Before we get into the rather aggressive title of this blog post, first let’s discuss what Impression Share is and how it can be a really useful metric. After reading this first section we hope you will have a moment that prompts you to think “Am I really measuring impression share correctly? It is mighty darn important after all. I think I’ll read on”.

What is Impression Share?
Technically speaking, Impression Share is a measure of the number of impressions your ads received relative to the total number of searches that your ads were eligible to show for. It is calculated as the number of impressions received, divided by the maximum number of impressions that you were eligible to show for. This is expressed as a percentage.
In other words, impression share is a % of the available impressions that you got your mitts on.

Impression Share Formula

 

There are many levers that may cause you to not receive a 100% Impression Share, however they can all be boiled down to 2 issues through which you could lose impressions:

a) Rank – This is where you lose impressions because your ad has fallen below the first page because of ad rank. This could be due to Quality Score, Max CPC or a host of other reasons.
b) Budget – Simply put, you have spent your daily campaign budget limit and can no longer appear within the search engine results page (SERP) because another click would take you over the maximum spend you set for yourself.

What’s so great about Impression Share?

This could be a blog post in itself, however I’ll try to keep it concise and boil it down to 3 reasons:

1. It’s a reliable indicator of changes in search volume.
Imagine you come into work on a Monday to find that clicks have increased by 30% and you want to know why. Your immediate thought might be, “Wow, my new ad copy is really smashing it,” however (depressingly), you find out that it is achieving a similar CTR to the others – you rule out your new ad. As CTR is unchanged, you will know that your impressions are also up by 30%. Further investigations tell you that average position and CPC are also unchanged. So what is it?

Impression Share is the next metric to help you explain things. If Impression Share is the same, it means that you are not taking any more share of the impressions out there than you were the previous week. Therefore the explanation for a rise in impressions is simply that the search volume on that selection of keywords has increased. It explains how impressions can increase without any change to rank, budget or competition. A lack of change in Impression Share can help you infer a change in the market!

2. It provides a sneaky peek into what your peers are doing.
In a similar situation to that described above, you may notice a change in performance that simply can’t be described with basic metrics like Max CPC and Average Position and that it where Impression Share becomes a key metric in your arsenal.

A rising Impression Share simply indicates that you are receiving a higher proportion of the impressions available to you and your peers. But what it implies is something much more powerful… an increase could mean a competitor has fallen out of the rankings or that some change you have actioned has helped you encroach into your competitor’s space. A decrease, however, could lead you to investigate which competitor is taking impressions from you.

Having said all this, there is a large assumption here that all other things are equal. Of course, they are often not… but wait… there is a cure for that! Not only does AdWords provide you with Impression Share, but they also provide you with 2 other metrics. Remember the 2 factors earlier that influence Impression Share: Rank and Budget? Well, these are also reported in AdWords as a percentage of impressions lost due to Rank or Budget. By comparing two time periods, you can see the root cause of your change in Impression Share – Rank would suggest competitiveness and Budget would suggest pacing changes.

Competitive Metrics

 

Finally, the AdWords Auction Insights report will provide you with further information on your competitors’ Impression Shares – as well as some other handy metrics such as “Top of Page Rate” and “Overlap Rate”, see below:

Auction Insights

 

3. It’s a very useful metric for forecasting.
By understanding how many impressions you are losing to budget, you can infer how many you would be able to generate by increasing your investment. All things being equal, you can weight up (or down) clicks, cost and conversions accordingly for a quick, reliable snapshot of alternative spend levels. Obviously, forecasting and projections get more complicated as you take into account how metrics like Impressions Lost Due to Rank, Max CPC, Average Position and CTR all interact with each other. Either way, Impression Share is a go-to metric for any paid search expert worth their weight.

Why you are measuring Impression Share wrong!
For those of you who analyse and optimise multiple campaigns and accounts, you will likely have a solution that can slice and dice the information in various ways. You may want to look at different combinations of campaigns in order to gauge performance i.e. all brand campaigns, campaigns targeting Australia or broad match campaigns.

A common way to analyse performance is to combine metrics into a summary table in an Excel spreadsheet or custom dashboard, often using an Excel formula such as “sumifs” to pick and choose different campaigns and summarise their metrics. This works fine for simple metrics such as clicks and cost etc. However, with Impression Share, values cannot simply be averaged across the desired campaigns as some have seen more impression and search volume than others. These campaigns would rightfully demand a bigger influence on the final figure. For example, a campaign with a 100% Impression Share but 1 impression should not have an equal weight on the overall impression share of the account as another campaign with a 10% Impression Share but 10,000 impressions. See method 1 below.

Impressions Share Data Table 1

Method 1 2

 

 

A method many businesses choose is one where each campaign’s Impression Share is multiplied by the number of impressions it received. This creates a “Weighted Impression Share” score. To calculate the total Impression Share of a group of campaigns using this methodology, these Weighted Impression Share scores are totalled for the campaigns and then divided by the sum of impressions from said campaigns. This reverses the calculated the Weighted Impression Share total back to a total Impression Share value. The idea behind this method is to arrive at an accurate overall Impression Share figure which is weighted proportionately towards the campaigns which delivered more impressions. See method 2 below.

Impressions Share Data Table 2

Method 2 2

 

 

Some of you may find method 2 very familiar. However, we’re here to say that, unfortunately, you are wrong. This is not an accurate figure for the Impression Share of these campaigns combined.

Let’s calculate the overall impression share using the equation that we defined at the top of this article:

Impression Share Formula

First, we will need a new column in our data table called ‘Available Impressions‘- which is calculated as follows:

Total Available Impressions Formula

Here is our new table:

Impressions Share Data Table 3

And here is our true Impression Share:

imp share 11

Now our 2 previous methods gave us 70% and 45% in a situation where the true value is 42%!

The Reason?

In short, the “weight” you give to an Impression Share should be based on the total number of impressions available in the market rather than the number of impressions you actually received from it. In other words, the number of impressions you receive is related to your performance and is an arbitrary number – it is not related to the size of the market. Therefore, you should not be using it to weight your performance.

For example:

Imagine a market where there are 1,000 impressions available. Brand A gets a 10% Impression Share, Brand B gets a 65% share and Brand C gets 80%. The “weight” of the Impression Share for each of these brands is equal. It would be incorrect for Brand C to weight their Impression Share 8 times more than Brand A – they are of equal worth for this market. In method 2 above, this is what is happening.

In fact if method 2 is being used, then the inaccuracy of the measurement grows with the number of campaigns involved. With each campaign being incorrectly “weighted”, there is another inaccuracy being considered in the final output.

Impression Share x Campaigns

 

The Solution

The fix is relatively simple. Instead of using the additional column to calculate Weighted Impression Share (Impression Share x Impressions), simply use it to calculate the Total Available Impressions using the aforementioned formula:

 

Total Available Impressions Formula

This should provide you with the 2 metrics you need to calculate the true Impression Share – Impressions and the Total Available Impressions. Simply sum up both from your selected campaigns and divide one from the other.

The following two tabs change content below.
mm

James Pendleton

I am a Digital Solutions Specialist at Reprise with over 5 years’ experience in digital marketing and analytics working in both the UK and Australia. I have a love for all things digital, data and technology and an unappreciated dislike of coffee.
mm

Latest posts by James Pendleton (see all)