Consumers can choose whether to open an email or catalog, but don’t have choices about viewing online ads. Targeting the right shoppers is critical.

Max Jacob, vice president, client development, Conversant

Max Jacob, vice president, client development, Conversant

As competition heats up for consumer dollars, marketers are focusing on A/B testing to drive conversions according to a poll of client-side marketers conducted by Econsultancy. The poll found that, around the world, 61 percent of respondents were using A/B testing and 31 percent were planning on using it to drive conversions. This is short-sighted. While traditional A/B testing methodologies do have their place in advertising, marketers should be cautious when trying to replicate this strategy in digital advertising.

It’s easy to fall into the trap of thinking that all marketing channels are the same, and as a result apply tactics that work in one channel to another. While channels are similar, people interact with them in different ways.  Customers get to choose when they open or look at an email, bring in the mail or open a catalog; they aren’t afforded the same choice with online advertising.

When you’re delivering personalized messages to consumers, A/B testing goes out the window. The focus instead shifts from the presentation layer to the types of conversations brands are having with consumers. Because what works for one consumer will not work for all.

Imagine you’re a retailer that specializes in women’s clothing. Winter is approaching so you show a woman as series of ads for three items—a dress, a jacket and blouse—based on her profile. She purchases a dress after seeing the blouse ad. Does that mean the blouse ad was not successful or does it mean the entire customer journey was a success since she made a purchase? The end goal of advertising is revenue so, objectively speaking, this scenario would be a success.


Obsessing on the order of the messages—a dress, jacket or blouse or a blouse, dress, jacket—misses the point.  The order of the messages might not matter.  Every consumer is different.  The right order for one might be the wrong order for another. Remember, delivering the right message at the right place and time is part of a broader journey. The goal is to inspire your consumers to engage with your brand.

The same is true for evaluating the effectiveness of different ad sizes. Ad sizes represent different opportunities to connect with buyers and prospects. Consumers may be on one device or several at the same time. It’s our job as marketers to get in front of people at the points where they are most likely to engage.

If you focus on the individual, you can’t go wrong.

Rather than worrying about testing creative order or ad size, concentrate on finding the right individuals to message, and stitch together the most appropriate customer journey for each of them. Explore if it is better to message consumers who bought in the last six to nine months or those who bought in the last 12 to 15 months. And then compare the return on ad spend (ROAS), or performance, so you can determine which audience helped move the needle and which fell flat so you can make adjustments as needed. The easiest way to gain this insight is by utilizing a tried and true methodology that works across all channels: test and control. By separating customers into either a test group, those who receive your ads, or a control group, those who see an unrelated PSA [public service announcement] or other message, you’ll gain insight into customers who are likely to make a purchase.

In the end, if you focus on the individual, you can’t go wrong. No matter what type of product, service or idea you are trying to sell, concentrating your efforts around individuals is the most effective way to realize ROAS and overall marketing ROI.

Conversant provides personalized digital marketing technology.