How to turn around your Google Ad Grants account
What do you do when a Google Ad Grants account goes wrong? When performance is in decline, you're getting no conversions and only a handful of impressions and clicks. That was the problem we had to solve for Freedom from Torture, and our work turning performance around changed the way we think about Google Ad Grants accounts.
Freedom from Torture
The client
Freedom from Torture
Freedom from Torture is an amazing organisation supporting people who have survived torture to heal and feel safe and strong again. Their work includes therapy, rehabilitation, legal support and campaigning so they have a broad range of topics promoted in their Google Ad Grants account.
We started working together in August 2023 on Google Ad Grants, paid media and SEO support. Their Google Ad Grants account had been in decline throughout 2023, which was part of the reason they reached out to us. We rebuilt their Google Ad Grants account in August, the new ads went live in September, and performance tanked…
We were getting 1-4 clicks a week. As you can imagine, we were very worried.
Had we made a mistake in the rebuild?
Had something been disapproved without us being notified?
The actual problem was far more fundamental.
The problem
GA4 transition challenges
We rebuilt the Google Ad Grants account just after the switchover from Universal Analytics to GA4. No historic GA4 conversions had been imported into the Google Ad Grants account prior to us taking over, so when Universal Analytics turned off in July all the account's conversions stopped working. The last conversion recorded was 120 days before we had rebuilt, which signalled to Google that this account was not working well.
When we first started working with Freedom from Torture, we immediately built GA4 conversions and imported them into the account, but they hadn't had enough time to gather that data when we went live a few weeks later. We turned on all the new ads and suddenly Google had no historic click or impression data. Google's AI had no way of knowing what kind of people would click on the ads or convert, so it stopped showing them. We knew we had to take extreme measures to fix this quickly.
The immediate solution
Switching back to old campaigns
We temporarily turned on the previous ad groups that were working well. When we rebuild a Google Ad Grants account we don't delete all the old campaigns just in case something like this happens, so we were able to quickly switch the old ads back on. While we still had the problem of not getting enough conversion data, at least those ads had historic click data and historic impression data so Google had a bit more information to go on. The old ads performed better than our new ads which had no history behind them.
We also looked at what new conversions we could add in GA4 and how can we make sure they really are set up to be conversion focused going forward:
Traffic quality metrics
- Scroll depth
- Session time over average
- Pages per session over average
High priority conversions
- Email subscribe
- Petition signature
- Donation
- Purchase
Lower priority conversions
- Outbound click
- Call
- Event sign up
- Legacy sign up
- Add to cart
We set up the metrics for traffic quality first because we also knew that we could get quite a few of these conversions in a short period of time without them being completely meaningless. Then, we focused on conversions that were really important to the organisation in terms of delivering revenue or supporting their mission. Finally, we set up the conversions that are useful for both signalling engagement to Google and the organisation, but weren't top priority. As soon as the new conversions were set up we imported them into the Google Ad Grants account.
We also broadened our keyword targeting, not worrying too much about overlap between campaigns. Normally in our accounts we use negative keywords to make sure that we're not trying to show ads for two different ad groups on the same searches, but in this instance where we had very few impressions it was more important to show up as much as possible.
At that point, the only thing we could do was wait for 30 days and hope that Google would give the account another chance… and it did!
The results
September to January performance improved significantly
- Clicks up by 7% compared to the previous period
- Click-through-rate up 48% YoY for Sep-Jan
- Conversions up 167% YoY for Sep-Jan
Next steps
Long term growth
Once the new ads had gathered enough data, we were able to turn off the old ads and add negative keywords to avoid crossover between ad groups without having an impact on performance. But we're not stopping there.
We are now looking at building long-term sustainable growth for the account in the following ways:
Applying learnings from SEO work
We have conducted organic keyword analysis to see what Freedom from Torture's website is ranking for organically. We reviewed competitor rankings to see where similar organisations are successful. We also used organic keyword data to identify content gap opportunities. While we're working on improving our organic search visibility for these terms, we can use the Google Ad Grants account to get a boost in the search results and provide learnings for valuable content.
Improvements to website
The SEO work also included an E-E-A-T review and a value of content audit. Freedom from Torture are planning on making changes to their website to ensure that every page leads the user towards a useful action for the organisation.
Focusing on meaningful conversions
Google Ad Grants accounts were once seen as a way to get a lot of impressions or clicks without having to spend a lot of money. Now Google is much more focused on conversions as a marker of quality and if your account is not driving conversions, then your ads will not be shown. This means that we have to change the way we approach setting up ads.
Instead of optimising towards clicks, we can look to set up conversions that signal top-of-funnel engagement: scroll depth, session time over average, and pages per session over average. We need to be careful not to set up events with too high conversion rates as that is against Google Ad Grant policies. It's a balancing act between something that will send meaningful data to Google, without being too infrequent to get learnings.
We also need conversions for middle-of-funnel engagement; people who are interested in the organisation but not ready to support financially or give too much time. For Freedom from Torture, these included: email subscribe, petition signature and outbound click. We should experiment with writing ads with these events as a call to action.
Finally, we still need our most valuable bottom-of-funnel conversions: donation, purchase, event sign up, and legacy sign up. We should optimise some campaigns towards these but recognise that it's not right for every campaign.
We've enjoyed working with Torchbox on our Grants setup and strategy, their expertise has been extremely useful. Performance is on the up and what's more we're not stopping, building through 2024 to make sure we get the best out of the channel.
Final thoughts
It can be distressing when performance is in decline, whether that's due to donations dropping in the cost of living crisis, cookie restrictions making it harder to measure performance, or something like GA4 coming in and changing the way we need to measure conversions. But we believe that this focus on conversions will ultimately be beneficial. The charity sector is under more scrutiny than ever to prove the impact of our marketing; focusing on what works and what drives value is the best way to achieve real meaningful growth.
Discover how to reach the full potential of your Google Ad Grants account with our rapid audit service.
Book an audit