The path to building AdXpose — which sold to comScore today for $22 million — was definitely not a straight one. With startups, it rarely is. (I recently gave a presentation on this topic at Casual Connect, which you can read here).

Backed by Ignition Partners, the company was founded in 2004 under the name Mpire by Dave Cotter and Greg Harrison. It started out as an eBay seller tool. Fantastic product, but too small of a market.

I joined the board in 2005 and got to see first-hand the trials and tribulations of a team trying to navigate a crowded market. When I joined, we basically started over.

Here’s the back story on that transformation.

We took all of our knowledge around shopping data and built a shopping search engine with analytics on when to buy items — not unlike Farecast (now Bing Travel) and Decide.com.

Matt Hulett

Desperate, and without any revenue to speak of, we started to do work-for-hire jobs for eBay. That experience led us into the advertising business.

We started by building data widgets for eBay, which became the underpinnings of an ad network. That led us to create an ad network around shopping. None of us had built an ad network before, and we really didn’t know how to run one.

We launched the WidgetBucks ad network, learned a lot, made mistakes, iterated, and improved. But we quickly recognized a problem: there were too many ad networks and we were concerned about punching through the market noise.

While we built a huge network of publishers and bloggers, we weren’t making the big dollars we had hoped for.

That led to a transformation where we created ads so that they could dynamically optimize themselves.  We collected information on where the ads were placed on the page; what text was around them; how they were performing; and what creative was working.

This emerged for one simple reason: We didn’t have a large team.  Necessity was truly the mother of all invention.

WidgetBucks exposed massive amounts of data that publishers and ad networks had never seen before.  They were intrigued (and maybe scared).  Eureka.  A new product was born and, yes, another business model pivot.  This being the fourth and final one.  (We internally joked that we changed the name every year so that we would win industry awards).

We launched AdXpose as a software-as-a-service product for agencies, brands, networks, and exchanges.  CTO and founder Greg Harrison built the team and had a crazy passion for providing truth and justice in advertising.  (Some members of the original team from 2004 are still at the company.  And, before I left,  I hired some fantastic business leaders in CEO Kirby Winfield and CFO Jeff Bergstrom who both know startups.  These folks took AdXpose to another level).

As we all learned at Mpire/WidgetBucks/AdXpose, most startups don’t actually start with the product or service that they end with.  The key ingredient to success, in my mind, is a great team.  And that includes investors, which in the case of AdXpose included Bill Bryant from DFJ; Mike Vernon of Zulily and Jeff Lanctot (former chief strategy officer of Razorfish).  In addition, we had some notable board members like Rob Solomon (who just recently left Groupon) and Michelle Goldberg and Rich Tong of Ignition Partners.

There certainly were a lot of twists and turns along the way.  But after multiple pivots and transformations, we got there, and comScore is lucky to have AdXpose.

Matt Hulett is the former CEO of AdXpose and current president and senior vice president at RealNetworks’ games unit. He blogs at the Startup Whisperer.

Like what you're reading? Subscribe to GeekWire's free newsletters to catch every headline

Job Listings on GeekWork

Find more jobs on GeekWork. Employers, post a job here.