We have updated our Privacy Policy and Privacy Options

Got It

5 Impossible Things You Can Do for Your Clients (If You Test)


Sometimes it seems like understanding the demands of a client is impossible. But we have found that a marketing campaign that involves testing can make even the most seemingly impossible demands possible. Here are five impossible things you can do for your clients if you are testing:

Read customers’ minds.

A common request by any client marketer is to read their customer’s minds. Of course it is extremely valuable to understand a customer’s wants, needs, anxiety points, etc.

This isn’t impossible. It is possible to read customers’ minds without the use of wires connected to brains, soothsayers, or mediums. The answer lies in asking the right questions and analyzing the data.

We start off every client relationship with an Understanding Phase that involves lots of questions to understand their business. Our audit includes hundreds of questions all designed for us to understand the current landscape of the client’s needs including competitive research, current data streams, and customer list health. We frequently see some clients have already done the research on this, but agencies forget to ask.

From those answers, we develop profiles of the target customer segments most likely to convert. But those personas are still only assumptions. Through iterative testing, we sharpen our demographic information and form a crazy-awesome, high-converting, highly profitable target persona.

Improve success metric by 50%.

We frequently hear of clients with unreasonable expectations for lift numbers. The problem is not in the % lift expected. Potential lift numbers greatly vary by conditions, but if you’re continually optimizing your conversion pages, any number is achievable. The problem is most of the time the metric they want to improve on isn’t the right one to worry about.

The best way to handle this is to reframe the goal. The key to hitting your goals (and meeting expectations) is to truly understand the value of each metric point. This will create an understanding of what is the most important metric goal to strive for. For example: You could test a landing page designed to get more clicks to the next page in the shopping cart. But if the real metric is sales, not clicks, you could be spending too much time optimizing to the wrong group of people.

Consistently improve.

It is tough for everyone to see a losing test. But every test is a good test. Failure to beat the control simply eliminates that test element as a viable way to increase conversion.

Don’t base the success of any relationship with one test. At Brooks Bell we understand you can’t win every test. We track every detail of every test so that while the specific test may not increase conversions, the information gained will help fuel the next test.

It is important to remember that a losing test doesn’t mean you aren’t consistently improving. By keeping track of overall lift numbers you can show the value of testing overall. We provide an analytics dashboard for our clients, which gives them Year-To-Date numbers to show lifts throughout the year. This has been incredibly valuable in pointing out consistent improvement to the metrics that are most important.

Take big risks… without all the risk.

Big risks by their very nature are scary because they are risks. That bold new design or edgy copy might hurt conversions big time. One way to control that risk in email programs is to batch test to a small segment of the total list. If your big risk bombs, you’re not hitting the full list and lowering the overall impact. But if it wins, you can roll out that new look with confidence.

Get peace of mind.

So, now you’re reading customers’ minds, improving success metrics (not just once, but consistently over months or years), and taking big risks without all that pesky risk… Finally, peace of mind can be achieved because of an optimized process, a systematic approach to testing, and trust in the data.

And that’s priceless to clients and agencies.