Why You Shouldn’t Target Your Ads
Every marketing expert will tell you that targeting is everything. But there’s something important, which they aren’t telling you.
There’s a new religion in the advertising industry. In the past, advertisers started their campaigns by targeting ads according to their customer research. They had to hope that their research was accurate.
Nowadays, there’s a far more objective philosophy for targeting advertisements. Experienced advertisers initiate campaigns without enforcing any targeting. Sounds crazy right?
Flying Blind
To inexperienced marketers, the idea of paying just for data is an anathema. After all, we’re in the business of minimizing advertising costs. Trying to lose money seems like insanity.
Let’s imagine a scenario that we have no customer research. We might have some preexisting views, on who our customers are likely to be. Perhaps we have an existing customer base. However, it’s important to realize that these are actually internal biases, which might be fundamentally incorrect.
A beginner to advertising might adopt the “suck it and see” approach, by trying out different targeting and seeing which method was the most cost effective.
However, there are some obvious inefficiencies to this approach. You can’t test every possible targeting combination — so you’re always going to miss out on the best targeting.
The Control Experiment
Let’s imagine we start our ad campaign with no targeting whatsoever. We target every demographic, every audience and every device equally.
Scientists call this the “control experiment”. Essentially, it’s an experiment which proves what doesn’t work.
The longer we run this ill-fated experiment for, the greater our confidence level regarding what doesn’t work. We will know with a high degree of certainty, precisely which customers aren’t engaged in whatever we’re aiming to promote.
Initially, the results of your experiment will swing violently. When starting out you’ll have smaller data sample, volatility will rule supreme.
While tempting, you should resist the urge to end your experiment prematurely. Keep letting the machines do their work and try to distract yourself from the expense. I suggest you dust off that old Gameboy, or watch a movie on Netflix.
Eventually, you’ll see all your data metrics settle down. This is the point when you know your experiment has reached an accurate conclusion. There’s no longer any point in paying for more data, because the conclusions are unlikely to change. The pain to your wallet is finally over!
Monetizing Data
Having concluded your control experiment, you now know exactly which demographics and audiences are interested in your product or service. You may even be able to cross-target your audience — using segmentation, such as household income, education, etc.
Perhaps it seems like you’ve merely spent a fortune on poorly targeted ads. However, you will have learned something significant about your audience, which you didn’t know before. You can now apply highly targeting to your campaign.
Successful ad campaigns can be expected to run for months, or even years. Over time, the money you save through accurate targeting will recoup your losses, and more.
This is the sort of strategic thinking which differentiates advertising professionals. They value data above all else, because it gives them a cost advantage over their competitors.
Scaling Up
But remember, even though you’re targeting your audience correctly, there’s still plenty of additional targeting work left to do.
Scaling up your ad budget, gives you the opportunity to conduct a wider experiment, with higher-resolution detail. For example, use your long-term ad budget, to gather further testing on:
- Good and bad placements.
- Best time of day, or day of the week.
- Blacklist keywords.
- Bid adjustment rules.
Also remember, ad networks are applying machine learning behind the scenes as well. You can input conversion data to conduct split testing on an array of ad copy and landing pages. The possibilities for improvement are virtually endless!
Ultimately, there’s no reason why you can’t keep refining your targeting. But remember to always be objective, be scientific.
