5 Rules for AB Testing in Startup Companies

There's a lot of pressure on startups to perform well in their initial marketing and advertising campaigns. If you succeed, you'll earn a wealth of new business, but if the campaign fizzles, you'll have wasted a significant portion of your already-tight budget.
This post was published on the now-closed HuffPost Contributor platform. Contributors control their own work and posted freely to our site. If you need to flag this entry as abusive, send us an email.

2014-12-18-01BuildBrandBlastArticleHeader.jpg

AB testing is critically important when launching a new marketing campaign.

There's a lot of pressure on startups to perform well in their initial marketing and advertising campaigns. If you succeed, you'll earn a wealth of new business, but if the campaign fizzles, you'll have wasted a significant portion of your already-tight budget.

AB testing is one way to alleviate some of that pressure. Rather than producing one massive campaign, you'll start with two small samples, which are similar, but differ in at least one key way. By running these smaller campaigns against each other, you'll be able to identify which elements of your campaigns work and which ones don't.

There are five key rules for AB testing effectively in startup companies:

1.Limit your differentiators. Keep your A test and B test as similar as possible. Run them at the same times, on the same platform, for the same demographics (unless one of these is your key variant). Otherwise, one test may be better than the other, and you'll never be able to pinpoint why.

2.Start small. Don't run both campaigns fully in a live environment. Instead, run them on a smaller portion of your audience, or even better, in a test environment so you don't overspend or tap into your potential lead pool.

3.Ask questions. If you can, follow your test up with a short survey. Objective results are extremely important to determining which version is more advantageous, but the qualitative approach of a survey can give you even more insights.

4.Measure everything. Double check all your analytics and reporting--you'll want to be sure to capture every piece of data you can. It's better to have too much than not enough.

5.Consider multiple rounds. Timing and budget concerns may not always allow this, but if you can implement multiple rounds of AB testing, you'll wind up with a superior end product.

AB testing is exceptionally productive, and can address the problems of your campaign before you even take it fully live. Make sure you start implementing it in all your marketing and advertising campaigns.

Jose graduated from Goldman Sachs' 10,000 Small Businesses program. Goldman Sachs is a partner of the What Is Working: Small Businesses section.

Follow Jose:

Close

What's Hot