The (In)Exact Science of Online Marketing

Modern marketing has always made use of statistics and data to some degree. The Internet has only provided more data in addition to offline sources. As the digital landscape evolved, we were able to build far more accurate buyer personas than any advertiser in the past.

The online world now affords marketers a rich variety of tools with which to track, analyse and report metrics relevant to a strategy or campaign. In fact, these tools aren’t just available to marketers. Anyone with wifi can see roughly how many times people from a certain country search for a certain keyword.

For our own websites, our analytics tell us who viewed what post. We can see where they click and depending on the tool we can literally see where their mouse hovered.

But wait, there’s more: we can see their gender, their age group, what device they’re using…

Early digital marketers must have been simply intoxicated as these new KPIs came into being. So much information! So much data (read: profit fodder).

Yes – I don’t contest that marketing is more analytical and more technical than ever before. Just think back to the 70s: how ridiculous do you think the combination of advertiser/programmer as a profession would sound?

Things have changed. One thing that hasn’t, however, is that…

…digital marketing is not an exact science.

“Excuse ME!?” the data analysts shriek. I’m sure there are several programmatic media buyers who want my blood now. And I’m definitely at risk of a SEO hurling their laptop at me.

I don’t care. Digital marketing is not an exact science and I stand by that.

Yes, marketing is far more accurate than it has ever been. That’s true. But it’s still not perfect. Why? Because…

You’re still dealing with humans.

Humans are imperfect by nature. Many of us don’t fit into neat little boxes (though some of us do, and that’s fine).

As the analytics show, there are definitely strong trends. For example, certain age groups are more likely to have particular views/taste in shows/trivia knowledge/whatever. Some age groups are more likely to fill in correct information about themselves.

Yep, correct information.

Not all of those metrics you get in your analytics is accurate. How many people are visting your site via a random country through a VPN? How many people jot in any old information into their personal details when creating an account? I never, ever do – except in very rare cases.

Depending on the channel, you cannot fully rely on “personal data”. Although I’m sure most people tend to must theirs into the form and be done with it.

“But of course,” you say, “you can still gain a lot of insight from general behaviour and optimize your campaigns for that, over time.”

I would certainly agree with you. However, remember: Humans can be utterly unpredictable.

Digital media buying and the loss of human perspective

While data, analytics, past behaviour etc. are all great metrics to optimize your campaign or sales funnel around, pinches of salt should be taken from time to time. People are not machines. In fact, people can be downright irrational. You can look at your beautifully-crafted sales funnel and think, “They’ve followed these steps, so obviously (Step 5) is the next logical development. Finally, we’ll get that conversion!”

And then, suddenly, boom – they’re gone. You never see them again.

Of course, this is coming from a B2C perspective. Naturally, with B2C you are dealing with primal wants and desires. I especially saw this in the online dating industry. There’s definitely a reason why “casual dating” made so much money.

The same issues can, however, crop up in B2B. It just happens at a slower pace as several (potentially) irrational humans are making the purchasing decision.

I’ve worked on a lot of media buying strategies. The campaign managers themselves got excited about certain trends, and then utterly confused when a lead didn’t convert. They cited possible reason in numbers, figures… And many times, completely forgot that the person was human and might’ve just changed their mind.

Think about it this way: How many times have you clicked on an ad in an app or on a site, when you were really trying to just X out of it? It’s happened to be me a lot (big fingers, small screens!).


I’m not saying that the data and insights we as media buyers have is useless. On the contrary, it allows for the perfection field of optimization. What I am saying is that a healthy dose of human perspective is key, especially if you’re running campaigns all day. It can be easy to lose sight of the things that make you human and act he way you do when you’re always looking at the figures.

Conversion optimization, overall, is not an exact science. It never will be, unless we start marketing to robots.