Facebook likes to keep us on our toes.

No matter how much you prepare; how well researched your targeting, how perfectly crafted your ad creatives and copy - you just can't predict the success, or indeed failure, of your ads.

And if you don't keep track? You might well find that even if one or two of your ads is taking the world by storm, another one or two (or three!) of them aren't doing so swell.

Who cares? My favourite ad is crushing it!!

Cool story bro. That's great news and everything, but those ads that are underdelivering might well undo all of the hard work of your winning ads.

Let's take a look at a basic example

Say you're running three ads, and the first one has already doubled your initial investment:

Wahey!! Party time!

Okay, while we were busy celebrating, let's take a look at how our second and third ads got on:

Ah. Not so great. But hey, we're still up £50 overall right? Where did I put my party hat -

Oh. Right.

Overall spend? £300. Overall profit? Err...none.

In fact, that's an overall loss of £50. (Even more if you consider the cost of putting those ads together).

But I like partying when an ad is doing well! I don't have time to look out for bad ads!

We get it. (Sort of).

We don't want you to be worried about unprofitable ads destroying the success of your best ones. We want you to be confident that any profit you make is safe, so you're free to do other things (like party, if you must - it is nearly Christmas after all).

So how do we identify and kill off those dangerous, unprofitable ads?

Automation, of course! With a little help from our old friend, historical data.

If we can automatically kill off any ads that aren't bringing in our expected results (based on historical data) within the first £1-5 of spend, we still give our ads the chance to get going - but stop them before they have the opportunity to completely drain any profit we make!

It's the magic of the early warning rule in action. Let's show you how it's done:

Step one: Analyse your past results

We took the historical data from all of the Facebook and Instagram campaigns run on an e-commerce account.

We considered the most profitable campaigns, and the non-profitable ones, looking at the average results we achieved, which you can see below:

Real data collected from an e-commerce company's ad campaigns

Step two: Identify benchmark figures

By comparing our profitable campaigns with our non-profitable ones, we were able to produce benchmark figures to use in our automated rules.

These figures could be used to serve both as warnings for poor performance (i.e. those from the non-profitable campaigns) or indications of opportunities to scale (i.e. the figures from our profitable campaigns).

Step three: Create your early warning rules

Now you have that all important information gathered in steps one and two, you're ready to automate.

You'll want to tweak your rules based on your objectives, but here are a couple to get you started.

Example early warning rules for Traffic Objectives

If your goal is to drive traffic to your website, you might set up your early warning rules based on CTR or website views, e.g.

  • Pause ad if spent > £5 and CTR < 2.5
  • Pause ad if spent > £5 and Cost per website view > £0.73

Don't forget, it's not just bad results you need to watch out for, it's no results, too.

  • Pause ad if spent > £5 and No. clicks = 0
  • Pause ad if spent > £5 and No. website views = 0

Example early warning rules for Conversion Objectives

Given that our data came from e-commerce campaigns, it seems wrong not to include example rules for conversions. So here are a couple of example rules you might want to set up based on Cost per purchase or Add to carts:

  • Pause ad if spent > £5 and Cost per purchase > £15.80
  • Pause ad if spent > £5 and Cost per add to cart > £4.90

Again, you'll want to make sure you're covered for ads that bring in no results:

  • Pause ad if spent > £5 and No. purchases = 0
  • Pause ad if spent > £5 and No. add to carts = 0

Profit protected with money to spare?

Let's consider how these automated rules would have impacted our example ads.

All of our rules are set up to capture and kill underperforming ad sets within the first £5 of spend. Had these rules been in place when those example ads were running, this is how things might have looked:

That's more like it. Oh, and that also leaves you with £190 of your original budget leftover, for you to channel into that profitable Ad #1... which you could also do with automated scaling rules.

With scaling rules in place, upping the bid or budget (or even duplicating) that first ad, you might end up with a result more like this:

That's right folks, with early warning and safe scaling automated rules, we could have prevented a £50 loss overall, instead turning an overall profit of £280.50.

And we could have been partying the whole time...

It's of course up to you how strictly you adhere to your historical data, and what limits you want to work within. You might want to kill and capture within the first £2-3 of spend or wait until after £20+, depending on the metrics and budget you're working with. It's entirely up to you.

Summary

Using simple automated rules and historical data, you can set up early (or even super early) rules to kill off your unprofitable ads before they do any damage to your overall profit.

Not only that, with the money you save by killing off your poor performing ads, you can use safe scaling rules to automatically capitalise on your best performing ads!

That means you're free to do whatever you want with your time instead (yes, you're safe to go and celebrate now).

Want to kill off your unprofitable ads, scale your best ones and still have time to party hard? Get your free trial now!

Already a member? Get those rules created!