[Technology] New tool uses AI to predict ad performance

Estimated read time: 3 minutes

Software as a Service (SaaS) certainly is a highly competitive field and it’s hard to rise to the top.

Many companies offer similar things and the differentiators can be small. Our green button is better than our competitor’s blue button. #snark

And then we have the never ending struggle between conversion and branding campaigns. I recently shared this on social media:

“The funnel is a mess,” – Sandra Fancher “There are branding campaigns and conversion campaigns and they aren’t interchangeable.” Book time here to talk conversion strategies.

The struggle is real: Ads get designed without calls to action, with unreadable text and unclear messages.

And then when nobody clicks – who knew they were supposed to? – we throw up our hands and blame somebody else that it didn’t work.

But of course there’s a method to the madness. The problem is that it is now a moving target. What worked yesterday may not work today.

And a branding campaign isn’t a conversion campaign and vice versa.

Read next: How Strategic Use of Color in Web Design Can Increase Conversions

There certainly are reasons to run one or the other. For example if I am re-branding I want to make sure to get my new brand and message in front of as many relevant people as possible

Branding campaigns can be considered top of the funnel as they make sure we know about a brand and hopefully we remember it when we need a certain service.

Conversion based campaigns can be at different parts of the funnel because even at the top of the funnel conversions happen.

For example, when people click to do something which in this stage of the funnel includes actions of further connectivity as opposed to buying something. Examples include:

  • Signing up for the newsletter
  • Registering for a webinar
  • Leaving a comment with a question

And to get people to do things we try to create useful content and creative that drives engagement and the action we’re looking for. Of course, we often do that with our experience and what we might describe as gut feeling.

“That ad seems like it could work.”

Of course that’s a highly subjective comment and can be debated and is debated by people.

With that in mind I was intrigued when I saw the news of the use of artificial intelligence to predict performance based on creative and audience.

IgnitionOne now uses AI to predict a display ad’s performance — before it runs, according to Marketing Land.

I should say that I have no relationship with that company and I’ve not tried the tool yet.

But to truly move to a data- driven culture using predictive tools is super helpful.

In the traditional model we plan a campaign and then run it and then look at the results. Similar to what I discussed during the Publishing Forum at Content Marketing World.

Everything gets measured but it mostly gets measured after the fact. And then we try to use what we’ve learned-and sometimes that’s hard-in the next round of content creation.

Sometimes it’s nearly impossible to know why something actually performed well. In those cases good luck trying to replicate that.

Using tools that help us predict performance based on the varying variables for a project can be super helpful.

Of course, we need to remember that there’s a cost involved as there is with any Software as a Service.

But, if it can help us drive results quicker it might be a worthwhile investment. Whether it’s with that company or another.


Here is an expanded audio version of this blog post that also was live streamed on Twitter initially.