Do Campaign Metrics Improve after "Cleaning" them by Removing Keywords that don't Contribute?
Updated: May 20
You wouldn't need to be "cleaning" your campaigns very often if you are regularly bid managing them, that is essentially part of cleaning. There are situations where simple bid management, campaign budgets, and ACoS control doesn't work. This year, for example, I know many sellers who have suffered at the hands of algorithm browse node/sub-category changes, or SKUs have changed for whatever reason.
These changes would mean that the Amazon algorithm has to relearn your product so the keywords (and how it trusts your item will get a sale for them) are affected which means that where your ads show and how much you need to pay for that is also impacted upon.
Other external factors can bring about a need to have greater control over cash, such as increased inventory costs, the need to make more profit/improve cashflow, and so on. These put more stress on cleaning campaigns.
Tricky Keywords that Don't Always Respond Well Following Bid Management
If, for example, you had a set of keywords in a campaign that were enjoying high sales, but also at a high ACoS, bid managing them could be quite tricky. You might reduce the bid by 10% but the ACoS increases over the next fortnight. So, you reduce the bid again by a further 10% and see over the month that the keywords still have a high ACoS and high sales volume. And these keywords could be affecting the ability for others in the campaign to perform well - so there comes a point where you have to look at the structure of the campaign and see how it contributes to the health of all the keywords in it - and then change it accordingly.
When you remove keywords so that only good ones remain in the campaign (as long as there is a group of them - once you get below three keywords/ASINs in a campaign the performance can falter) - then all of the campaign budget is going into those keywords, and none of it is being spent on the others that were poor/borderline. When you see campaigns with good overall metrics, it gives you greater confidence to increase their budgets knowing that they should get more sales but be quite stable.
Reassigning Keywords to New Campaigns
An important part of this process is not to pause keywords and then leave them. You should be retargeting them in other campaigns. So, I often call these "second chance" campaigns to see if the keywords improve in performance once they have left the shadow of the ones in the original campaign. They often do. And the process is then repeated because not all of these will perform well.
Got Trust Issues?
Why we want campaigns with good metrics is not so much to do with algorithm "trust" but because we don't waste so much money investing in keywords with poor metrics. The data gives us an idea of what works for the shopper, and we base our marketing strategies around those.
So, whilst people like me might mention "trust" when your campaign metrics are better, the keywords tend to support each other as a family. If one doesn't do so great this month, well five others might. And when we get more sales, well the algorithm is simply seeing a good connection between the buyer and your product.
You might have amazing campaign metrics but that, in my opinion, isn't giving you a competitive advantage. The algorithm isn't going to look at your ad campaigns and improve your rank because they are in better order than someone else's. But if you are getting more sales, and you can personally keep investing in them at that cost, and put more money into them - well, you're going to get more sales, and the algorithm will trust your product more for the search terms used.
If you want to understand more about Amazon Ads I have several PPC courses available. Check the above menu links for further details. Click the following image to be taken to the course or navigate via the above menu to Amazon PPC courses.