“We just started our digital ad campaign, and we aren’t seeing the results we want.” If this sentiment resonates with you, you aren’t the only one.
In today’s world of instant gratification and rapid results, it’s far too easy to rush judgment on your digital ad campaign results. But, trust us, the best results take time (and we’ll show you concrete examples that might just change your tune).
The Learning Phase
When a new digital advertising campaign launches, especially on platforms like Google or Meta, ads enter into a learning phase. During this time, the platform tests a variety of placements, creative combinations, and audience groups to gather data on what delivers the best performance.
The platform’s algorithm doesn’t know who is most likely to perform the desired conversion action and what ads are best to get them there, until the learning phase is complete. This data is used to optimize the campaign toward the combination that will provide the strongest results as the campaign continues past the learning phase.
This phase usually lasts a few days to a few weeks, depending on budget and audience size. A larger budget or broader audience provides the opportunity to acquire more data faster, making the learning phase window smaller. This window ends up being 1-2 weeks on average for most campaigns.
The Work You Don't See
While results start coming in, the platform’s algorithm is doing a lot of work behind the scenes. During the first few days, the ads will be shown to a wider section of your targeted audience. There may be larger impressions spikes and more cost fluctuation.
We began a Google search campaign for one of our clients, Insight Memory Care Center, at the beginning of February. The first day the campaign was live, it served 310 impressions, while the next day it served 69. The other days in the first week ranged from 39-203 impressions. We saw fluctuations each day throughout the first month, as the campaign worked through the learning phase.
When we compare this to the first week of March, the ads were then serving on a more consistent level. This campaign has a limited budget, so it took about 4 weeks to start seeing consistent daily results.
The algorithm was able to figure out which searches were driving traffic to the website and began showing our ads to the people most likely to visit the client’s website. Our conversion rate increased and our cost per conversion dropped when we compare March results to February results.
Understand From The Beginning
While your ad is in the learning phase, it is important not to make any significant edits. Doing so will restart the clock of the learning phase, delaying the ads for full optimization.
Making changes too quickly is also based on learning phase data, which is not necessarily accurate in terms of possible future performance. Letting the ad complete the learning phase before making changes allows you to make informed decisions about what changes could maximize results.
If we had immediately started making changes to Insight Memory Care Center’s ads when we saw impressions drop to the double digits, we would have been making a change too early. The campaign wasn’t out of the learning phase, so the daily variation of impressions was normal and didn’t reflect consistent data to use to influence strategy.
We like to begin campaigns by optimizing for clicks leading to the website or landing page, before switching to specific conversions (usually form fills or calls). This strategy allows the platform to gather data on what the current audience looks like before going right to conversions, helping to cut down on wasted impressions later on. This strategy also builds an audience to use for retargeting ads, which reminds the user to engage with your brand after their initial website visit.
A Meta campaign that begins as an ad optimized toward leads like website form fills, will take more money to get quality data from because the cost per lead is higher than the average cost per click of an ad with a website traffic objective. The ad focused on website traffic, which gathers data on audience engagement without the higher cost. Then, when switching to ads with a leads objective, the campaign already has some data on who is most likely to fill out a form.
Learning Phase Best Practices
Here are some best practices for the ads in the learning phase:
- Don’t edit your ad during the initial learning phase and avoid edits after the learning phase unless necessary for optimization. Remember that most edits cause the ad to re-enter the learning phase.
- If you do edit your ads, avoid making too many changes at once. Doing so will make it difficult to tell which edit resulted in any campaign performance changes.
- Keep campaigns separate within each platform to avoid confusing the algorithm. This means Google Ads, as an example, has separate ad groups for branded keywords, service keywords, and competitor keywords. This allows for audience learning on a more specific level and can help ads serve in a more targeted manner.
- Make sure your budget is large enough to be competitive within your industry. Use Google’s keyword planner or Meta’s audience estimate to ensure you are spending enough for the campaign to gather adequate audience data during the learning phase.
Need support with your digital advertising strategy? Get the most out of your ad campaigns with Devaney & Associates expertise in multi-channel monitoring, reporting, and strategy.