million developers at over sixty thousand organizations already use them. Those folks are probably on to something.

And be sure to use Shortcut if you don’t already. We’re modern Project and Product Management, without all the management. Lead your software teams with Shortcut!

<!-- Code for CTA component starts here-->

<div class="post_body_cta"><div class="post_body_cta-heading">
Give us a whirl
</div><div class="post_body_cta-text">
Because you shouldn't have to project manage your project management.
</div><div class="common_buttons_wrap"><a href="/signup" class="button primary text-color-white w-button">Get Started - free forever</a>
<a href="https://api.app.shortcut.com/oauth/signup" class="button is-google w-button">Sign up with Google</a></div></div>

<!-- Code for CTA component ends here-->

No items found.
Share this Shortcut story
Enjoy your work
Project management software should be helpful, not a hassle.
Check out our words
Shortcut is modern project management without all the management. 
And this is our blog.
Read more stories
No items found.