Unless you have been living under a rock for the past decade, you will know that the number one thing most marketers talk about is advertising on social media. And a good chunk of the time, that means Facebook advertising (which also now covers Instagram). The proposition is tantalising too, fully targeted advertising, right down to how often someone blows their nose. Your adverts will hit your target audience squarely in the face.
So it was some years back that we had decided to start using the Facebook system to advertise our own products [we ran our own online based product business selling globally for close to 20 years]. The cost was low and the promised reach was, well, high. Over a period of months we ran modest campaigns and according the the numbers being fed back, the adverts were making good reach. Then we decided to tweak the campaign and run a ‘click for like’ advert. If you are unfamiliar, click for like is a campaign that encourages people to like your business page, you pay per click.
And the likes started coming in.
Being the stickler for ROI that I am, I decided to forgo relying on the numbers being fed to me by the Facebook ad manager and actually backtrace who was liking us via the advert; our account was not so big that it was an issue and after all, we were paying for these likes, so I wanted to see who was liking as it would be easy to tell if they were target market.
The picture that came back was, at very best, disturbing.
To be clear here, I was physically clicking through to the accounts liking our page via the advert. That meant I was looking at the people’s actual Facebook pages. Sounds a bit mad but…
What I found was the majority of the ‘accounts’, and by majority I mean in the vicinity of 50–60%, were what could be best described as either bogus or inactive accounts. And they all followed the same pattern. A ‘person’ (a term I use loosely) somewhere in Africa or the Middle East, mostly with one or two posts and nothing else. There were also accounts (supposedly) from the US that followed the same pattern.
At first I thought it was something I had, or more to the point, not done. So I ran another campaign, making sure to diligently check all the regional, demographic and interest settings carefully to make sure we were 100% on mark. The results once again reflected the same patterns. Now it’s easy to sit in isolation, see things like this and make up all sorts of theories but luckily for me a friend of mine in New Zealand was doing the same thing for their magazine, so I asked him to do the same level of checking I had done. Lo and behold, what came back were almost exactly the same results I was seeing.
Keeping in mind this was all before the dirt started rolling out about Facebook, but between the two of us, in different countries, marketing different products, the picture forming was that the Facebook ad system was delivering false positives. Worse than that, it was systematically delivering bogus results that it was charging us for and if I had never been so pedantic as to actually drill down to the individual page level, instead relying on the stats delivered by the management system, I’d never have known.
Suffice to say, we promptly stopped all advertising.
Fast forward to now, several years on. With some stock that really should not be sitting in boxes, I decided to try Facebook advertising once again, as it now also runs campaigns across both the Facebook and Instagram platforms. With everything that has transpired about Facebook over the past year, one would think what I had been seeing would have been sorted out.
With changes in the ad creation system, I decided to follow Facebook’s recommendations for a few of the settings. I dialled in the age, gender and interests, then used Facebook’s own recommended location setting; I also set a very modest budget, as there is not that much stock to clear. And the results came flooding in. In one and half days the advert hit a reach of 1.5k people with a engagement of 50%. In anyone’s books that’d be great, right? But not a single sale. In fact, I saw a better sales result in the single Instagram post I did.
So once again, it was time to forget the numbers served by Facebook and drill down and actually look at who was engaging. And guess what?
The below screen grab reflects what I had originally stumbled upon a few years back. The majority, and this time I am talking around 80% or more, of the engagements are coming from an audience so far removed from targets I had set it’s not funny. The accounts seem to be a mix of potentially bogus and what I call ‘off target’ accounts, but the commonality is that the bulk originate in the Middle East or Africa. Further, when clicking through to the account pages, it becomes instantly clear that these people (if indeed they are people) would have zero interest in the target ‘interests’ I had set for the campaign. The same advert on Instagram revealed almost identical results.
So what’s going on here? Considering I have found near identical results for campaigns separated by several years, one simply can not say this is a strange coincidence. It very much appears Facebook’s delivering false positives in order to look like it’s achieving results. What’s more, when I reverted to my own geographic settings (AU, CA, NZ, US), as opposed to the recommended, engagement plummeted. What does this tell us?
Asking those I know who have run Facebook ads, most admit to only relying on the statistics from the ad manager to measure success; and those numbers make the ad system look like a very solid proposition. In today’s world of low ROI for any form of paid advertising, the ability to look at metics that tell you that at least your stuff is being seen may be enough. But the equation changes when you start to think that maybe 50% or more of those ‘looks’ may not actually be genuine. Throw enough money at your campaign, as the ad manager likes to keep suggesting, and sure, your ‘genuine’ engagement will increase , but I suspect so will the bogus; what that ratio will be I have no idea.
It’s easy to think that with all that’s been revealed about the the machinations over at Facebook, that this view of their advertising system is a skewed, cynical one. The fact is we very much want to move this product, so opted to give the system another go in the hope it would deliver. What we found though were the same highly questionable results we had seen several years ago; results that only become apparent if you take the time, and effort, needed to drill right down to the individual level. Based on this, all I can conclude is that like everything else about Facebook, there are untoward things going on that are less than honest…
My advice to anyone thinking about using Facebook advertising? Think long and hard about it. If you do decide to give it a try, do yourself, and your budget, a favour and drill right down to the individual interactions. If what we found is across the board, you won’t be impressed.