Web Metrics: Don’t be slave to the next HIT

Early website management was obsessed with volume. Today, an increasing number of page impressions can mean a website is failing rather than succeeding.

I still hear senior managers and journalists quoting HITS when they want to say something impressive about the Web. (HITS stands for How Idiots Track Success.) A HIT is a totally meaningless measure, so why is it still being quoted?

Many years ago, when the Web was still young, the web team was desperately trying to prove their worth. A message came down that a vice president was going to give a speech and in that speech he’d like to mention the website. He needed something to say that would sound impressive.

So, the web team got together and brainstormed. Somebody got out the website behavior data and began to pore through it. Then suddenly they shouted: “I have it! I have it! I know what we can tell them. We’ve got a zillion billion HITS!! Look, look, have you ever seen a number like this?” And then they all gathered round and marveled at such a big number. Then they danced around and drank orange juice.

Senior managers don’t particularly like being made to look like fools. One of these days, someone is going to explain to them that HITS is worse than useless as a measure. Okay, so page impressions or visitors are a little better. But are they really? If you are a site that lives off advertising revenue, then yes, but if you are any other website, then no.

The Web has given rise to the phenomenon of the “word farm.” Websites that generate their revenue from advertising pay as little as possible to get writers to churn out low-quality, high-volume content. Multiple versions of the same article are created with just enough modifications to fool the search engine.

“Collections of articles are sold in packages, for as little as $1 per 500-word piece,” Danny Bradbury of The Guardian writes. “Customers spin them into thousands of articles designed to draw traffic to sites laden with adverts or other profitable payloads such as email collection forms or credit card payment systems.”

I have talked to people who ran websites who refused to take down out-of-date content because it would reduce their page impressions and mean that they were less findable in Google. Think of a website filling up with products that you no longer sell, analysis that has been proven to be wrong, conferences that were cancelled, events that happened in 2002, and all because someone is a slave to volume, desperate for another HIT.

Let’s say you launch a new product and there’s a problem with it, and you get lots of confused and annoyed customers coming to your website. Is that a good thing? Let’s say your website is so confusing that it takes 20 clicks to do something that can be done on a competitor’s website in 5. Is that a good thing?

Word farms of the web: The Guardian

 

10 responses


  1. So if a ‘HIT’ isn’t a meaningless measure, what should I be looking at to determine the ‘popularity’ of my site (and the content within it)?


  2. Shona… one thing to consider is how long people stay at a website page after they “hit” it, especially when its a page that you’ve made a marketing effort to get people to visit. In most cases, fewer people spending more time will tell you that they like the content they find rather than more people who peek and scoot.

    Another element is whether or not people take an action you want them to take, whether that’s completing an inquiry form, making a purchase, or following a link to another place on your website.


  3. Hi Gerry – you’re talking sense [on both subject areas] as usual.

    Re HITS : a sure sign that the user of the term has no idea about e–commerce/e–marketing – they may as well be wearing a hat with flashing lights and a sign that says ‘nuggett’. Not always their fault of course – just last week a BBC article [on TV] included reference to a web site’s hits. Unless someone tells you – or you try to find out – why should you know it is a useless measurement?

    Re word farms : I know the argument that if you have 1000 [or a million] such sites and they all get one clickthrough per day you make money, but who does click on ads on these sites? Often the contextual element is miles off, and even if there is some synergy, the leap between the reader and the ad is a giant one for mankind. [eg why would a student in the UK searching and reading up on ‘online security’ click on a link to a community security service in Chicago?].

    That said, it can be hard for the search engines. A poor example, because it doesn’t carry ads, but some would consider the content on my web site to be un-worthy, but I have the right to publish it. Who are the search engines to penalize me in the rankings because my syntax might not be spot on? At least the word farm publishers have made some effort with content, it’s those ‘ad farms’ that really bug me when they appear at the top of SERPs.

    And don’t get me started on the rubbish that is on some web sites [perhaps word farms] that students take as being accurate and then reference it in their assignments.


  4. Re word farms - I forgot to say: surely the ad networks don’t pay out on an ‘ad impressions’ set up do they? If this is the case - kerrrrr-ching.


  5. What combination of other trackable metrics can provide a more holistic view of the effectiveness of a site and its content? The key here is that the process needs to be repeatable and can be done with scheduled frequency.


  6. I’m with Shona on this. I need to know what measures we need to look at - if any are valid?


  7. The key measure I believe is task completion. If a student comes to your website looking for a course, did they find it quickly? If someone wants to renew a passport, did the website help?

    How do we do this? I think we have to regularly usability test. We have to get our customers in front of our websites and ask them to complete top tasks.


  8. You’re right, Alan, it’s all about volume. You get to this word farm website, you’re allowed. But there’s this ad that at least looks something like what you’re trying to find–you might click on that.


  9. Problem with page views and even how long (time) they view can be easily manipulated.

    I for example load up several pages at a time and then as I break slightly from work or wait for a file to upload, download etc I will ‘alt tab’ and read a bit. So the page may be loaded on my browser for 2 hours but I probably only spent 1-2 minutes on it.


  10. Gerry, i think you’re making a good point here. Metrics itself are useless. Always ask yourself, why am i reporting, what do these figures mean to me, and what action should i take.
    The ultimate key measure i believe should be ‘added value’. Popularity of a site is not a measure. Only in context of strategic objectives and user expectations this measure is a meaningful figure.
    Metrics should drive you to take action to optimize your key business processes and tell you something about the value you or your website added to the chain.

One trackback

Leave a Reply