Interaction Optimization vs. AI: A Three-part Series
Part 3 of 3
Why Man and Machine Make Digital Marketing Better
In the last blog, we’ve established all the problems machines can solve by themselves, what does that leave for the human marketer?
As it turns out, plenty.
Machines Make Humans Better and Vice Versa
For years, the time and focus we spent on testing and optimization left little bandwidth for anything else. With machines and specifically self-driving AI and ML doing the heavy lifting, we can finally get back to the strategic part of marketing that brought many of us to the field in the first place. Instead of spending our time creating and deploying campaigns, we can actually leverage the huge strategic insights that the technology gives us, to make smarter decisions and investments, guide which channels and or segments to focus, and change the business in ways that delight and create loyal customers.
Does that mean we’re totally off the hook when it comes to testing and optimization? No, of course not. Machines are smart but they’re not completely autonomous. AI can’t decide which things to optimize around or create the combination of assets that constitute a personalized experience. It can only take what you’ve given it and try to do the best that it can.
That brings me to one final client story. The client, a large hardware and home improvement site, was using our tech to help their customers find DIY inspiration, as well as relevant home improvement or maintenance products and services. Now, this client was incredibly happy with how things were going, but there was one sub-segment for which the AI and ML always underperformed. It turns out this sub-segment, which was made up of single mothers, was one the client hadn’t accounted for with its content. So, no matter which combination of assets or promotions the machine served, it could never find one that appealed to this demographic. On first glance, this might seem like a machine problem, but it’s really one for the human marketer. Without the machine, the business may have never known about this potentially valuable segment, but absent human intervention, the former wouldn’t be able to do anything about it.
In the end, the client decided this was too important of a group to ignore and developed a strategy and accompanying marketing assets and promotions to serve it. With the new assets in place, the machine was finally able to do its job and find the right mix to please this audience too.
If there’s a lesson to be learned from all this, it’s that AI and ML, like all technology, is just a tool. And while, like all tools, it will inevitably surpass some of those that came before it, it has very little value without humans with the expertise and knowledge to wield it.
In my next blog, we’ll move beyond optimization to look at how infusing AI and ML into marketing is helping businesses advance in other ways, such as with auto-segmentation and increasing loyalty.