The Hillwatch Analytics Briefing Series for Public Sector Managers

Getting Down To the Right Numbers, Part 1: Hits are good in hockey, but not so good on the web

"We have a web site and I am absolutely amazed at how valuable a tool it is getting. We are getting about 7500 hits a day, which absolutely staggers me."
    Leadership Candidate, Paul Martin speaking at the 2003 Crossing
    Boundaries Conference

In this brief we will:

• Explain why �Hits� are not a useful measurement of web activity;
• Demonstrate how hits provides a distorted view of web traffic; and
• Show how to look beyond �Hits� to focus on �Visitor Sessions�

Use of the term �hits� has become engrained in peoples mind when they refer to Web site traffic.  Most senior managers when asked about a site almost immediately start referring to the number of �hits�.  Some programs now measuring �hits� as one of their performance indicators in their RMAF process. When using the term most managers believe they are referring to the # of visitor sessions or # of page views but these are not what a �hit� is.

Hits are an outdated technical remnant from the days when the sole purpose of web logs was to measure server activity. When you or I download a single web page, our web server log file separately registers all the downloadable file images and hidden functions and records each one as a �hit�. Each page we view may generate a dozen or even hundreds of hits.

In other words, �hits� are a totally useless reflection of site traffic.  Worse, they are as useless as a measurement metric because they provide no information that can be used to understand or make improvements to a Web site.  Despite this, the term has stuck and is now part of the vernacular. 

We don�t consider people at the senior level of the organization misusing the term �hits� a serious issue, as long as the 'sherpas' below are consistently supplying them with the correct data.   However, if unfiltered data from web traffic floats around the organization, it would be easy for many managers to mistakenly conclude web visitor traffic is far higher than it actually is.

To illustrate how misleading hits can be, Table I below presents a case study Web site with six-months of cumulative traffic data.  The numbers have been taken directly from a report generated by the leading provider of web traffic analysis software.  As we can see, the number of hits is very impressive, but since it doesn�t deliver any information that can be tied to actual visitors, it is meaningless from a site management perspective.

Table I: Comparison data for hits, page views and visitor sessions (6 months)

Instead, of �Hits�, Hillwatch recommends that you get in the habit of referring to �visits�, �visitors� or more accurately �visitor sessions� or �user sessions�.  A �visit� is what a visitor is engaged in when viewing site pages.  It could be repeat visits by a single individual or a one-time visit but cumulatively it is one of the key measures of broad public interest in your site.

That said, there are some real reliability issues with visits.  First and foremost is ensuring that your numbers accurately reflect visits from real people.  Sound odd?  We�ll explore this subject in our next brief, Getting Rid of Spider Infestations: The Need for Scrubbed Traffic Data

Hillwatch�s Previous Analytic Brief

The Public Sector's Problem With Web Traffic