Validating new customer segments: The campaign experimentation lifecycle
It can—and will—take more than just one try to evaluate the efficiency of a given campaign. Don’t stop with a single test.
Finding, validating, and targeting new audience segments is a powerful growth strategy. Great marketers innovate by experimenting with new audience segments to discover high-potential audiences for their brand.
The starting point of the process, on which much of its propensity hinges, is the audience analysis effort that goes into delineating a high-potential segment. Sifting through the data, establishing the weight that needs to be allocated to different audience attributes, balancing reach and homogeneity, and formulating an understanding of the segment’s unique needs and leverages is—as the cliche goes—an art and a science, and a prerequisite to the success of any campaign. Once such a segment is defined and established, targeting it with the right campaign can work magic.
And yet, over and over, marketers assume that customer acquisition campaigns that are not immediately effective are a waste of time and money. They give up without refining their methods, ultimately depriving their brand of potentially profitable new customers.
Campaign iteration is key
The main reason customer acquisition campaigns often fail to produce early on is that they're affected by many factors, including:
- The size/scope of the targeted segment
- Proper (or improper) matching of messaging/images with the targeted segment
- Bidding issues
- Ad format issues
- Misalignment between the audience and the product/service being promoted
These are just a few of the possible reasons why your campaign might be falling flat, but you shouldn't jump to the conclusion that an audience segment won’t ever work for your business because, for example, a particular Facebook campaign performed terribly in your experiment.
One failure shouldn’t stop you from experimenting with this target audience again, immediately or in the future. Any good customer acquisition campaign takes a page from the iterative design playbook. Far from being a "one and done" proposition, the most valuable, profitable campaigns undergo an iterative process like this:
- Define your audience segments
- Test a variety of campaigns
- Review your results
- Adjust your tactics
- Build upon your wins and shore up your losses
Iterative design works on a simple principle—software designers assume their first attempt at designing a piece of software is going to have issues. They assume their will be a huge potential for improvement. Just as writers don't rely on the first draft of a novel to be perfect, designers don't assume their first attempt at creating the next Facebook is going to be perfect.
We've only listed five possible issues above, but it could be any combination of those issues (and others) that are causing your initial failure.
That's why iteration is so important. Your first campaign is quite literally a science experiment. Your hypothesis is that these particular images, these specific words, this product or service, in this ad format, with this bidding strategy, is going to strike a chord with this audience. If it's the case that, for example, the image is the problem, but your guess on everything else is right, then giving up right away could prevent your brand from tapping into a huge new source of revenue.And even if you get a campaign mostly right, that doesn't mean a few tweaks couldn’t have a huge effect. It's so common it's practically a running joke that slight changes in the microcopy of a button or CTA can have bizarrely and absurdly enormous impacts on conversions. The same is true in the world of customer acquisition campaigns.
Use data to refine your efforts
The hard truth about creative ads is that the design of imagery, videos, and messaging usually comes from the gut intuition of the creative in charge of creating it.
The job of marketers is to use the results of those gut feelings as a base, and then to test, optimize, and tweak, building on your performance (or giving up and trying a different approach after sufficient testing makes it clear you're not making headway).
However, before giving up on an audience segment for good, try these fixes:
- Change your creative and messaging: Try altering your creative, messaging or both to better match the target audience. Even slight changes to copy can have enormous impact. A different image or video (or even video thumbnail) could be the difference between an ad that flops and an ad that garners clicks.
- Bids: Try changing your bid-optimization strategy. You never know what the right amount really is until you hit it.
- Promotion: Your promotion may not be well-matched to the target group’s lifecycle (or may be too high, too low, or even entirely inappropriate—some audiences respond better to no promotion at all).
- Ad format: Videos generally perform better than static images, and certain types of images generally perform better than other types, but every audience is different—switch things up if your format seems to be flopping.
- Channel: Once you’ve ruled out all of the above, take a step back and ask yourself if this audience segment is better suited to a different channel.
The Phar Lap moral
Iteration means it can—and will—take more than just one try to evaluate the efficiency of a given audience or channel. Don’t stop with a single test.
Remember the Phar Lap moral: the famous Australian racehorse came from a distinguished racing lineage, and was expected to excel like his forbearers. But when his moment came, he finished dead last in the first race he ran, and failed to place in his next three races.
Thankfully, his owners and trainers understood the importance of continued investment. They changed his training regime, and concentrated on learning what makes him tick. Eventually, he won the largest prize ever offered (in 1932) when he set a track-record time in the Agua Caliente Handicap, and went on to become one of the most prized horses in history.
Take what you’ve learned from your first experiment, make some adjustments, and then test again. On the other hand, going back to the Phar Lap metaphor, make sure you’re not flogging a dead horse. If despite your efforts you’re still not getting the results you’re expecting, move on. Stay in touch with the data. Learn when to double down and when to walk away, and always keep experimenting. You never know what you might miss if you give up too quickly.