A conversation with Dan Williams from Instacart

During this week’s Product to Product episode, Amy Chyan, Product to Product podcast producer and host, chatted with Dan Williams, Senior Product Manager, at Instacart.

Dan shared some hurdles he encountered through his various product management positions, from working on the creation of a new product to transitioning to working with enterprise clients. He highlighted how he solved for user problems when developing a new product and how he adjusted his product management working style for enterprise clients.

We highly recommend watching Dan’s full talk but if you’re tight on time, we’ve pulled out some highlights below.


(The highlights have been condensed and edited for clarity.)

Congrats — you’re a Product Owner now! (2:15)

Amy: Dan is currently a senior PM at Instacart, and he spent four years at WorkMarket and over five years at Kobo. Dan, can you give us a quick three to five minute summary of your background and how you got into product?

Dan: Yeah, absolutely. For me, it goes pretty far back, I would go as far back as university. I was in the computer engineering program at University of Waterloo, which has a co-op program, and I got to try out various roles through that process. And so it spanned QA roles, a little bit of dev, and some other roles around project coordination and management side of things. And what I learned through that process was, I like the tech space obviously, but I really enjoyed being in the middle of projects, not just on the QA, not just on the dev, and I liked helping communicate requirements over and why we were doing certain things and if anything, helping shape those requirements where possible. And so it wasn’t product management, but it was just that idea, I didn’t even know what product management was at the time, but just that idea of being in the middle and being able to see all players and what everyone is doing and how it all came together. And so that’s what I learned through my co-op.

Then I started at Kobo, and what was good about Kobo was that we had a great QA manager, Carolyn, that gave QA a voice, and we were allowed to go beyond our role. We were allowed to make suggestions, we were there to represent the users. And so we just continued to do that as QA. We’re very technical as well, so we understood how changes would impact things downstream. And so I was QA on various teams there, from the website, to a little bit of mobile and then ultimately on our content management system team, which is where all the content came in and processed before it went live on the site to sell and would respect all those rules that publishers wanted to follow. And so through my QA role, I became QA lead, I started doing a little bit of the product ownership side, where I would help create the stories. I made sure that there were signs of developers and manage that side of the backlog alongside the existing PM there.

And then the opportunity arose at one point for that role to be taken on by me. And it was a little bit funny, so I was already doing the job. We were in a meeting where our VP RD was giving updates on teams, and I hadn’t been talked to at that point, but a slide came up where it said I was a product owner for that team now. And that’s how I found out.

Amy: Congratulations.

Dan: All right, I’ll start doing the job properly now, I guess because I wasn’t really sure, I was just taking on little bits of tasks before. And so I would say that was my first phase into product management was that product ownership role. And it was pretty straight forward to start because we had one main stakeholder, our publisher operations team, which represented the publishers and did a lot of management of books on our platform and pretty much what they felt was the most important next thing to work on was the most important thing to work on.

Adjusting to working with enterprise clients (7:48)

Amy: So for those who don’t know what WorkMarket is, what is it?

Dan: WorkMarket is an enterprise software as a service, and it helps clients effectively manage their on demand talents, like freelancers and workers, who can even find them, even send them work and pay them and do it in an efficient and compliant manner. So at a very high level, you have your enterprise clients and you have your workers on the other side.

Amy: So you mentioned that at WorkMarket, you transitioned into working with enterprise clients. How did you have to adjust your product management work style with enterprise clients because they’re pretty different from just regular users?

Dan: So when I moved to the enterprise side, I joined some ongoing research around how companies manage their talent. And it was just about wrapping up, and we were going to look at the insights and I also had the advantage of, a new team was going to form based off this research. So I had some time to understand what was coming out of this research and how to best approach it before this team was actually formed. It’s a huge advantage and privilege there. We saw some major objectives that we wanted to achieve out of that research, and how it was more about, okay, how are we going to execute on this the best way possible with enterprise clients? And traditionally, what I had been told before taking on this kind of work was that enterprise clients are not open to being tested on and freelancers, the workers, they were very open to it.

We can AB test on them, we can try new things with them, we can move faster with them, but they said, “No.” They’re not as open to that. They want a lot of lead time if a change is going to happen, their processes are very sensitive. If they don’t feel like work can get done, that’s a big problem for WorkMarket, we’re stopping what that company is trying to achieve. And so I felt like a core part of this though, was just showing them that we’re genuine about what we’re trying to do, we’re really trying to identify and solve their problems. And so part of that was showing them the research results and what our understanding of it was and what insights that we have and what objectives that we would want to try and achieve and why these ones were prioritized over the others.

This is what we’re hearing the most, this would solve the biggest problems for you and do you agree? And then from there we asked them if they want to continue to be part of this process, if we have prototypes or mocks created next, do you want to see them and provide feedback? And once we went down that path, they started being more open to providing feedback and wanting to be part of this process because they see that we’ve identified a need of theirs that we’re trying to solve. So we put these mocks and prototypes in front of them, we would continue to ask them at the end of every session, do you want to continue to be part of this next step, this beta program. It’s probably going to take X amount of time, but we’ll stay in touch in any case.

And this went all the way through to the actual production release. So that was some of the changes we wanted to make were very dramatic ones, where we weren’t sure if it would honestly work until it was out there. So one thing that we did that we’ never had on the platform before, was to create an opt in experience, where let’s say you’re managing your talent, you go to that screen the way it is in its current state, but a banner would appear at the top and saying, “Hey, do you want to try this out?” And we only released it to the clients that said they wanted to see this first. And then on a per user basis, they could opt in to the experience, try and complete their task and if they couldn’t, they could opt back out and when they opted back out, we prompt them for feedback on why they’re opting back out.

And sometimes it could have just been like, “Hey, I’m just not ready to fully test this yet. I want to complete my task and then maybe I’ll come back later.” Or maybe we’re missing a field that’s really important to them to accomplish their task. But I think a lot of the time, what we heard was they wanted to accomplish something, they said we didn’t have it, but we knew we did have that feature, and so it was more about a discoverability issue that we needed to improve on. So you never know what you’re going to learn sometimes that way.

Process of creating a new product (13:55)

Amy: So going back, you talked about when you got to WorkMarket, one of the projects you worked on was to create something new, which was the mobile app, and you had the privilege of a new team, new frameworks, new structure. Can you tell us about some of the process of creating this new product? And I guess, how do you make a decision on what to include, what not to include?

Dan: Yeah, that’s a great question because this was a new space for all of us on the team, and so part of it was just understanding what the product was already. There was a lot of features on the website, and so we had some data that we could look at to see where the freelancers were going on the website already. So actually, we had a little bit of a guide. We knew there were three main areas that we wanted to have the app achieve, and that was finding work, getting the work done and getting paid. If someone fell within those buckets, they became candidates to be in that first release. And we first included everything, we made an estimate and it was way too long. And they said, “Oh, it’s got to be shorter than that.” So we had to ruthlessly prioritize what will go in and go out.

And so it was various features that we needed to pull out based on data of how often they were being used and how effective it would help you achieve those three aspects of finding work, finishing work, getting paid. Finding work being the most important, getting paid was actually pretty straightforward, but finding work was very, very sensitive. So there were two main ways that you can get work out on the platform, and one is, there’s a public marketplace companies can post jobs and you can go find them as a freelancer, apply and get the job. There was another way where you’d be directly invited to a job. So a company may have a previous relationship with you or have you in a group or something, but some way they’ve invited you directly. And that way was the more effective way of getting a job by far.

So we decided not to put the marketplace in the first version, since it wasn’t an effective way of getting a job. That’s what the data said. So we released on Android first, that’s where most of our users were. And we just saw the rating starting to tank on the Google play store-

Amy: The data got it wrong.

Dan: Or we used the data wrong, something, but we got it wrong. And the rating side of the tank, we knew we were going to put the marketplace in.

Amy: Was it within a week of release? What’s the timeline?

Dan: It was pretty immediate. Yeah, let’s go with within a week. And it continued. So we knew immediately, all right, prioritize the marketplace feature and it would definitely need to be in the first version of the iOS app as well. And as soon as we put that feature out, people started rating it a lot better. iOS has always traditionally had a high rating, I think it’s in the 4.5, 4.6 range. But when we thought back on it, we were wondering, why did we get this wrong? We do a retro and everything that we do, and there could be multiple reasons, but in general, we’re a company called WorkMarket and people probably expected a marketplace to be there. That was probably a major one. And just the traditional expectation of how you get a job, is the job boards and listings of jobs and things like that.

And the app was well suited in that first version, for very mature users on the platform, people who are already established, getting work on the platform, getting invited, it was working fine for them, but for a lot of users that are newer and looking for that first job or those first set of jobs, then they’re not really sure what to do, the marketplace isn’t there. They may not suddenly get an invite if they haven’t filled their profile with the right information. So we learned from that and we fixed it quickly for Android and we knew not to make that mistake for iOS.

Audience Q&A Period

(17:59) Can you talk more about the work you had to do to enable the opt in function for enterprise users to test a new workflow? In practice, I found that while we want to do this, it’s complicated to do for one piece of a workflow or to be able to navigate back and forth within task between the new and old experiences.

(19:24) What are some other challenges you’ve hit as a PM? And what were some of the major learnings? I think, as a junior PM or an APM, you just have to go through those things. What were some of those for you?

(21:40) How do you balance the business needs and product/customer needs?

(25:28) What new problems in product do you think will be introduced with faster iterations?

(26:28) How often do you bring users in to test your product and do you only test with high fidelity prototypes or also with low fidelity prototypes?


Hi, I’m Dan!I have been a Product Manager for a little over 7 years and have always enjoyed solving problems, finding out what people are truly trying to achieve and creating experiences that make it easier and fun to do.I have recently joined Instacart where I work on the enterprise side. With a focus on mobile, I’m helping to create white-label e-commerce solutions for grocery retailers. Previously, I worked through exits at WorkMarket, a SaaS platform to find, manage and pay freelancers and Kobo, an e-reading platform.

Amy’s a Content Marketing Specialist at Roadmunk on the Marketing team. She produces Recess, the Product to Product podcast and video content. Prior to Roadmunk, Amy worked as a journalist in various Canadian newsrooms and wrote for publications like NBC, CBC, Vice and more.You can follow her via her website, Twitter, and LinkedIn.

Product to Product | RoadmunkFor product people, by product people.Follow1

Try Roadmunk for free

14-day trial | No credit card required | Get started in minutes