“Impressions” is an Undervalued SEO KPI

This article originally appeared over at https://www.distilled.net/resources/impressions-is-an-undervalued-seo-kpi/. That URL 404s at the time of writing, hence the new location here.

This post is about a hill I’m willing to die on, but might not need to – I actually have no idea whether the title of this post is a controversial statement, or not. I’m keen to find out! However, from what I see in talks, posts, pitches, and business practices, our industry has definitely not taken to heart the value of an impression.

I’m going to lay out five reasons why I think impressions are not just a valid indicator of SEO success, but actually an unusually good one.

Before we go any further, it’s probably worth clarifying what I mean by an impression – I’m talking about the number of times someone has seen a site in search results. The most common place this is measured is Google Search Console, who write about their metrics in more detail here.

Reason 1: Clickless results are still worth having

If organic search was ever just a performance channel, to be measured by directly attributed conversions, it isn’t any more. Organic search is also, increasingly, a brand channel – something we all implicitly recognise when we invest in top of funnel content for our sites and clients. 

Furthermore, even if you were working on top of funnel content entirely for the remarketing list, you probably also work for a business that invests in branded advertising in other channels, whether it be billboards, sponsorships, display ads, radio, TV… the list goes on. All of these are channels where the primary objective is to get the brand in front of a potential customer, often at great cost.

What this means for organic search is that as much as it’s annoying that Google is interpreting more and more search terms as informational, or delivering more and more clickless searches, we probably need to stop complaining and start playing the game. This is value that we can provide to our clients and businesses, that they’re probably already paying a great deal of money for in other channels.

Indeed, if you just aim for search results that deliver a click, or a converting click, you’re leaving open space to your competitors.

Most SEO metrics, however, don’t capture this value. The one that comes closest is rankings, which brings me to my next reason…

Reason 2: Why not rankings?

There definitely is value to rank tracking – it’s easily communicated and understood, it’s precise and controlled and easy to break down, it provides a whole load of analytic depth you don’t get in other places. However, some of those strengths are also weaknesses – rank tracking, even if it does include the numerous location-based, personalised, device-based, or result-type, or synonym differences, only tracks the keywords you ask it to. It’s limited, therefore, by your budget, and your imagination.

Impressions, as reported by Google Search Console, include any keyword you might happen to rank for – even if you never thought of it. Even “search visibility” trackers like SEMRush, Ahrefs, Sistrix, and Searchmetrics have a limited corpus of keywords, often biased heavily towards high volume terms.

Impressions from keywords you care less about might be less valuable than any you garner for your target keywords, but they still have some value.

Reason 3: PRs have been fighting this battle for years, with worse data

Public relations (PR) agencies are paid a great deal of money to get their clients mentioned, preferably in a positive light. The general consensus, certainly among clients I’ve worked with, seems to be that this is valuable and useful, and it’s obviously an industry that’s been going for scores (hundreds?) of years, and doesn’t seem to be going anywhere.

To demonstrate and measure the value of their work, PRs have historically reported on metrics like the circulation of publications they have achieved placements in, number of column inches placed, or “advertising equivalent value” (AVE). These metrics are, compared to what we have online, terrible. There’s no real way of knowing how many people saw a print article, and yet, businesses found the notion so important, they were willing to tolerate that, and invest billions.

As SEOs, and digital marketers in general, we sit on far, far better data. Impressions are a pertinent example – we know exactly how many people saw a page mentioning our content, and can describe the light in which it was seen. But, we too often don’t bother communicating that.

Reason 4: Impressions are a ranking factor (well, not quite, but…)

This is not literally true – I do not think that Google looks at the number of impressions that a site gets and then factors it into that site’s rankings. That would be a bizarre and circular system. It did get your attention though, didn’t it? And if the principal benefit of impressions as separate from clicks is brand awareness, there’s lots of reasons to suspect that that would indirectly impact rankings.

Brands which are well known and trusted will find it easier to garner links and clicks, which will in turn make it easier to rank.

(I’ve written a bit more about this ranking factor nuance in this recent post.)

Reason 5: Straight from the horse’s mouth, absolutely free of charge

Lots of the metrics we have access to in SEO are critically flawed. Because Google is so cagey, we end up paying rank tracking providers a small fortune to operate an enormous batch of IP addresses with which to scrape search results. Even then, getting a real picture of what results are like in different parts of a country, and weighting your metrics by the different volumes in those parts, is an absolute nightmare. Or you can use an on-site analytics platform, which seems to get shakier with each passing day.

Impression data, on the other hand, is free from Google Search Console. There’s an API, which has a free Data Studio integration. The data is minimally sampled, robust to tracking protection, and not biased towards any location or device type. Sure, there are occasional updates to Google Search Console, which can be a bit of a spanner in the works of your year-on-year comparisons, but that’s true of most other search data platforms, too.

Caveat

I’m hoping to read on Twitter and in the comments below the various ways in which I’m stupendously wrong. However, one flaw I’d particularly like to draw your attention to is that sometimes fewer impressions is better. This is particularly true if it comes as a result of better focusing your site’s targeting, such that it actually gets better rankings and clicks for the keywords it’s relevant for.

To give you an extreme example – I had a client several years ago that ranked 8th for the keyword “Facebook”. This resulted in a huge amount of impressions, and when they dropped out of that SERP, the impact on their business was at worst neutral. That said, many other tools (looking at you SEMRush etc.) were similarly thrown, and it was a bit of an edge case.

What this means is that impressions needs to be used alongside other KPIs, hopefully including clicks. That’s true of any good metric, though.