Social media and location based data

After some discussion and prompting, this post is looking at social media and the idea of location-based data. The concept itself is of great interest to many and from what I have seen to date, the services are somewhat inaccurate or very high level (unless of course, you own the social channel)(Side note: a small number of people do identify where they are, but is it enough?).

While social location data is very valuable, no third party has really nailed it yet. That being said many are working on it, but as it stands it is very basic and high level.

Let’s look at how location data is determined currently:

  • Geo-coordinates (not always shared and numbers at one point indicated less than 5% of users enable this)
  • Location text (e.g. Twitter bio) (Not a great identifier of location. In some instances “the moon” or other random suggestions, entirely lack credibility or accuracy)
  • Timezone (this is kind of broad)
  • Top-level domains (e.g., .ca, .mx, .es) (again like the time zone, can be kind of broad)
  • Geo-IP of the website host (this is a shot in the dark, I host my sites in the UK and Canada)
  • Infer (Software guesses based on what you share and/or who your network is)

If we look at Canada as an example, separating the social accounts by province would be challenging and isn’t currently done by anyone. There is no automated process that really does this and this is why we don’t see regional reports, only high-level country based reporting.

While companies like Geofeedia do sell location based services/data, I suspect that a lot of the data has been inferred from conversations, friends list etc or one of the other indicators I referenced above. An educated guess of sorts (Social networks had terminated their contracts with Geofeedia, not sure if that is still the case).

As the technology continues to evolve, we might get to the point where AI can make a guess as to where someone is, but as it stands I don’t think we have that yet and I certainly haven’t seen anything that would suggest anyone is close at this point.

So be wary of anyone selling fine-tuned location based data, it likely has issues or some creative liberties were taken.

What are others saying?

Mashable – “Unfortunately, studies indicate that more than half of mobile location data is inaccurate”

MarketingLand – “Accuracy of location data declining”

MMA – Variables that negatively impact location data quality. Culprits include a “lack of accuracy standards and market education,” “urban density,” “inaccurate interpretations” of location data that have been translated into a latitude/longitude coordinate and poor “data freshness.”

And there are many more articles, studies etc. Search it on Google.

Quick example. I pulled down 1,500 tweets relating to “Chronicleherald” for a 7 day period and only 5% of people identified being in Atlantic Canada. That is self-identified and that is likely high, as the Herald is a bit of a hot topic right now.

If you have some insights (maybe I overlooked something), know of accurate third party software that captures location data or just have some general thoughts/comments, let me know.


Leave a Reply

Your email address will not be published. Required fields are marked *