Negative Keyword Expansion

Developing your negative keyword lists can have a huge impact on your account performance, appearing against irrelevant searches not only increases costs but reduces CTR and Conversion Rates. Developing lengthy yet considered negative keyword lists will help drive improvements in your account performance and save you money!

split testing

There are many ways to develop your negative keyword lists from downloading “common negatives” from third party websites, manually scanning through Search Query reports or with a more data orientated way! In this blog we are going to show you how we use Microsoft Excel to interrogate our search query reports and generate extensive, relevant, Data Driven negative keyword lists.

What is the goal of a negative keyword:

“to stop ads from showing against searches that don’t produce results?”

We need to find the words that simply don’t produce results!

Strategic PPC Planning

A search query report can contain thousands upon thousands of search queries and these can’t confidently be interrogated manually, we need a data driven process to do this.

For example: a single search containing the word “free” might only cost £0.10 but a thousand searches containing the word “free” could cost £100.

We need a solution that can isolate the component words of each search and then combine performance together to analyse and determine the right course of action.

“does the word free drive conversions or not?”

We need to identify all the times the word free has been used in our searches, collate all the impressions, clicks, costs and conversion data together and determine:

if the word “free” had been a negative keyword what would the impact have been? Would it have saved us money? (well yes of course), would we have lost any conversions? (to be seen) and should it then be added as a negative keyword?

How To Stand Out

The Process:

The first step is to download a search query report from AdWords, use a decent time frame such as 30 days so that the data is developed and removes the impact from any daily influences.

This will produce a list of search terms used to match and show your ads, for example:

Search Query Example

We now need to isolate every word used in these search terms; this will produce a list such as this:

Component Word Example

With the component word list compiled we can now use a formula to gather all the data from the Search Query report:

=SUMIF([Search Query Column],”* “&[Component Word]&” *”,[Data Set])

By expanding this to include the impressions, clicks, costs and conversions we can produce the data sets for all the words:

Component Word Results Example

We can now clearly see that the word free was involved in 290,000 impressions, involved in 25,000 clicks and was involved in £1,786 of costs, but it did not produce any conversions!

The word free should definitely be added as a negative keyword!

We can also see from the data that the word for could be added as a negative keyword too as this produced no conversions either, and we should also consider adding the word daily as a negative; daily did generate conversions but these come at a very high price and high cost per conversion.

By adding these 3 keywords to the campaign as negatives we could have potentially saved 62% of our budget and only lost 6% of the conversion volume. This is surely a win for the client.

Applying this approach across other accounts, campaigns and ad groups we can make considered savings across the board.

Success at PPC

If you would like to start saving budget please get in touch today and one of our experienced team would be happy to discuss your requirements at length.

Written by Rick Tobin Managing Director at Circus PPC Agency

Rick Tobin
Managing Director

Over 20 years dedicated PPC experience working with some of the world's biggest brands.

Share on