Generally I'd suggest that testing should be done AS you're launching - and running - your campaigns. Of course it depends on how big (and how much you're investing) in a particular campaign. If it's a Super Bowl ad, then try to get as much focus group feedback beforehand.

But if it's social, direct, email, etc... - then the testing should be incorporated into the campaign itself & focused on continuous refinement and optimization. Test lists, test offers, test subject lines, test landing pages, etc... Use A/B and multivariate testing methods to continuously optimize your campaigns.

Matt Heinz Replied on Jan. 5, 2012

Totally agree with this. Jim Collins addressed this in his new book Great By Choice when he talked about bullets and cannonballs.

Simply put, a bullet is a test. It’s done to learn something with speed and minimal risk in mind. After a few successful bullets hit their mark, you conceptually have enough data or market validation to fire the cannon instead.

Collins, backed up by years of research, demonstrated that companies who grew 10X over several years during difficult market conditions did so by firing a lot of bullets, all the time, before carefully choosing their cannonballs. Similar companies who stalled or faltered too often bet big on the cannonball first, which took more time, attention and resources away from the business with a lower rate of success and return.

You aren’t going to grow at high rates year after year without firing cannons. At some point, you have to make the big bet.

But there’s no reason to do that without first testing the market, validating your idea with data, and ensuring your cannons will hit their mark.

Put another way, spend most of your time at target practice. Figure out what works. Then put the pistol away and get out the big guns.

Mike Turco

Consultant, Mike Turco

Posted on Jan. 4, 2012

I really like Chris Selland's answer. To add my own thoughts, you should build testing into your marketing plan. For example, if you place an ad online, you can send the people who click on that ad to a special page on your website. That way you can see how many people clicked through. If you're really good at it (or work with someone who is), you can go further and see exactly how many of those people who clicked through actually bought something. In broader terms, this is called advertising tracking.

This way, if the ad you've placed doesn't pan out, you can cancel it right away.

Not everything is this easy, though. For example, if you shoot out a press release with a link to your home page, that can be really hard to track. Sure, you can look for an uptick in hits, but its not an exact science.

Mike

Chris Selland Replied on Jan. 4, 2012

thanks!