Pixly

Pixly

Mar 2017 – Apr 2025 9 min read

Building Pixly: from a cafe to acquisition

In 2016, Gustav and I were taking lunch walks around the ironSource office, riffing on business ideas, and we kept circling back to the same problem. Every gaming company on the planet (Supercell, Machine Zone, GameLoft) was competing for the exact same users on the exact same ad networks. Facebook CPMs (the price of 1,000 ad impressions) were through the roof and Google was getting more expensive by the quarter, with all these brands fighting over the same male 13+, English-speaking mobile gamer and driving each other's costs up.

So Gustav and I looked at each other and basically said, "What if we took everything we know about performance marketing and applied it to influencer marketing?" Content creators had massive, loyal audiences, people who actually trusted their recommendations.

That was the whole pitch. We called it Pixly. I started working on it in mid-2016 while I was still at ironSource, left in February 2017 to go all in, and we founded the company in March 2017, with Gustav as CEO and me as COO. We started grinding out of Workshop Cafe in San Francisco with two people, no funding and no safety net.

The first move was Gustav's cold email to Machine Zone in late 2016, pitching what he called "Direct Influencer Campaigns" (no deck, no warm intro, just a note from a no-name two-person shop). While he worked that side of the door, I was already on the phone with TUNE/HasOffers, working out the tracking and attribution integration I knew the pitch would eventually rest on.

Workshop Cafe, where Pixly started

The Battlecats (where it all started)

Our first real test was a mobile game called The Battlecats, published by PONOS. Gustav and I spent six months building a database of YouTube creators by hand, going to each channel and entering its name, URL, subscriber count, average views and engagement ourselves. We analyzed tens of thousands of influencers one by one, looking for the best creator-to-cost ratios we could find.

We landed on an emerging creator called Reaction Time, who had never done a sponsorship before, was priced reasonably and had an audience that matched perfectly. That first video took off, with tens of thousands of clicks, real conversions and strong ROAS (return on ad spend). The flywheel clicked: viewers became players, players became spenders, and spenders funded more marketing budget. Behind Reaction Time, the early Battlecats roster filled out with Guava Juice, MessYourself and LaurDIY, emerging creators who each brought a different audience, and every one of them went into our hand-built spreadsheet before they ever saw a contract.

We had proof this thing worked.

The attribution moat

The part of those first months I rarely talk about is the unglamorous work that made early publishers trust us. While Gustav was meeting creators and closing campaigns, I was deep in the TUNE/HasOffers integration, building tracking links, postback URLs (the server-to-server ping that tells the advertiser an install happened), install attribution and conversion windows. That plumbing turned a creator video into a measurable performance channel.

Influencer marketing in 2016 mostly ran on screenshots and vibes. So when a publisher asked us which video drove which install and we had a real answer in real time, that accountability became our moat. Repeat business comes from making the path from video to click to conversion repeatable, and we could show that path.

Almost didn't make it

Pixly nearly didn't survive its first summer. When I left ironSource to go full-time on Pixly, the plan was that Gustav would follow a month later. Then he and his wife found out they had a baby on the way, and suddenly there was real pressure to make this work, and fast.

The company wasn't making enough to support the two of us. Gustav told me straight up that if revenue didn't pick up soon, we'd have to close up Pixly and take jobs elsewhere. So we scraped together what we had and flew to the Postback marketing conference in Seattle (July 2017) to meet potential clients.

That trip changed everything.

The Fortnite break

At Postback we met the team at Epic Games. They were working on Fortnite Save the World, a game six years in development, and we were tapped to lead the influencer program for its launch. Then, one week before launch, Epic called to say they were pivoting the whole thing to a free-to-play Battle Royale mode with a way bigger budget, so we reworked our messaging on the fly.

Fortnite launch

At first influencers didn't want to touch it. Cartoon graphics, a weird building mechanic, and everyone was already playing PUBG. So we focused on shooter creators (the PUBG and Call of Duty crowd) and convinced them to give Fortnite a shot, starting with names like Fitz and NoahJ456 at 50 to 100 influencer activations a month. Then viewers started demanding more Fortnite content. Creators' Fortnite videos were outperforming everything else on their channels, and some converted their whole channels to Fortnite. We scaled to 500 influencers a month, brought on Wilbur Soot, eventually MrBeast Gaming and dozens more, and expanded beyond shooters into Minecraft, Roblox, lifestyle and vlogging.

The other deliberate wedge was geography. While every other agency was knife-fighting over the saturated English-speaking audience, we pushed hard into Latin America, MENA and non-English European markets, promoting Fortnite mobile around the world.

By January 2018 we'd hit 1 million Fortnite installs. Fortnite was our first major breakthrough, and it put us on the map.

1 million Fortnite installs

Building the machine

In 2018 we hired Berkeley interns to keep up with the volume of activations and brought on Zack as CTO to build out our tracking software and performance analytics. Monica joined later that year, Garrett came on in 2019 to lead sales, and from there came Austin, Josh, Mitch, Jackie, Ewic, Morgan, Ryan, Michael and others.

As COO I built out the backend of the company: finance, HR, onboarding, documentation and platform automation. Gustav and Garrett ran sales and campaign management. We automated workflows so we could stay lean, and grew from 2 co-founders to a team of 15 that operated like a team of 150 (a small crew punching way above our weight).

Garrett built our DTC (direct-to-consumer) business from scratch. He sourced, negotiated and closed the deals that pushed us beyond gaming: MANSCAPED, Keeps, Bespoke Post, Warby Parker.

From there things snowballed. We landed Opera GX, Opera Browser, House Party, Genshin Impact (miHoYo), CASETiFY and NordVPN, and the roster kept widening with Drink LMNT, Incogni, Honkai Star Rail, Kingdom Maker (Scopely), Star Trek Fleet Command and Fall Guys as we expanded from gaming into brand and DTC.

By the time the dust settled, Pixly had activated more than 10,000 unique influencer videos and racked up more than 8 billion views across campaigns.

The operational reality

Bigger business, bigger problems. By 2021, running hundreds of concurrent activations meant constant fire drills: payment routing breaking, tracking links going missing in the middle of campaigns, PayPal refunds, wire reconciliation, creators going dark mid-deliverable, and a steady drumbeat of "URGENT" subject lines in my inbox. The team grew because the operational surface area grew. That's what scaling a services business looks like in practice, and we got better at it the same way anyone does, by absorbing the chaos until the systems caught up.

Going remote

When COVID hit in 2020, we moved out of our downtown San Francisco office and went fully remote, and we actually thrived. The team adapted fast, and some of our biggest influencer programs (Opera GX, House Party) ran during that stretch. Opera GX in particular became the flagship of that era and the most repeatable case study we had. The systems we'd built held up.

The acquisition

In June 2022, Acceleration Community of Companies (ACC) acquired Pixly. ACC was building a community of specialized agencies: MKG for creative, Pink Sparrow for design and fabrication, Stripe Theory for data and analytics. Their client roster included Pepsi, Target, Nike, Ralph Lauren, Meta, HBO, Delta Airlines and T-Mobile.

ACC acquisition announcement

That network is why we chose them. Our performance and influencer expertise complemented the brand agencies they had brought together, and their deep brand marketing experience opened up a whole new world for us.

Michael Nyman, CEO of ACC, called Pixly "a proven leader in the influencer marketing space" and "the perfect agency to add to our community." GP Bullhound served as our financial advisor on the deal.

Gustav summed it up in the announcement:

Most advertisers lack many of the necessary components required to build successful influencer campaigns and we're thrilled to bring our technology and capabilities to ACC's growing roster of leading global brands. Joining ACC is an incredible opportunity for Pixly to collaborate with other best-in-class agencies as well as ACC Advisory's team of marketers, cultural insiders and category experts. The GP Bullhound team was instrumental in helping us navigate through this process with ACC.

Alec Dafferner, Partner at GP Bullhound, framed the bigger picture:

With brands increasingly looking for a more personal relationship with customers, influencer marketing has accelerated. As a result, we expect to see continued investments in technology and services that allow brands to access and connect with their customers via influencer marketing.

Right after the acquisition we launched Pixly 2.0, adding brand marketing for Fortune 500 brands alongside our performance work.

What we built

The part I'm proudest of is that we bootstrapped the whole thing with no outside capital. We went from the two of us at a cafe to 8-figure ARR (annual recurring revenue) with a lean team of 15, and managed more than $100M in influencer spend along the way.

We helped build a new category, performance-driven influencer marketing, by bringing the accountability, precision and relentless optimization of performance marketing to a space that was running on vibes and vanity metrics. We built a proprietary system that gave brands real-time access to insights and the KPIs they cared about. The influencer market exploded while we were building (by 2022, 72% of marketers were using influencers, up from 17% in 2019), and we were right in the middle of that wave.

The relationships are what I'll cherish most, with the hundreds of talent agencies, gaming publishers and brands we worked with over the years. People like Monica, who joined in 2018 when we were still a handful of people and grew into a senior partnerships leader. And Colin West at Solaro, who said our work was "instrumental to this industry."

That matters more to me than any revenue number.

The Pixly team

Stepping away

In December 2024, I told the team at our all-hands that I'd be stepping back. My wife and I had our daughter, my mom was struggling with her health, and I wanted to be fully present for both.

It wasn't an easy decision. 9 years of your life doesn't just switch off, but it felt like the right time.

I spent the first few months of 2025 hiring, training and onboarding Morgan, our new Operations Specialist, to take over my operational responsibilities, and handing the rest to Gustav and the team so the machine kept running without me in the seat. March 31 was my last day.

When I left, Pixly was thriving and poised for an even brighter future. For me, it was time to focus on family.