A client contacted us with a request to make campaigns for his company. He had a site with one product that could be bought in several options, but nothing complicated.
We made campaigns according to our technique:
- Collected keywords and negative words
- Wrote the ads, made campaigns for EX, PH, and BR match types
- Checked the conversion tracking and imported the data into Google Ads. Since the product was worth a certain price, we imported this value in order to optimize the campaign for better ROAS.
At first, everything went well, and the client received a large number of conversions at a favorable price for him. We tested the target CPA bidding model and, after testing the hypothesis that the price is holding, we haven’t worked with a client for a month and a half, as the campaign worked stably and brought constant income.
After a while, the client turned to us, as things went wrong. What we saw after auditing the campaigns was:
- An additional conversion code was added by the client, which calculated the same as ours, and therefore there was a double conversion calculation, which automatically ‘improved the results” 2 times, but only on paper. And the automatic strategy began to work incorrectly, as it received twice as high conversion data.
- The budget was divided, 50% budget was for exact match campaign, 30% was for phrase match campaign and 20% was for broad match campaign. With the previous campaign budget, this was a good separation, broad match campaigns provided ideas for conversion keys, exact and phrase campaigns generated cheap leads due to strong relevance and high-quality score. When the client decided that he wanted to increase revenue, he increased the budget without preserving the proportion – and the list of negative words was not ready for such a budget, so a lot of irrelevant traffic began to appear in the broad match campaign, which brought down the overall profitability of the campaigns.
The results at the time of the audit looked like this:
So at the request of the client, we took the account to work and restore efficiency. Redistribution of budgets, cleaning semantics, and other optimizations. We are still working, because the client wants to give out a budget about 7-10 times larger than it was at the start, but with the same efficiency. At the moment, we have improved its performance, but have not yet reached the best performance after the start.
- It’s worth developing an account gradually, testing hypotheses on suitable budgets and measuring the effectiveness of each test.
- All changes should be discussed with specialists who lead you, because the setup of campaigns may not be ready for shock changes. The goals of the business may change in the process, but then you need to take into account setup and, if necessary, change it for new purposes.