It’s the first rule for business and real estate. It’s an essential ingredient for success when building out-of-home advertising sites. And it’s also an often-critical factor in media planning when targeting high value audiences, ensuring reach with relevant consumers and for local area marketing.
When it comes to digital advertising, there are more tools and technology available than ever before to enable location specific targeting. Most digital platforms will quote a mixture of IP addresses, GPS data, profile information and ‘proprietary signals’.
That all sounds like a powerful cocktail of location targeting for the modern marketer, but who is verifying the accuracy of who we are actually reaching?
The short answer, no one.
Interestingly, when you ask a digital media platform about the accuracy of their targeting products the answers are as varied as the audience numbers they give (which we will get to next), ranging from “less than 20% error” to “100% at 100 metres” to “unspecified, but trust us, our testing is thorough”.
The reality is that when it comes to geo-targeting, the uncertainty is high. It is opaque at best and misleading at worst. There are enough tales of localised campaigns being served out of geo for the issue to be clear and the issue comes into sharp focus when reviewing the targeting tools within platforms like Meta, Snapchat and X.
Let’s take Sydney’s suburb of Castle Hill. The 2021 Census data records the population as 40,874. Using the lowest figure in the ranges given within the social media platforms, X claims to be able to reach 17,800. Meanwhile, Meta claims to reach 43,400 and Snapchat 57,000!
It’s no different when we look at Glen Waverly in Melbourne with a Census population of 42,642. Again, at the lower end of the range, X claims 36,900, Meta 58,000 and Snapchat an enormous 84,000.
It’s understandable that platform audience data doesn’t match with Census data but to claim reach is up to 206% higher than the population is an enormous stretch. That’s before taking into account that these platforms don’t reach everyone in the population.
Of course, part of the answer social platforms will give is that the audience picked up also includes those regularly visiting these areas or ‘other signals’. But what constitutes a ‘visit’ or passing through? Would that count using a social app whilst commuting through on public transport? How long is the data kept and included as part of that targetable universe?
It’s highly unlikely a global platform will give any sort of local insight or answer to these questions, stonewalled as we are by their US masters and endless references to existing URL links with pre-set answers.
If it feels like we have been here before, that’s because we have. Perhaps we haven’t moved on very far from 2017. But in a world that is now even more dominated by tech giants who show us black and white answers, we need to think instead in full colour.
It doesn’t mean we shouldn’t be actively bringing targeting solutions like geo to the table. But like real estate, buyer beware. Know what you are getting when it comes to location targeting. And if you can’t work it out, there is always that outdoor site.
Brett Elliott, Managing Director, Government Practices, UM