I don’t remember what the first Facebook post I ever put money on was. That’s what my charity called it when we had no idea about the sheer scale and power of paid social – “just put £75 on this post and see how it goes.”

So I shrugged my shoulders, hit boost, took a random guess at a few other charities someone who could be interested in our petition might like, and spent the equivalent of two weeks of groceries promoting a picture on Facebook.

A few years on, our agency now works with budgets anywhere between four and six figures. Deciding what to spend money on, when and how is central to my day to day work.

Through a lot of trial and error, messing things up, scratching my head, and discovering all sorts of new quirks in the way Facebook allocates spend and prioritises advertisers, I’ve learned more than a few key things that I wish I’d known when I first started.

I wish I’d used Facebook insights.

Here’s a secret: the assumptions you’ve made about what your audience likes are in all likelihood not that accurate. Time and time again, we test ads targeted at pen portraits or supporter profiles provided by our charity clients against our audience research, and almost every time, insight-driven audiences perform better.

Here’s another one: no matter how good your ad copy and creative are, if the people you’re delivering to don’t care, it doesn’t matter.

You may not be surprised to find out that people who like your animal charity also like the RSPCA or Dogs Trust. But did you think that a large proportion of your fans would also like Dunelm or Center Parcs?

I’m big on data-driven insights. So if you’re not using Facebook Audience Insights, it’s a great start. When you input your organisation into the sidebar, the tool will tell you what percentage of your audiences have specific interests outside of your charity. Use that information to help you figure out who your audience is and what else you can target to find more people like them.

If you want to take this to the next level, when you’ve got a nice little collection of people who’ve taken action on your ad, create a lookalike audience and deliver ads to them as well.

I wish I optimised campaigns for the actions I wanted people to take.

Facebook has a few different options for delivery optimisation depending on what the aims of your campaign are. Basically, this means that Facebook only delivers the ad to people it thinks are the most likely to complete the action that you’re asking them to undertake – whether it’s clicking on a link or watching a video.

It can be a little overwhelming, when you’re first starting, to be faced with all of these options. But it’s important to optimise your campaign for what you’re actually trying to achieve because it makes a huge difference to who Facebook decides to deliver to – and once it’s live, you can’t change it.

You’re probably familiar with these if you’ve done some advertising before – this list pops up when you’re setting up at the campaign level:

But did you know there’s a secondary list at the ad set level?

Have a play around with these to see what suits your campaign best: for example, if your page is really image heavy, or you want people to load up a page that has a video on it, you might be better off optimising for landing page views rather than just link clicks.

I wish I tracked… anything, really.

If you’re running Facebook ads without the Facebook pixel installed on your website, go do that now. Here’s why:

1. It helps you track how far people are getting through your website journey
2. It helps you keep your advertising low by helping you to retarget people who have already been on your page
3. It helps you optimise for conversions – when you do this, Facebook will try and find people who are also likely to convert, based on the people who already have.

If you haven’t got a Facebook pixel set up, it’s fairly straightforward to do it – check out Facebook’s guide here.

I wish I’d excluded people who’d already converted.

You spend advertising money on every person you reach. So once they’ve signed your petition or made a donation, remove them from your audience – otherwise, they’re just going to keep seeing the same ad, and you’re going to keep spending money asking them to do something that they’ve already done.

If you have a Facebook pixel installed, you can do this quite easily: create an audience of everyone who’s reached your thank you page URL, and then exclude that audience from all your ad sets.

Doing this isn’t just about putting your advertising pounds where they matter most – more importantly, it’s also about maintaining a healthy, happy relationship with your supporters.

I wish I’d split tested more.

We split test with the goal of long-term learning, first and foremost. But our second, and often more immediate goal, is to figure what performs well for this specific campaign, so we can keep delivering it.

Start with the biggest variables, the ones that result in the biggest changes, and work your way down in successive versions of the campaign. For example, you could launch your campaign with three different types of creative: a video, a photo, and an illustration.

If illustrations are performing the best for you, launch a new round of ad sets with three illustrations in different colours. Turn off the ads that aren’t working, and keep launching new versions until you work out which combinations work best for your audiences.

You need to temper the number of variations with the budget that you have. Larger budgets allow for more variations and more iterations – but just because you have a small budget doesn’t mean you can’t test at least 2 or 3 variations.
Are there any other mistakes that you made when you started using paid social advertising? Let us know in the comments!