Deep Learning for…Marketing?

If you’re doing any marketing, you’re probably using some form of Marketing Automation tool, something like Constant Contact, Eloqua, Marketo, and so forth. These tools aren’t just cool, they’re necessary. I mean, how else are you going to keep track of your user interactions and interpret the firehose of data that you have? All this, mind you, on top of the constant arms race that you’re in when it comes to filling your marketing funnel, maintaining brand awareness, etc!

The tricky bit here is that there are just an insane number of factors involved. Things like changes to the business climate, fiscal year variations, customer behavior around holidays, user dissembling, etc. these are all factors that can theoretically be tracked and maintained.

That said
• The rules go stale really really fast (probably faster than you can code them)
• The ground-truth around the rules can change in a blink (“new tariffs drive down customer demand” etc.)
• Your own biases will affect your interpretation of the data
• There are factors that you aren’t even aware of.

The point here is that you need marketing automation to work with the world the way it is, not the way you believe it is. And that requires that the rules reflect the “truth on the ground” as it were.

This is where Machine Learning comes in. Using it for data interpretation allows it to infer patterns, trends and relationships in your data — in ways that you may not have expected.
So yeah, we may not quite be at a place where we have customized flows for everybody in the lead funnel, with real-time data analysis leading them through tailored purchasing decisions. But hey, it’s a start!