Say Hello 👋

If you’ve worked in social, you’ve likely come up against at least one Facebook or Google analytics attribution showdown. Google Analytics can make your Hercules of a social campaign look more like 80’s Tony the Tiger. But look a little closer, and you can see they are telling you subtly (but vitally) different information, both presenting it in a way that shows themselves in the best possible light. Martin Burns, Head of Media and Analytics, weighs in on the biggest power struggle of our time: Facebook vs Google. Who is telling you the truth?

Advertising platforms are doing a better job of opening up opportunities on their channels, and now money from your social budget could go in a number of different directions; from TikTok to Spotify to Snapchat to Reddit, just to name a few. For the best part of a decade Facebook and Google were really the only serious gig in town and what you didn’t spend on one, you’d likely spend on the other.

Google is questioned for their motive less often than Facebook, but they make their money the same way – and are often competing for the exact same spend. It’s understandable; if you are a manager for a company with multiple sources of traffic, with multiple campaigns and a level of complexity, Google Analytics offers you a tantalising prospect of being able to attribute everything all at once… but consider the source! Paid search results appearing exactly above organic search results for the same term is just one example of not-quite-pure motives or results from the search giants. And that same information telling you that paid and organic search are doing the real heavy lifting for your conversions, is also the one to benefit from your subsequent choice to move your budget to Google search or display.

Let’s examine 2 ways to look at the same transaction:

Someone sees an ad on Instagram for shoes on their phone, they click on it and look – but don’t buy them. Next, they mull it over and decide they are going to take the plunge; later on their tablet they search for the shoes and then buy from that session. Facebook knows that it served an ad impression, and that the person who saw that then bought from the site. Facebook claims full credit. Google sees a bounced social session and then a search session that converted. Google claims full credit.

Giving credit to the last click for the above is like giving credit for your holiday to the pilot who flew you there – they were crucial for your journey but not the reason you decided to go. Take away the pilot and you’d find another way to get there. Take away your decision to go on holiday – you’d still be ordering a Friday night Domino’s at home.

Both are giving you accurate information. They look contradictory, but they both can be reconciled and cohabit. Facebook is saying this person bought an item and we showed them the ad that got them there, Google is saying the session that converted into a purchase came from their search. Both are true.

In our early days working with Deliveroo we would often see organic search spiking, corresponding with the timing of a campaign launch. At the time they weren’t the company they are now. People weren’t thinking, “What if a kangaroo could bring me stuff – would that be called Kangery? Let’s Google that… Wait, what about Deliveroo? Let’s Google that also.” Well maybe some creative souls would, but probably not enough to spike search traffic volumes. If your main or only source of information is Google Analytics, that spike might be what you’d see.

That being said, Facebook aren’t saints – there is a lot of suspicion thrown their way which is reasonably understandable. It’s common for a campaign to show success through conversion – which happens almost all of the time, but this is a case of imperfect segmentation of audiences for campaigns with Facebook taking credit for said success. If Facebook has opportunity and information, it seems to effectively run a small retargeting campaign – one that presents as prospecting results. That being said, effective exclusions can prevent this. At Puzzle we’ve worked with a number of different sized companies, and in instances where we’ve worked with smaller organisations, using Facebook campaigns to increase the audience – we’ve only ever seen Facebook understate the results from their ads in this specific situation.

If Facebook were found to be knowingly misleading advertisers (and there are a lot of people watching them very closely) – they would lose their business. The deterrents ultimately come down to two points: they make a lot of money and the stakes are too high. There is a cottage industry that offers you the chance to make sense of your attribution through the use of third party models, information and cookies (for a fee, or more often a percentage) but what if the issue isn’t inaccuracies, but understanding the information that you already have?

With all this being said, when you are looking to attribute Facebook ads, set up your exclusions and understand what the metrics are telling you. You can verify aspects using tools like Google Analytics, but don’t throw the baby out with the bath water. Use Facebook to attribute Facebook, because like it or not, they are the only player in a position to know.  

If you want to learn more about running paid social campaigns contact hello@puzzlelondon.com