Here is how Facebook Throws Advertisers Money in Gutter
As a Facebook Marketer myself, most people would advice me not to talk about the negative points in Facebook. But by sharing this evidence, I believe other advertisers also would become cautious about the Facebook insights data and Facebook would take action fixing these bugs.
About couple of years back, I wrote on how Facebook charge advertisers more than they should and shows more views, video views and web clicks than your ads actually received. Recently, Facebook admitted it and said it was a glitch and will be fixed soon.
I like that Facebook’s apology and I believe Facebook will fix all other pending glitches too.
There is a talk for years that Facebook showing advertisements to bots too and charge for all those views and impressions. As advertisers, we pay Facebook only to show our ads to the real humans who will be interested in our campaigns and take some action for us to justify our ads spending.
Evidence: Below is a screenshot of an account that Facebook suspected to be a bot and so blocked it. Still it allows the account to login to manage the pages it is administering on, which is a good thing actually if it’s a real account mistaken for bot. But what’s not so good is that Facebook, despite banning an account considering it a Bot, did not stop showing advertisements to that account. I’m sure with all the technology power, Facebook could have easily done it. But why haven’t they done it yet? Think yourself.
Now there are a lot of disadvantages that Marketers will face if their ads are shown to Bot accounts.
A/B Variation Test Becomes Effect-less
One of the reasons we run various advertisements to the same campaign is to find which ads perform better and which doesn’t. With results after some 1000s of reach, we’ll deliver only the ad that gets us a better result and pause others that doesn’t.
Now the problem is, as Facebook is not going to reveal how many bot accounts are there and how many bot accounts our Ads are delivered to, we could not properly compare performance of ad copies.
For ex: If we run 2 ad copies, assume the 1st ad copy shows 20 engagement against 1000 reach and 2nd ad copy shows 15 engagement against 1000 reach. Going by the numbers, we can be sure that 1st ad copy performs better. But what if most of the reaches that 2nd ad copy received are from bot accounts?
Impressions Could Go Wrong
Bot accounts negatively affects the reach and impressions you see. If you are paying for impressions, how do you know that your ads are delivered to the people and not to the bots?
If Facebook argues- how can a bot log on and see ads?(which would count as an impression),
Well in that case,
- A bot account that Facebook thinks is good enough to maintain pages is not good enough to log in automatically?
Well, there is no solution till Facebook fix this glitch.
I always suggest that it’s best to select different objectives like cost per impression or cost per engagement based on an advertiser’s goal for a campaign, but till this bot-issue is fixed, I’d recommend you to go with cost per engagement wherever possible.