Perhaps because we spend a lot of time researching the odd parts of Google AdWords, Yahoo!Search marketing (a.k.a Overture), etc, we are often called in to rescue an advert or an account, or at least explain why AdWords or some other paid search program is behaving as it does.
One of the problem types we’ve been asked to look at recently is why certain keywords have a zero or decreased volume of impressions. The causes of the problems? Highly varied, but one of the more recent inquiries was a consequence of the bidding strategy.
Zero impression keywords can be caused by all sorts of factors. Here’s some factors that we’ve looked at:
- another keyword is taking the impressions
- trademark suppression
- performance of the same keyword in a different geoterritory (campaign)
- bid levels
On investigating the recent case, a broad matched keyword with a higher bid was taking most of the impression volume from a low bid keyword. The account management approach had been to pour more money into broader matches because this is where the impression volume was seen to come from. The effect of this strategy was to increase advertising costs, decrease relevance and to remove impressions from keywords that should have been more effective.
The weaker strategy of focusing on a few high impression volume keywords almost certainly stems from the perception of cost of managing large keyword lists. It’s a false economy, at present. By taking impressions away from a large volume of keywords each with more words and low numbers of impressions, the effect was to increase the average paid price, by more than 25%. With the budget spent on excessively high priced clicks the company could have used automation or even employed someone to manage the larger lists.
Why do people come up with strategies like this? I think there’s something counterintuitive to search marketing for people trained in normal marketing practice. Normally marketing people talk about and manage activity using the Pareto Principle or the 80:20 rule. But in search marketing Zipf’s Law (a.k.a the power law a.k.a The Long Tail) is more important.
If you look at the Pareto Principle, you focus effort on the keywords that get 80% of the impressions. You increase their budgets and bids and you ignore the mass of other keywords. This will tend to emphasise keywords with a very few words - because “holiday” is a part of “cheap holiday” and that’s a part of “cheap holiday in (named location)”. However, a web page that answers the needs of a long form search are more likely to convert… so using the Pareto Principle to manage search marketing results in higher prices and lower conversion rates - or a lower ROI. And the symptom is some keywords that used to get impressions no longer do so.
S’funny how you can start off solving one problem, and end up resolving a completely different problem that someone didn’t know that they had.
