6-8 minute read ⏱
TL;DR
Sometimes startups take on client work outside their expertise to keep the lights on. RockitBoost was that project: an influencer marketing platform connecting social media creators with Kickstarter sellers, built by a construction analytics company.
I was solo everything: UX designer, PM, project manager. The product flow I created never changed through the entire project, but everything else was chaos: midnight calls from frustrated developers, a client designer who redesigned the UI five times, and devs who said the core feature was "impossible to build."
I don't believe in "impossible" so I started researching. Turns out Google Analytics could do what everyone said required complex API integrations. I documented it, convinced the skeptical dev team, and saved the project. Then my CEO pulled me off as a reward for solving the hard part. I never saw the final metrics, but I've done what mattered.
CLIENT
rockitboost
ROLE
product manager, ux designer, project manager
year
2020
TAKING ON UNFAMILIAR WORK
RockitBoost connects social media influencers with product sellers on platforms like Kickstarter, Indiegogo, and Amazon. It generates referral links that influencers use to promote products and earn commissions. Simple concept.
But all the commission tracking, click attribution, and analytics was being done manually before. The problem we were solving was two-fold: making it simple for influencers to get referral links and track earnings, and simple for sellers to find influencers and boost sales.
NAILING THE FOUNDATION (THEN WATCHING EVERYTHING ELSE BURN)
I researched Kickstarter itself and realized we were essentially building "Kickstarter for Kickstarter" — same core concept (generate a referral link) but with a layer of commission tracking and analytics on top.
I created wireframes for all the major flows:
Sign-up flows (for all three user types)
Analytics dashboards (commissions, clicks, engagement)
Marketplace where influencers browse available products
I presented them directly to the client. He loved them, and we never pivoted from those flows, they stayed exactly as I defined them initially. Nailing the foundation on the first try felt good.
Then came the UI design stage
The dev team designed something and presented it to the client, who immediately said he didn't like it. So he brought in his own designer to take over the visual design work.
They created the initial UI and the devs started developing it. Then they decided to update some colors. No biggie, that's a quick change. But then they redesigned some elements and completely changed the layouts, fonts, and colors again. The devs were furious because this meant redoing work they'd already completed. And then they did it a third time.
At this point, everyone was frustrated. I was frustrated. The devs were frustrated. And the dev lead started calling me daily at midnight on my cell phone, a number I'd given him strictly for emergencies. Apparently, complaining about the designer qualified as an emergency in his mind.
Eventually, I had to have a difficult conversation with the client about budget. If the designer kept redesigning everything, we'd go over the agreed-upon amount because each redesign meant additional development time and costs. We explained the reality: multiple full redesigns aren't just aesthetic changes, they're entire development cycles happening over and over. My CEO stepped in to reinforce the message, and the client finally agreed to hold the design direction steady.
The redesigns stopped after that.
WHEN "IMPOSSIBLE" ALMOST KILLED THE PROJECT
The ENTIRE concept of RockitBoost was based on referral tracking. The app needed to track referral links to calculate commissions, clicks, purchases, everything.
The developers came back with bad news: "It's not possible to do this without API integrations with Kickstarter, Indiegogo, Amazon, and every other platform the client wants to support." But those platforms don't just hand out API access, I don't think any of them had public APIs at all.
I looked at how Kickstarter and Indiegogo themselves handle referral tracking, dug through knowledge bases and support documentation… Look, I respect developers. But when someone tells me something is "impossible," my first instinct is "Are you sure?"
Turns out, it was just a Google Analytics thing. Here's how it works:
Users authorize Google Analytics (you know that modal that pops up asking you to "Allow" something? That.)
Google Analytics tracks clicks and purchases on the linked platforms
That data gets passed to our app
We calculate commissions on our end
Bam. Problem solved.
The devs didn't believe me at first. So I sent them the Google Analytics articles and Kickstarter's knowledge base on Intercom. Eventually, they saw it: yay, it actually was possible. I'd saved the project 🎉
THE OUTCOME
RockitBoost taught me that sometimes the biggest design problems aren't actually design problems: they're technical and communication problems. The wireframes worked because I understood the user flows and business logic. But the project almost died because of a technical constraint that everyone assumed was impossible to overcome. Solving that didn't require better design skills, but curiosity and refusing to accept "impossible" without proof.
This project taught me: sometimes being a good designer means knowing when the problem isn't a design problem and solving it anyway.
© 2024