Measuring Readership

I’ve been thinking about how we might more accurately determine how many times a blog post is read. Not just loaded in a browser but actually read.

Firstly, there are different levels of reading.

Level 1: Evaluate Source

This is just a quick look to see if the post looks worth spending time on. If the page is full of click bait junk then a discerning user might just back out here. If it looks like a decent quality site then the user will hopefully progress to a…

Level 2: Quick Scan

This is reading the post but at high speed and without taking in all the details. The user might skip over any introductory or background content and try to get straight to the meat. When I scan a piece I tend to navigate by subheadings or by reading the first few words of each new paragraph or section.

Level 3: Full Read

This either means the user knows the blog already and trusts its quality or that he/she has the time and interest to read it properly.


Using Google Analytics and filtering we can establish the type of visit rather than it just being a page load.

First up, try to filter out any bots, spiders or non human visitors. When you set up a view Analytics has a nice checkbox option for this.

The next step is to break the visits down into the levels described above by looking at the length of time on the page. It’s not quite a simple as setting fixed ranges as blog posts will vary in length, so what might be a full read for a short piece could be a scan for a longer one. So, we need to use the word count and the time on page to calculate the level. This means we need 3 ranges of reading speeds which reflect our 3 levels.

Setting the reading speed levels is largely guess work but I’ve tried to apply some logic to it. Some reading speed research shows that the average reading speed for an adult is around 300 words per minute. That’s five words per second. So, if we get the word count and divide by 5 we’ve got the rough amount of time in seconds that an adult should need to be on the page to read the whole article.

So, if the seconds spent on the page is greater than or equal to a fifth of the word count then we have level 3, the full read.

At the other end of the spectrum, what can we discount as an evaluation or bounce? How long can you be on a page before it’s considered that you might be scanning its content? I feel this can be a fixed value rather than relative to the word count as our initial evaluation is done very quickly. Unscientifically, this part is really just plucking a number out of the air but my gut feel is for something like 10 seconds. That’s taking in 50 words. Below that I find it hard to consider that any content has really been processed. I can’t think of a good way to test this out but I may canvas opinion to refine the figure.

So, we have a page visit time of less than 10 seconds as our level 1 evaluation. That leaves anything in between as level 2, the scan.

Fine Tuning

Of course users could spend time on the page without necessarily reading anything. We can add some other details to increase the accuracy. Using JavaScript we can listen for a scroll event or better see if the end of the post is shown in the viewport and then send this as a dimension. A lack of scrolling or content visibility would certainly help to qualify the evaluation visits.


Please comment if you think of any further refinements or see any flaws in the logic. Thanks.

Talking to Machines

“Hey Siri”, “OK Google”, “Hey Cortana”. It still feels weird to me.

As I’m a Windows user Cortana is my “virtual assistant” of choice. It’s an incredibly useful service and getting better all the time so I wonder why I don’t use it more. I’m learning slowly. For example, I just wanted to work out 500 divided by 52. In the past I would have opened the calculator programme or app, typed in my request and got my answer. Today, however, I just said this: “Hey Cortana, what’s five hundred divided by fifty-two.” And, about a second later I had my answer. Much faster.

I caught myself saying, “thank you”, and immediately felt like an idiot. But actually, that’s not shameful at all, just polite. So what if it was redundant. I’m still human.

I think there are 2 main reasons why I don’t use it more. Firstly, I just don’t think of it. I’m a bit set in my ways and am too used to typing. Secondly, I’m self-conscious about using it within earshot of other people. It still feels a bit weird, talking to your computer, and I imagine others who can’t hear clearly enough to make out my words wondering who I’m talking to. So, I tend to only use Cortana when I’m on my own, which is even weirder, right? Though, as I never see or hear anyone else using these assistant services I’m guessing I’m probably not alone in this.

I’m going to try to push through the weirdness and make it more normal in my circles.

Just for fun, if you’re using Cortana, say this: “Hey Cortana, what does the fox say?” Do it a few times – you’ll get a variety of answers. I guess it probably works on the other assistants too.



iOS vs Android vs Windows in 2017

Everyone has their favourite big web company and associated operating system (OS). We’re all either Apple, Google or Microsoft. Maybe not quite all of us. There’s always one weirdo who finds something “better”.

I’d say that for office desk based work most people are either on Apple or Windows (Mac or PC). I’m not sure if any significant numbers use a Chromebook and Chrome as their serious OS. I may be wrong. For phones it’s iOS, Android or, to a lesser extent, Windows. Do people still use Blackberry in big numbers? I’m assuming not.

People who use Windows for work might have any kind of phone. It’s actually pretty unlikely that they’d have a Windows phone as it’s smaller in the phone market. On the other hand, I bet almost everyone who uses a Mac has an iPhone. And without even considering any of the other options.

I’m in quite a lucky position where I work on a Windows laptop, use an Android TV box and an iPad at home so have access to all 3 operating systems. I thought I’d do a quick comparison, nothing scientific, or even particularly fair, just my own impressions. I should also point ut that this isn’t about bashing anything. I think all 3 are great. More or less they all do the same things. It’s the little differences and the UX of them that becomes interesting. Here are some pros and cons for each.

  • iOS and Android are miles ahead of Windows in terms of apps available.
  • iOS is on Apple hardware only whereas you have more choice with Android and Windows.
  • The point of entry in terms of price is much higher for iOS.
  • Android (Material) and Windows have more modern UI whereas iOS doesn’t seem to have changed very much in the last few years. Not necessarily a bad thing, just an observation.
  • iOS has Safari as its default browser, which is not quite as feature rich as Chrome or Edge. It’s easy to install Chrome though.

For what it’s worth, I have a Windows 10 phone (Microsoft Lumia 550) and I absolutely love it. The main drawback is the lack of apps compared with the other platforms but I don’t find that to be a major obstacle. Since Windows 10 there’s now a first class web browser in Edge so most services which don’t have a Windows app seem to have a web equivalent.

My phone cost me £55 (about $70) and does everything I need, whereas a new iPhone would probably cost me about 10 times that. My phone doesn’t have a great camera or loads of storage, as you’d expect at that low price but I have a digital SLR camera for that. I’m not going to pretend my Windows phone is anywhere near as good in spec compared to the latest iPhone or but I’m also not convinced that my overall mobile experience is 10 times worse. I feel I’m getting the better deal.

Don’t put all your eggs in Google’s basket

GoogleWorking on your website’s search engine optimisation (SEO) is all well and good but it’s a dangerous strategy to use this one source of web traffic in isolation. I’ve found this out the hard way.

Yes, there’s no disputing that appearing at the top of the Google search results for the right phrases will bring in visitors but using this as your sole strategy could be very risky. You shouldn’t be too dependent on Google as things can change very quickly.

I’ve experienced it myself a number of times. I’ve had my site sitting pretty in the top 5 results only for it to suddenty plummet down to the fourth page or lower or vanish from the search results altogether.

Google ranks web pages and other documents by following an algorithm, a complex formula. As they continually try to improve their search technology they tweak this algorithm from time to time. Depending on the change this can have quite dramatic effects on ranking. Through no fault of your own you can suddenly lose your foothold on a particular phrase or market.

Google indexes websites by periodically visiting them with it’s search robot, Googlebot. The frequency of these visits depends on the frequency of your site updates. If you’re unlucky enough for your website to be down at the time when Googlebot visits you’ll drop out of the index and have to wait until it drops by again before reappearing.

The other big factor is competition. In competitive markets your competitors will be working tirelessly to try to grab the top spot. Changes they make to their sites or search engine marketing activity can easily leave you playing catch up.

I’m not suggesting for a second that you should stop pushing for a high ranking on Google but you should be doing other things too so that should anything change your site doesn’t grind to a halt.