Algorithms Are Cool, Except They Don’t Work For Marketing (Heuristics Do)

Marketing, especially online, is more closely related to doing things based on heuristics, as opposed to algorithms.

It’s more about “Playbooks” than strict and rigid Operating Procedures (we’ll talk about how to fix that, too).

What does that actually mean?

Let’s Define Algorithms And Heuristics

An algorithm is a set of (well-defined) instructions for carrying out a particular task. It’s, for the most part, deterministic, predictable, and not subject to chance. It works for all cases and gives a (presumably) correct answer.

Lots of people approach marketing this way, especially with the lure of “Big Data”.

You look for predictability in buyer decision-making and behavior, so you can scale a campaign.

You follow a set of marketing “rules” that seems to have a history of generating results. It aims at making sense of, and controlling, a chaotic, unknown, and disorderly environment — that of buyers (clients, customers). If you can “figure out” how to market to them, then you can end up with a set of instructions that always produce results.

This approach is probably grounded in human psychology and how you can manipulate (negative connotations not intended) and influence it.

It does fail to recognize that human psychology is, well, human psychology.

We barely understand how our brain works so most efforts at generalizing human behavior into a set of rules and variables that you can play with, is largely futile  and might account for 7% of what is really going on in that brain of ours.

A heuristic, on the other hand, is an experience-based technique that helps in problem discovery, learning, and solving. It basically helps you come to a “good enough” solution — close enough to the best, optimal, solution. It’s like a set of educated guesses or “rules of thumb”.

If you approach marketing this way, you have a greater chance of success (in generating leads, sales, or what have you). Why?

Because marketing is, inherently, unknown and disorderly.

The Failure Of Algorithmic Marketing

I know engineers-turned-marketers (growth hackers) feel like everything can be reduced to algorithms, predictability, and numbers. It can be to an extent, but not completely.

Still, in how they operate, they rapidly test channels and tactics (heuristics) to find a scalable, repeatable process (algorithm) of acquiring customers.

Does that work, though?

Does it ever turn into an algorithm, where you can reduce a marketing tactic down to a set of well-defined instructions, input the generates a particular output, all the time?

No. Channels go dead. Customers and buyers change. Tactics go out of style.

The monkey-wrench of human psychology tends to screw with predictability models, especially when you try to force it on people.

Heuristics + Marketing = Winning

Also, let’s not forget, that any marketer who knows what he, or she, is doing will always espouse “testing” as the best way of figuring out your messaging, Unique Value Proposition, what ad copy resonates, what button color works, and so on.

Because testing is such an integral part of marketing, you’ll be better served developing a set of heuristics for your marketing strategies and tactics. The moment you think you’ve developed an algorithm, your buyers have already moved on.

Before you misunderstand me, though, this is not saying that data and analytics have no place. Of course they do. But they’re indicators of buyer behavior from which we try to understand buyer decision-making and thinking.

At best, you’ll be able to understand it up to 80%, perhaps. Not enough to develop an algorithm off of — but plenty to formulate a few heuristics that will generate those leads and sales that you’re looking for, faster.