The ABTs of programmatic

This article first appeared on iMedia Connection

If it ain't broke, don't fix it, right?

Well, it depends on what type of marketer you are.

It's no secret that digital marketers have now embraced data-driven programmatic as the new standard in digital ad buying. Many rely on its targeting effectiveness, consistent performance, and ease of measurement to help round out any traditional media plan. The best technology platforms leverage sophisticated bid modeling to make decisions on how to reach the right consumer with the right message at the right time, for a price that's measured down to the penny -- all based on data. Sounds great, right?

But because of strong results, even the savviest marketers often fall into the habit of depending solely on automated optimization tools, letting the reports do the storytelling without much tinkering.

What's wrong with that, you ask?

Maybe your campaigns aren't living up to their full potential.

Like search optimization before it, programmatic display is as much of an art as a science, demanding experimentation -- and the insights that come with it -- to get lasting results in a constantly-fluctuating market. I have seen the scenario all too often, when an advertiser or agency settles for "on target" campaign performance, when they could be knocking results out of the park.

Now is the time, before code freeze, to test and establish best practices in preparation for the upcoming holiday season. To get you started, here are three testing areas to consider:

Inventory

Is desktop the main contender for your display campaigns? By limiting yourself to only one type of inventory, you could be missing the mark in providing customers a cohesive, cross-channel experience.

Test new channels like mobile web, in-app, video, native, or other programmatic placements. Learn which sources drive the most engagement and ROI, and you may find new ways to reach a new crop of consumers -- like the ever-elusive millennials.

Creative

Creative helps build brand affinity, drive loyalty, and increase sales. With the sheer volume of ads served in most programmatic programs, you can quickly gain statistically relevant results from A/B tests.

Don't just change your creative when your promotions change. Always be testing. There are so many other creative attributes you can test including background images, colors, logo placements, button shapes, product details, calls-to-action, and more.

Would free shipping influence a sale over a product discount? Test it and let the numbers decide.

Do dynamic, product-oriented banners perform better than static for ecommerce advertisers? I believe so, but try it and see for yourself.

First-party data

Your first-party data is a gold mine of testing opportunities. Customer data in your CRM system holds the key to how to best reach different types of existing customers. Perhaps you're aggressively investing in reaching lapsed buyers, when your CRM data might reveal that it's more efficient to engage with recent purchasers.

CRM data also provide insights into the optimal frequency to serve those audiences. Work with partners who can offer frequency caps not only per supply partner but also on a universal and user-individual level. Test to find the optimal frequency, maintain control of how frequently you're serving your consumers, and avoid overspending and banner burnout (and keep your customers happy in the meantime). You can then reinvest those savings to reach even more people.

On the other hand, proprietary data from your website or shop can be used to drive demand in all the right places, helping you fine-tune your bottom-line revenue strategies. Let's say your website sells new and used bikes, but you make a larger margin on bikes sold by certain partners. See what happens to your revenue when you prioritize higher-margin products in your dynamic ads.

Keep it simple

Programmatic advertising is only as good as the people behind the technology. Digital marketers that build strategic test roadmaps are the ones who will see great results.

You don't need to be a rocket scientist to squeeze better performance and insights from your campaigns. Keep it simple. Work with your partners and start with a three-month road map packed with different tactical strategies. Prioritize tests that could have the most impact on your brand's objectives.

With Black Friday and Cyber Monday approaching, now is the time to test-and-learn with different optimization strategies so you're prepared with a best-in-class program when it matters most.

Agree? Disagree? I would love to hear what you think.