The ten things I've learned after spending over $15m on Facebook (part two)

20th December by Deniz Hassan

Published on


School blackboard

Last month I teased you with the first five musings on the things I'd learned after dropping a fair wedge of cash into Mark Zuckerberg's pocket, so here's the rest.

6 - It's all about the ads

Well, I'm just being facetious having left you with number 5 saying 'It's not all about the ads'... but what I'm saying is creative is rather important, just as it always has been for any fundraiser. I've recently been in a position where I've been able to really see the effect of this. On one of my largest jobs where budget is almost unlimited, we've had a real struggle getting the really high level stuff we need. And despite having the setup almost perfect - with nailed on audiences and hundreds of thousands of conversions to help the AI - as well as a super robust conversion rate optimisation programme supporting the ads, we struggled to get the returns we wanted.

Conversely, for a much smaller test campaign for a charity much newer to the game, we were able to shoot some amazing content. The tracking and landing pages were somewhat hacked together and it was all a bit 'seat of pants'... but the results have astonished me. Low CPAs, high average gifts... dreamy dreamy.

7 - Understand your ad costs

It's one thing people tend to forget about as it's one of the less showboaty metrics. But it's important to understand what drives costs to enable you to reduce them rather than accept them. Often when CPAs go up and ROAs down, you'll look at ad performance but it's very possible the ads are performing as well as ever but simply cost more.

It's important to consider things like seasonality that might ramp up competition in the auction. This is why advertisers that only go big during calendar moments like Christmas are disadvantaging themselves. You're essentially paying a premium to learn and perform when you could've learned during a cheaper time.

Also during busier times, ad quality is even more important so keep rotating and learning and keeping an eye on quality and conversion. We don't exactly know the algorithm but we do know these are massive factors.

8 - Focus on your objective

Is it acquisition you're after? Or just income? Single gifts or regular? This focus will determine your strategy. For example if it's just $$$ you're interested in you'll want to create a campaign which runs alongside email and work out how you'd want to attribute. If acquisition is your game, you'll want to set your exclusions otherwise you'll get a mix of data which isn't really easy to interpret.

9 - Make your campaign size appropriate to your budget

I've lost count of the number of campaigns I've seen with 10 ad sets, each with 10 ads and audience sizes of 100,000,000... but with a $50/day ad spend attached to the campaign. This will get you nowhere. Scale down. Learn slowly and make incremental changes. Start with fewer audiences so you can start to develop your bankers. And have less creative. This way Facebook will be able to learn much more efficiently and so will you.

10 - Customise your attribution

Don't just rely on what Facebook and Google tell you. If you do that, you'll find you've got half the conversions you think you do! By understanding your own data and how your supporters interact with your channels, you can create custom attribution models that show you the real value of your ads rather than simply relying on what the companies who want to take your money say! This one isn't that easy but is so so valuable.

So there we have it! Merry Christmas and I hope something on the list was vaguely useful as a stocking filler. As ever, I love chatting about this stuff so if anyone ever wants to ask anything, don't be shy.

Oh, and it'd be remiss of me not to remind you all that we've got a job going at AAW working in the digital team putting all of this fun stuff into practice!