Earlier this year, we all started to hear information about a major update that Apple were planning to roll out to their operating system. The iOS 14 update would make radical changes to privacy by allowing users to opt IN to tracking instead of opt OUT.

Immediately, Facebook was on the offensive, rolling out a major PR campaign aimed at painting Apple as the bad guys and Facebook’s users as the victims. 

What bollocks! But I digress…

What’s the true impact of Apple iOS 14.5 on Facebook ads then?

What we thought would happen…

When these announcements were made earlier in the year, we were told to expect Apple would offer app users a prompt like this. 

(Yes I doctored it ? It’ll actually say “Allow “Instagram” to track your activity across other companies’ apps and websites?”)

What to do with the Applei OS Changes

What’s really happening…

In actual fact, here is what is happening when Apple device users update their operating system.

Everyone who updates to iOS14 will be opted out by default, and instead of being shown the opt-in prompt immediately, it could be a day or a week or more before that prompt shows. 

So that means 100% of users will be opted out of app tracking at the beginning. 

Facebook is estimating that only 40% of app users will opt back in to allow tracking. And to opt back in will not be easy – that function is buried deep in the Apple settings and not easy to find… ?

Further to this, the ‘Conversion API (CAPI)’ that we’ve all been told to install as an apparently more reliable method of tracking than the Pixel, actually has to adhere to the same app tracking transparency (ATT) rules as pixel tracking. 

(Most advertisers don’t understand this and, sadly, Facebook mis-marketed this aspect…)

Is this only going to affect advertising on Facebook? 

No. It impacts tracking on ALL apps and all digital ad platforms will be impacted. Google, YouTube, Pinterest, Snapchat, LinkedIn – any other platform where your customers hang out and you’d like to advertise on will be affected.

Will my ads stop working?

We are already seeing and hearing of advertisers hurriedly turning their ads off in a panic. 

Honestly – when it comes to making any business or marketing decisions, the best thing to do first is to Not Panic! 

Your ads will not work as well as they may have in the past if you do not adjust your strategy to suit the new environment we are all operating in.

But, if your audience is spending time on these platforms, then advertising on there can definitely continue to work. With the right strategy. 

What to do next with your ads

Step 1: Do not rely on the algorithms to know exactly who your buyers are and to place images of your products in front of them expecting them to buy. 

This kind of advertising will not work. 

It is also unlikely that the recent common strategy of using pixel-based retargeting, such as Dynamic Product Ads (DPAs), will be as effective at retargeting people who have been on your product pages or adding products to cart. 

Step 2: Use more video views and engagement advertising at the top of the funnel to attract, engage and ‘warm up’ your ideal customers first. This keeps the engagement on the platform, meaning you CAN retarget them with ads designed to get them to your website to buy.

Step 3: Get more creative with the rest of your ad content so that you are way more appealing to your audience. No more boring images – the algorithm is not there to cover for that anymore. 

Step 4: Be prepared for Facebook’s performance metrics to be incorrect. I’ve already seen people make a knee-jerk decision to pull ads that were suddenly showing a lower Return on Ad Spend (ROAS), when in actual fact their revenue was not down. 

We are now using a combination of UTM tracking codes and cross referencing orders shown in website analytics, alongside Facebook’s own performance data. We’re also building our own custom-developed tracking and reporting system so that we can collect more comprehensive performance metrics than Facebook will be able to give us. 

*For those of you who like to geek out on this stuff – below you will find a break down of how Facebook are now measuring results.

Step 5: Get stuck into Klaviyo!!! I cannot emphasise this enough. Klaviyo is the email marketing platform purpose-built for eCommerce. No other email platform offers the functionality for eCommerce like this. 

But for the purposes of assisting your Facebook ad results – make sure you have Klaviyo lists and segments set up and synced across to your Facebook custom audiences. 

This is one way you can work around the lack of pixel-based retargeting audiences so that you CAN retarget your warm audiences with ads.

Step 6: Always spread the risk. This is not a new piece of advice from me – I always advocate that the best marketing results come from sharing your marketing messages through a number of different channels. 

At a bare minimum, this means advertising on other platforms (Google Shopping ads are an excellent pair with Facebook ads), and search engine optimisation, along with email marketing.

Need some help?

Keen to get moving on an Ad Strategy and would rather get some help instead of spend ages trying to nut it out on your own?

After working with hundreds of eCommerce brands to help them with their advertising and marketing, we have access to so much more data than any one person has at their fingertips. 

So we know what’s working and what’s not – and we’re able to navigate these major platform changes much more quickly as a result. 

My team has the experience and know-how to ensure your online store achieves its potential.

Click the button below to schedule a time to speak with us about your goals and requirements.

*The geeky stuff…

Reporting on the 8 conversion events will be split between Modelled and Partial.

Modelled vs Partial

1. Partial = 7 day attribution. Campaigns will be optimised to this. No alternative.

Include Android users and people who have opted in within Apple iOS (aka no one).

2. Modelled = 1 day attribution. Only used for reporting.

This is the same as above, Partial, plus statistically modelled portion of people who opted out. Historical data used.

Why are 4 events used for purchase value event? 

It’s going to be a Modelled (as above) ROAS number, and the 4 events are the 4 tiers of purchase value categories to your store. Example for me is $20 hat, $100 tie $500 shoes, $1200 suit. Apparently super important and without it your ROAS will likely plummet.

Turned off = Partial data only.

Moving forward the campaigns need to be simpler, and you can forget website visits/AddToCart retargeting as that won’t be available any more. Keep it broad.

The “Learning Limited” is more important. Previously you could ignore it, but now you’ll see huge peaks and troughs as raw data is so limited now, you need every bit you can get so take it seriously.

Important highlight: Opted out user’s data will only now be able to viewed on the 1 day click attribution! It is important to work out the impact of this within your account.