In the old days (well, 10 years ago) everyone thought that "hits" were the way you measured success. After all 15000 hits sounded a lot. Then someone pointed out that a hit was pretty meaningless as it just meant the number of items downloaded from your server. This meant that someone viewing your home page might be counted as 15 hits. So page views were a better way of measuring success. Now you had 4000 pages being viewed. But again that is deceptive -- is that 400 people looking at 10 pages or the other way round? And does it matter how many pages they looked at? OK, you say, let's count visitors instead; and because someone might visit a few times, let's count only unique visitors.  So far so good. We've now got what most people consider to be a reasonable way of measuring the effectiveness of your website.

Lies, damn lies, and . . .

You're now able to tell your boss that the Adwords campaign you're spending £400 per month on is working well. After all you've increased the number of unique visitors per month (UVM) by 28%. Your boss is of course very pleased (although a little perplexed as to

why sales don't seem to have increased by nearly so much).

The problem with the UVM figure is that not every visit is equal. First of all there is the worst case -- the dreaded hit-and-run. Typically 60% of your visitors will turn around at the entry page and immediately leave the site. Such visitors "bounce" out and never see additional pages. Given the high bounce rate, you need to stop counting UVM as a success metric. In fact the bounces should be considered a negative as it means that you've failed to convert them to even a second click.

Prospects?

So this leaves around 40% of the UVM as quality visits. But think about that a bit more and of course you'll realise that this can be equally misleading. What you really need to do is ask yourself the question, if I was a visitor looking to buy what pages would I look at? Measure those visitors and you have the proper benchmark of success.