We specialise in paid search management. The largest budget account we managed was in the finance sector, up to that point previously handled by many different (large and small) PPC agencies over years. Both the new industry for us and account history meant the project was a little daunting. The account was spending £150k per month.
But given the client was enthusiastic about our philosophy, we set about completely reworking every aspect from scratch. The complexity level is the number of different keywords associated and not really by the volume. Within several weeks we had reduced the cost per acquisition by 40%. Our predecessors had messed things up.
Improvement Factors
It will entirely depend on the current account structure. If there is a lot of keyword history, then that's all we need to work on
Since 2019 when we started that account
- Google changed their targeting rules and made management a lot harder, we had to rework it to counter, and we did.
- If google has been influential in account management, then performance may well have declined over the past two years. We had to work hard to prevent that.
- Automation is seductive but serves to hide the source of success - if you don't know the keyword, you are not measuring anything.
Our philosophy is measuring keyword 'elements' accurately and separately. Averages hide performance, and every google tool shows averages.
- Switching off the lower converting phrases
- Bidding high on high converting for maximum market share
The real skill is separating and grouping terms - many think it is easy or unnecessary, but we are still learning 12 years later. Keyword elements combine exponentially, bringing ambiguity.
The reward is to find the relatively small number of terms responsible for the vast amount of profitable enquiries. It leaves a lot of wasted clicks.
Hundreds of thousands of keywords need filtering and negating from each other. If keywords show up in multiple places, then you are just not measuring them. Returns on a keyword can vary hugely. We have seen a 10x difference even between singular and plural variations.
The problem started when Google reserved the right to show any keyword on any ad. If you switch off a targeted keyword, they'll show it elsewhere. Unless you actively negate those words in every other campaign and as a group.
Once this is all done, most of the work is to match the site with the keyword for conversion rate improvement. We unfashionably believe in first-click attribution. It is considered less accurate, but we think it's most actionable and reveals the value.
Chargemaster PLC
This project had the most complicated keyword set for a national campaign. The electric car chargers had a complex set of tech specs with brands and location elements and the mixture of intent between people who wanted to buy one, find a space to recharge their car or research the topic.
We quickly understand the topic and intent - it's part of what makes us good at all this. We measured things for so long across many different industries that patterns emerged. Comparisons enable quicker analysis.
Within 14 months, we had doubled the results and then doubled them again. Our work was before the company sold - something a few of our clients have done.
Summary
The large account may be the equivalent of many smaller projects, but equally, it may not. The keyword set doesn't always equate to volume. Focusing on high volume keywords is generally a mistake. We focus on the smaller volume, closer to buying keyword cohorts, and the larger volumes eventually look after themselves.
