A Year of Product at Creative Market
In October of 2015, I stepped into a product management position at Creative Market. In the last 12–14 months, we’ve completely restructured the team, fundamentally changed how work gets done, dramatically reduced project timelines, and integrated user feedback into our workflow at a much deeper level than ever before. We’ve also built some fantastic stuff.
Redefining How We Work
At around 30 people, we’re a remarkably small team when you consider what we’ve accomplished and built in the last few years. Even at our small team size though, we’ve experienced some significant growing pains. Tripling your team from a group of 10 to a group of 30 is a double-edged sword. You have way more resources at your disposal, but you also learn the hard way that processes built for 10 simply don’t work well for 30.
The Old Way: Triage Project Model
Historically, we only had a single project team: a handful of designers and developers who were assigned work in a sort of triage model. Our single project manager assigned project resources as they became available to the most urgent tasks. This works great at an early startup scale and it allows you to work quickly to accomplish the most critical tasks.
The more resources we threw at the triage model though, the more strained it became.
Here are some of the issues we encountered:
- We started juggling too many disparate projects simultaneously
- Designers and developers were always pulled off large, long-term projects to put out fires, squish bugs, and build last minute promos.
- Context switching is a very real threat to productivity. When you have designers and developers stepping away from big projects, even for a day or two, getting back isn’t immediate. They have to reaquaint themselves with the work and get back in the right head space, which costs days. Do this multiple times and you can bet on weeks being added to your timelines.
- Talented people with valuable insight and opinions started to become cogs in the machine. Small teams can work through everything together, but in a fast-paced triage model with lots of resources, you end up with product or project managers scoping projects and designers/developers simply taking the work that gets assigned to them without having any opportunity to help shape the scope in a meaningful way. This is by far the largest offense. Not only does it reduce the overall staisfaction of your team, the product work suffers greatly as a result. Great things happen when you invite the people who build your projects to the planning table.
In light of these issues, Priya Kothari (our Operations Manager) and I set out to kill the triage model and rework how things are built at Creative Market.
The New Way: Cross-Functional Project Teams
Our solution was to reorganize our project resources into two small, cross-functional project teams: the platform team and the conversion team.
My team, the platform team, focuses on building the core product value and user experience, things like large-scale new features and equipping our sellers with the tools they need to succeed.
The conversion team focuses on things like funnel optimization, A/B testing, scaling marketing efforts, etc.
We’re continuing to grow and evolve, but each project team started with:
- One product manager
- One designer
- One frontend engineer
- One backend engineer
Our goals for creating these teams were simple:
- Keep the teams small and lean. Allow them to figure out how they work best and don’t bog them down with outside requests.
- Eliminate the problem of project roadmaps coming from executives on high. Execs should set goals and each cross-functional project team should be free to build roadmaps and prioritize projects of their choosing to meet those goals.
- When you start a project, you see it through. No more getting tossed around from project-to-project, starting a million things and never finishing them.
- Every member of the project team has a seat at the table for every step of the project, from roadmap planning to scoping and through building and launching.
- Every member of the project team has a voice. For example, if you’re a backend engineer, you’re present for the design reviews and can call out UX problems in a comp just like a designer could. Egos are checked at the door and we all collaborate to make every part of the project as good as it can be.
Important Note: The platform team includes Noah Stokes, Seb Armand, Jesse Street, and myself (shout out to Ryan Weaver who also sat on our team for a while). Any time you read the words “we” or even the word “I” throughout this piece, know that it was a team effort. These guys (along with their counterparts on the Conversion Team) build the things that makes Creative Market awesome.
Fixing Runaway Timelines
We went through a period where everything felt really slowed down and stuck in the mud. Projects that were small and large alike were all competing for resources, and timelines across the board were far beyond what they should’ve been.
In light of this, one of the things I sought to fix right away when we launched the platform team was the runaway timeline issue. We now complete large-scale projects in weeks rather than months. Here’s how we did it:
Everyone Scopes
I mentioned this above. It’s not fair to hold people to deadlines when they didn’t have a say in scoping out the work. Everyone scopes the project together, everyone knows what’s involved, everyone commits to a realistic timeline.
More Roadmap Ownership = More Timeline Ownership
When designers and developers are treated as grunts doing tasks, they don’t feel the weight of missing a deadline. They miss a deadline, we push it back. No big deal. Only the people with a full view of the roadmap were really feeling that pain of other projects getting pushed back.
With cross-functional teams who build their own roadmaps, this dynamic changes. Everyone on the team knows what’s ahead and feels the pain of shifting everything else back when something isn’t done in time.
Commit to Deadlines In Front of The Rest of The Staff
We have a Monday team call with all local and remote staff. I make it a point to talk about our deadlines and what we’re shooting for in these calls, and to point out when we’re behind. The entire project team feels that. We don’t want to let each other down, and we don’t want to admit to everyone else that we screwed up, so we dig in and work hard to complete our work on time.
Always State Your Current Deadlines in The Daily Standup
Everyone does standups, and just about everyone I could find asks roughly the same questions:
- What did you do yesterday?
- What will you do today?
- Is anything blocking your progress?
This is a useful start, but I added two more:
- What deadlines are you working against?
- How confident are you that you’re going to meet those deadlines?
The guys on my team got clever and simplified these two questions into one statement by using emoji in Slack to represent their deadline confidence.
💚 = On track and confident I’ll hit my deadline
💛 = I’m a little behind, but I think I can make it
🔴 = I’m not going to make it
So a typical morning Slack standup from a team member might look something like this:
- Yesterday: Completed V1 Dashboard comps
- Today: Dashboard V1 Comp review at 9am, then I’ll start on V2 comps
- Deadlines: Wrapping up dashboard V2 comps by Friday morning. We’re a bit behind due to the holiday break, but I think we can still finish in time 💛
- Blockers: I have to take my car to the shop so I’ll be out for about an hour around 11am
Project Sprints
The biggest impact to timelines was adopting a sprint framework for projects. There are some important things to note here:
- We use a homegrown sprint model that evolves based on how we work best. We have process check ins, and when something isn’t working, we improve it.
- We’re not agile. We’ve picked up a couple of useful agile methodologies, but we don’t subscribe to the strict agile religion.
- We also don’t strictly follow the GV sprint model, though we did all read and love the book, and continue to pull important lessons from it.
- We try to limit sprints to three weeks.
- Project research and discovery is done outside the three weeks, often heavily by me and the lead product designer as the rest of the team is on the home stretch of wrapping up the previous project.
As we hit our groove, we’re finding that we can pull off most project ideas in three weeks. Keep in mind projects were often taking multiple months before, so this is a huge improvement!
That being said, we’re not 100% fixed on three weeks, and ultimately let the problem definition and project scope decide. We’re wrapping up a project now that we decided to do in two weeks becuase it was smaller than our usual projects. Especially large projects get separate three week sprints for design and development.
Working Prototypes Over Flat PSDs
As a product manager, I hate seeing web work represented as a flat comp. A folder full of PSDs or JPGs shows me pixels, not product. It forces me to imagine a user flow. It doesn’t let me feel a user flow. Nevertheless, that’s how it had always been done at Creative Market. No matter the project, we always jumped straight to pixel-perfect PSD work.
When I came on as product manager, I challenged our designers to change this. With the help of InVision (or Adobe XD as it continues to get better and better), we now build clickable prototypes for almost everything we do.
This has helped us drastically cut down on the number of UX “gotchas” that are typically not unearthed until the development stage when we have live pages to work with.
Making User Feedback Critical
This is one of the accomplishments that I’m most proud of in all my time at Creative Market. Historically, we built things with users in mind, but we did so on a mostly anecdotal basis. We’d check to see if there were any support tickets related to a project, maybe browse the forum, but not much more than that.
With my first big project, a new dashboard for Creative Market shops, I knew user feedback was absolutely vital to our success, but there was no formal system in place for me to get that feedback.
Fortunately, my wife (Kelley Johnson) is the community manager at Creative Market. Her entire job is working with and talking to our users! So we teamed up and built a user feedback process.
- Core User Feedback Group: A growing group of over 100 users who have volunteered to answer surveys, share their opinions, and test features for us.
- Surveys: We use simple Google Forms to survey users before, during, and after projects. We often start with the user feedback group and cast a wider or more targeted net from there as needed.
- Prototypes: We’ll send along a link to an InVision or XD prototype, give the user a list of basic tasks to perform (without instructions on how to perform them), then ask questions about their experience.
- Recorded Google Hangouts: We watch and record a few users working with the prototype or early build. The community manager hosts this and is there to provide help, while the members of project team are simply watching the private stream without being able to interact. It’s good to have someone not on the project team run the interview (this is one of the tips we pulled from the Sprint book). The community manager is in a more neutral position to offer assistance without leading the feedback in a given direction. It’s SO frustrating to watch users stumble through something you built, clicking all the wrong things while you just sit there, unable to jump in. This is a good and necessary frustration though and it leads to much better product work.
Choose Your Own Adventure
In truth, this is more of a toolkit than a process. We pick and choose from this list as needed. For the dashboard project, we used all of them, but for smaller projects, one or two often suffices.
Getting user feedback takes time, and then that feedback will inevitably add even more time to your project as you fix all the problems you see users encounter. But it’s the single most valuable thing you can add to your project process.
No Surprises On Launch Day
Hearing loads of negative feedback when you launch a big project sucks. People are angry and confused that you would build something so frustrating, and you’re embarrassed that you missed seemingly glaring UX issues or bugs. It’s terrible.
If you move all of that to before launch though, the people spotting problems aren’t angry or ranting publicly about your incompetence, they’re partnering with you and know that they’re a valuable part of the process. When you release to the world, you can be so much more confident in what user reactions will be and can be at ease knowing that you smoothed out all the major concerns before most of your users ever laid eyes on it.
Some Cool Stuff We Built
We built tons of great things in 2016. Here are some key highlights. Where possible, I’ve tried to show the progression from early wireframes to finished product so you can get a feel for how our ideas evolve over the course of a project.
Shop/Affiliate Dashboard
Creative Market shop owners are a dedicated and hardworking bunch. For many, Creative Market is their primary or even sole source of income. When you consider this, our simple table of recent sales started to feel pretty inadequate. These are small business owners and they need to be able to view their shop statistics in an informative way. They need to spot trends, notice sales spikes or dips, learn what products get purchased the most, and see their progress charted over time.
Our new dashboard is a treasure trove of information and really helps shops dig into what’s going on with their sales. It also offers context-sensitive education and tips that help you at every step of the shop journey, from creating your first product to building a dedicated following.
Shop Updates
Shop Updates accomplishes two things for our users: it gives customers a place to check out what’s new and on sale from shops that they follow, and it allows shops to directly market their products in an unprecedented way right on the site.
Shops post new “Updates” daily and customers see them in a constantly updating feed. It’s basically our own social media feed where our shops can talk directly to the people who have expressed an interest in their products.
Product Rearranging
In 2016, the top requested feature from our shops was the ability to manually rearrange the items on their shop pages. Historically, every product simply appeared in order of date published with the newest at the top. We rebuilt shop pages to be interactive with both drag-and-drop rearranging and fast, automatic sorting.
The challenge here was to build an interface that worked well for shops with only a few products and shops with hundreds of products. We accomplished this by allowing shops to select multiple products to drag them around or even quickly send them to any page.
Pay It Forward 2016
One of my favorite things about working at Creative Market is our yearly Pay It Forward campaign, where we raise money for natural disaster victims all over the world. This year we partnered with our amazing shop community and raised over $40,000 for All Hands Volunteers to help people in need. Our staff also flew to Louisiana to spend a few days helping with some demo work for flood victims.
Here’s to an Amazing 2017
Working as a product manager at Creative Market has been an unbelievably rewarding experience. I’m honored to work with such talented people who are dedicated to seeing our community grow and succeed.
We have huge plans for the year ahead and I couldn’t be more excited to be a part of it!
We’re always looking for amazing people to join the Creative Market team. We value our culture as much as we value our mission, so if helping creators turn passion into opportunity sounds like something you’d love to do with a group of folks who feel the same way, then check out our job openings and apply today!
