Traction: How Any Startup Can Achieve Explosive Customer Growth

by Gabriel Weinberg

One thing I learned at Smart Bear is that I have zero ability to predict what’s going to work. There’d be a magazine where I thought, “This is just some piddly magazine, surely no one reads this,” and sure enough it was cheap (due to small circulation) and it’d do terrifically! Our ROI on some of those were incredible. And you just couldn’t predict, whether on circulation size or media type, how it was going to go. And it changed over time—an ad might be good for a quarter, or a year, and then decay slowly until it wasn’t valuable anymore. It was unpredictable and decayed over time: so the only thing we were left with was trying everything and measuring what worked. TARGETS Run cheap tests by first targeting local markets. It is hard to predict what will work, so it is often useful to run several small offline ad tests in parallel. Each offline ad medium is testable locally. Then you can scale up to regional or national campaigns if warranted. Seek out remnant ad inventory for the highest discounts. You can use remnant ads for both initial tests and scaling this channel. The downside is less targeting ability, both in terms of demographics and timing. Use unique codes or Web addresses to track the effectiveness of different offline ad campaigns. Make sure before you set up tests or campaigns that you can trace conversions back to specific offline ads.