All the social media metrics bloggers have disowned impressions as a useful metric, just like all the Web marketers before them. Impressions data is actually very important to understanding how effective your marketing campaign is.
Of course, there are some wrenches in the monkey works when it comes to analytics data these days: all those bots out there are clicking on everything and inflating your page views, visits, bounce rates, time on site, etc.
Actually, bounce rate is a terrible metric by which to gauge Web marketing success (all Websites have a 100% bounce rate because eventually everyone leaves the site). But analytics tools make bounce rates even less useful than they could be by showing you aggregated bounce rates and expecting you to “drill down” to the more informative levels of information.
But let’s back up here. Bounce rates and impression counts go hand in hand. You cannot have one without the other and numerous attempts to filter out rogue impressions compare multiple points of impression per page.
That sounds pretty fancy and sophisticated, doesn’t it? “Multiple impression points per page” just rolls off the tongue like it was made to be said since Tim Berners-Lee decided he hated gopher.
If you want to calibrate your “page views” with real people who actually see your page you have to count impression points. An impression point is every secondary file that the browser should be requesting when a person visits your page. These impression points should occur within a very narrow time frame. Of course, some bots can now extract all the file references and send them back at the server just as fast as the browser (in fact, some bots now use browser rendering engines to do just that and more).
But let’s assume that the team down at the analytics lab have figured out to eliminate the bots from your reports. You are looking at pure, genuine traffic data now. Real people hitting real Web pages with real intent to do something.
Here is how impressions data can show you what is going on:
- On long pages, impressions data drops off the further down the page you go
- On interactive pages, impressions data drops off as you require more action from the visitor
- On pages optimized for multiple platforms, impressions data should vary by viewing device
Impressions are the First Stage of User Engagement
I like to point that out to people, especially when they tell me they don’t waste their time with impressions (or “page views”) any more. Trust me: if no one is looking at the page then no one is engaging with it.
So any trend you see in your page views data sets a baseline against which you should be computing or comparing other engagement metrics. Valid engagement metrics include things like:
- Clicked on a link
- Left a comment
- Shared the document (a subset of “click on a link”
- Bookmarked the URL
- Scrolled down the page
- Kept the page loaded in the browser for more than (10, 20, 30, whatever) seconds
You really cannot measure bounces and time-on-site very accurately because you actually never see the user’s last action. Without a record of that last action your analytics software is just making (often bad) guesses at what the user did next.
So much as you may love “time on site” and “bounce rate”, these are NOT valid engagement metrics. They are semi-engagement metrics, which is not necessarily a bad thing but they are definitely less useful than whatever other engagement metrics you can capture.
If you’re not seeing a dropoff between impressions and one of these other (secondary stage) engagement metrics, something is wrong. Every visitor should see the page for at least a second before deciding what to do next. Studies indicate that most people figure out what they are going to do next after landing on a page within the first 3 seconds. If it takes your pages 5-10 seconds to load and render you should see a higher bounce rate than if it takes your pages 1-2 seconds to load and render.
Bounce rate is the ratio of impressions to “no other actions taken” (or to “other actions taken”, depending on how you want to look at it). All sites have 100% bounce rates because everyone leaves eventually, but the bounce rate you want to use is the per-page bounce rate that tells you whether any secondary action was taken.
Impressions Measure Reach Better Than Anything Else
Another great way to use impression counts is to map reach against them. Here you need to look at Unique Impressions, but there is some disagreement among analytics experts whether you can rely on IP address or need to use cookies. For example, you may want to measure multiple contacts with the same person in your estimate of “reach” because you may need to talk about a “reach” where your page is utilized as a resource, not just a “reach” where someone visits the page.
You could also track “bot reach”, which (roughly defined) refers to the number of bots that are hitting your page (assuming you and the analytics team can agree on what is a bot and how to track them). Some people like bots. For example, suppose your RSS feed is integrated into multiple feed aggregation accounts. That could mean that someone is reading what you write through another Website or an app (although most publishers would prefer their readers see the ads they publish rather than the ads the aggregators publish).
Reach is one of those funny marketing words that has been used in more ways than it should have been. At one time many publishers sought to entice advertisers by claiming to “reach 80% of the Web”, meaning that 80% of all Web users might see their content. This kind of reach was considered misleading and people stopped talking about it after a few years.
In truth it was more like a potential reach. Reach should really only be used to measure who (or what) actually sees whatever it is you are “reaching out with”. If you consider yourself a content marketer your reach should be measured in terms of the number of new people you bring into contact with whatever you are marketing (this is just a measure of consumer awareness, not sales).
But if your content marketing strategy requires multiple contacts/connections per consumer then you have to define a more sophisticated “reach” to measure.
Your regular readers can also help you define a marketable reach statistic. If you publish an article today, how many of your regular readers can you expect (project) to read the new article? 100? 200? 1000? This kind of impression count is important for people who sell advertising in newsletters. It’s also an important metric for people who practice affiliate marketing.
If you know how many regular readers you have to reach in order to trigger a referral commission then you know how much you have to build your readership before you arrive at a level of income that is comfortable for you.
So reach is not one thing; it is many things in many different contexts. But all of these different reach measurements begin with impression counts. Or maybe I should say “qualified impression counts”.
The Last Impression
Hopefully by now you understand just how ubiquitous impressions are in the field of analytics and user engagement. This kind of data is fundamental to what we do. You cannot ignore it, although you can certainly enhance it by bundling derivative value into it.
A good analyst understands that everything starts with impression counts. We just need to do something meaningful with the data.
Read More about Search Engine Optimization
How Long Does It Take Google to Credit A Website with Links?
Natural Backlink Profile: Endless Ways to Build One
Website Not On Google? Why Some Internal Pages Aren't Indexed
RankBrain and Neural Matching: What Is the Difference?
Follow Reflective Dynamics |
Click here to follow Reflective Dynamics on Twitter: @refdynamics. Click here to follow SEO Theory on Twitter: @seo_theory. Reflective Dynamics' RSS Feed (summaries only) |