By Brant Cooper on January 28, 2015
Some companies adopt Lean Startup to become more agile on the sustaining side of the innovation continuum. Others do so, because they can see disruption on the horizon. Edmunds.com’s disruption had already arrived so for them, business model innovation wasn’t a should, it was a must.
Our November 2014 Lean Lunch featured the President of Edmunds.com Seth Berkowitz. Edmunds.com is an online publisher of automotive information that helps buyers make purchasing decisions. The company has been around since 1966, and got its start printing booklets of automotive specifications for car shoppers. The company launched its website in the mid-1990s, with the same objective—to provide people with car specifications, pricing, and reviews to help them decide which car to buy. Over the years their web traffic grew to an audience of over 18 million people per month, generating hundreds of millions of page views. Car brands represent one of the largest advertising segments in the world. The Internet has been good to Edmunds.com. The recession of 2008, not so much.
Business Model Innovation
The late 2000s brought two huge wake-up calls to Edmunds.com. The first was the auto industry: “Our largest [advertiser] clients, GM and Chrysler, declared bankruptcy and [were working through] restructuring with the federal government,” Berkowitz said.
The second was the introduction of the iPhone and Android smart phones. “The second [disruption] was the shift to mobile. [Today], forty percent of our page views come from mobile devices.” Edmunds.com faced a one-two punch of a significant decline in Internet advertising and a shift to mobile devices that does not support the same advertising model.
“The first thing we did is we said, ‘How do we optimize our desktop and laptop [online advertising] business to give us more time to figure this mobile thing out?’” Berkowitz said.
Instead of just using best practices, which in changing markets are necessarily stale, Edmunds.com implemented lean startup practice to institute continuous learning from their customers and uncover new ways to create value for them. This sustaining innovation depends on testing ideas, just like other kinds of innovation.
“For a company of our size, and for the dollars that were at stake in these decisions, using the lean method to get out there immediately with clients, without something baked, throw it against the wall, get their feedback, come back, do it again and do it the 3rd time—doing this over a 90-day period was pretty quick,” Berkowitz said. “We completely collapsed and destroyed and then rebuilt our ad model between 2014 and 2015.
They even stumbled onto their own version of Eric Ries’ problem and solution teams to think boldly:
“We have this thing called an Ad Summit,” Berkowitz said, “Where we invite six people from six different groups in the company to present for 10 minutes or less on their perspective on the problem as it was framed. Then we have 90 minutes of free flowing discussion, with a notion that we are going to rethink our business model in that context. I don’t think you could do that every week in every area of the company, but we’ve been trying this internal summit on thorny issues since we were happy with the way it worked with respect to rethinking our ad business.”
Out of the summit, Edmunds.com assembled a cross-functional team to address the need for better ad performance. They also had leadership that inspired the teams to solve the problem by experimenting.
Based on conversations with customers, who didn’t always see the same value in web page advertising Edmunds.com thought they should, the team decided to make some major changes to the way they packaged, marketed, and sold their advertising inventory. Between June and September of 2014, Edmunds.com collapsed their 83 retail sales units into just six segments, each of which was defined by a particular car-buying behavior. When Edmunds.com pitched this new segmented model to their customers, it wasn’t immediately embraced.
“We learned that our customers were using rules of thumb that had been developed many, many years ago,” Berkowitz said. “We started to work with them on A-B testing to show them that their ad performance could be improved if they would embrace our new audience-based model.”
Through those conversations and learning with customers, most advertisers were convinced to adopt the new model. However, they did receive push-back from a few “challenger” brands whose goal was to use a “whole page-based” advertising option on Edmunds.com as a messaging hub to their consumers.
Customer Development is not easy
It was frustrating not to be able to convince those advertisers. “I watched my team […] bang their heads into the wall,” Berkowitz said. He told them, “One of the things we need to do is listen.” It was a good lesson for us and part of the lean process is to throw stuff against the wall quickly and then learn. For the challenger brands, we developed a completely different model which created these clear channels for them, that gave them a 100 percent share of voice on this specific set of pages that they wanted to create as their hub.”
The Edmunds.com team then offered both of the advertising models to all of their advertisers.
“Essentially, we gave advertisers a choice: you could buy our 6 segments and we believe that it will yield the highest performance for you,“ Berkowitz said. “But if you’re determined to own a specific area on Edmunds and have that hub, we can offer that up to you as an alternative.”
“I think the biggest thing is that people have to overcome the fear of not having all the answers in front of the client, or saying to them, ‘This is an experiment or this is a provisional plan that we have. Let’s get your feedback,’” Berkowitz said. “It used to be that the thought was everything had to be perfectly polished, then you take it out there, and you didn’t want to show any kind of vulnerability. What we learned is that our clients actually want to work in partnership with us and feel like they’re part of the process.”
Lean Startup & Cultural Transformation
Going through this process in a time of crisis inspired new innovation processes throughout the company as well.
Edmunds.com clearly made some big changes to its business model, and to its innovation processes along the way. But the biggest change has been rethinking its core purpose and strategy. Throughout all this, and through evolving its relationships with car dealers, Edmunds.com has positioned itself as a platform provider that improves the car buying process. Not just providing consumers with information, but changing how buyers and dealers communicate, and providing tools to help the buying decision.
“In terms of the control of chaos, we actually have something that we’re rolling out this month [November 2014], that wasn’t even on our road map at all in the beginning of the year,” Berkowitz said. “It wasn’t even in our mindset. We ran a hackathon, and a company called Car Code won. Car Code is a simple service that enables consumers to send outbound text messages to dealerships. We invited them here for an accelerator program, and at the end of it we bought them. Now we’re offering Car Code for free to our 10,000 dealer partners…We’ve developed the insight that you can’t develop all the ideas and you don’t have the monopoly of all the ideas. We’re really embracing the notion of more hackathons, and APIs for the developer community to build things off of our data that we could then take a peek at.”
Using lean innovation changes expectations of what leadership is as well. Leadership must not be about having a clear vision to simply execute on. It’s about keeping communication channels open, exposing the big problems that have yet to be solved, and demonstrating progress toward solving them. It’s about a willingness to “fail”, being open to experiments that take things in a new direction. It’s about welcoming chaos.
Watch the entire discussion below. Also, tune in Friday, January 30 for our next Lean Lunch featuring Canadian Telecom giant Telus. You will be amazed by their lean innovation story! —>