Share of search is a metric that’s been accessible for well over twenty years, but its merit has come on leaps and bounds across the past few months, capturing the attention of the whole marketing spectrum. So, whether you’re starting from scratch or well versed in this area, this article will be of value as I answer:

1. What is share of search?
2. How to measure share of search?
3. Why is it significant?
4. Where does it fit into AMEC’s Integrated Evaluation Framework?
5. What are the issues?
6. What’s next?

1. What is share of search?

Put simply, it’s the number of people searching for your brand, divided by the number of people searching for all brands in your category – and it can predict share of market.

To clarify, by “searches”, we’re referring to those that take place using a search engine. And given Google’s huge market share in nearly every country across the globe, it’s safe to say we’re referring to Google searches in nearly all cases.

That’s the basics, let’s go a bit deeper.

2. How to measure share of search?

If you need a free option, then Google Trends can do the job. James Hankins has written a good guide on how to do it, although I’d caveat it that unless you’re comfortable with data and excel, you may find it tricky.

Another free option is Google’s Keyword Planner which, if you’re set-up for Google Ads, will show you how many people are searching for a particular keyword across a particular month in a range (1-10 searches, 10-100, 100-1,000 and so on). That is, unless you’re spending a lot on Google Ads, then it’ll give you the exact numbers. Either way, the method I’m outlining below will work with data from this tool too.

However, the reality is this task is far easier with a proper keyword research tool, and a decent one costs about £500 for a year (well worth it) and that’s what I’m going to run with below.

Taking a fictional example, let’s say you’re Pip & Nut peanut butter and we’ve found that your brand name has been searched for 10,000 times in Google across October 2020. Then, after repeating this exercise for all your peanut butter competitors, we’ve found they’re searched for 90,000 times. Using this data, we’d calculate share of search for October 2020 like so:

Then you can repeat for as far back in the past as your tool allows and map your graphs!

3. Why is it significant?

Because it is a “leading indicator” of market share, according to Les Binet, head of effectiveness at adam&eveDDB. His study revealed that not only is there a relationship between share of search and share of market, it can also be a predictor of it too.

Here’s a great example from his presentation at this year’s EffWorks, where we can see that the share of search for LG correlates closely with its share of market, even predating it by six months:

In other words, if the LG team were monitoring their share of search, they would have had a six-month warning of their dip in share of market. However, it’s worth noting that the lead time between changes in share of search and changes in market share varies by industry. Binet studied three categories and found the lead times were:

• Cars: up to 12 months
• Phones: up to six months
• Energy: three months or less

So keep this in mind if using it yourself.

Another reason this metric is so significant is that, when compared to other forecasting methods, it was found to be more accurate. This example for Volkswagen being exceptionally so:

And considering that plenty of marketing activity (including communications) can have an impact on share of search, it’s an incredibly valuable metric to add to your arsenal.

4. Where does it sit within the Integrated Evaluation Framework?

So you can, and should, be measuring share of search, but where does it fit in the AMEC framework? Well, it’s an out-take: an initial response from your audience to your activity; but it’s also an indicator towards an impact. Therefore, this one metric has the potential to break down data silos and rally the whole organisation towards effective measurement. Seen a rise in share of search as a result of your campaign? Show those data keyholders the rise, explain its significance and tell them you want to investigate the impact that rise had on the business, those walls should come tumbling down…

5. What are the issues?

There are a few, but that’s not unusual given the developmental stage of this concept.

Firstly, searches for a brand can increase for negative reasons. For example, following on from the highly accurate Volkswagen prediction referenced above, the emissions scandal came to light shortly after and – as a result – lots of people searched for Volkswagen to find out about it and the impact it’d have on their car. Crises can spoil this metric if looked at in such simple terms.

Secondly, what do you do if your brand is already in the vernacular? Let’s say you’re operating in the video doorbell space, that means one of your competitors would be “Ring”. Deciphering how many people searching for “Ring” are actually looking for the brand, and not searching for the word (because a surprisingly high number of people search for seemingly random single words) without access to Ring’s primary data would be very difficult indeed.

Lastly, whilst someone searching for your brand does express interest and can express purchase intent too, it doesn’t promise conversion; some brands will be much more effective at converting than others, so whilst the relationship between share of search and share of market is there, it will never be a silver bullet (but then…is anything?)

6. What’s next?

Binet has a follow-up webinar on 18^th November, where this concept will be built out further. What I’d like to see from that is twofold:

1. A study that combines sentiment and intent analysis of all brand keywords studied. It’ll be more time consuming to carry out, but will tackle the first issue head-on and no doubt lead to even better results
2. A methodology for analysing vernacular brand name searches

You can sign-up to that webinar here and, in the meantime, hit me up on LinkedIn or Twitter as I’d love to get your take…