Analytics 101: Hits, Pageviews & Visits

What is a hit? How is different than a page view? What constitutes a visit? Let’s will explore the definition of these basic building blocks of web analytics, and also explore how these basic metrics misbehave, causing misleading data and reports.

Definitions

Hit – A hit is a request from a viewer for a file from the server.  For example, when a webpage loads it typically include numerous requests for files such as index.html, style.css, script.js, logo.jpg, image1.jpg, image2.jpg and image3.jpg.  Each request is a hit, for a total of 7 hits just loading your homepage!

Page view – A page view is one of the top five metrics and measures the number  of times web pages are loaded. This metric corresponds to a recording each item a customer picks up while in a department store.

Visit – A visit (or visitor) is a sequence of page views by a single customer.  This metric corresponds to a a customer walking into a department store.

How to use these metrics

Hits, page views and visits are seemingly straightforward metrics, but an analytics guru knows to take each of these metrics with a grain of salt.

First and foremost, forget about hits!  Hits are a holdover metric from the early days of the internet.  At that time, webpages were a single page, with no images, style sheets, javascript or other files.  At that time, a hit measured the number of page views.  In today’s world, if you hear someone talking about hits, RUN AWAY! (okay, they may server one purpose when defining server specifications, but if you ever need to worry about this, you will surely know it).

 Second is page views, these are a top-five metric because it tells what legitimate pages (as opposed to hits) your customers viewed.   There are many things you can do with page views:

  • Look at a total count for the past few days, weeks, months and years to see how you are trending.  This is a report (vs actionable analysis) but it puts things in perspective.
  • Segment by page and see what pages are popular and which are not.  This is still a report, but knowing what is popular and unpopular means that you can focus your efforts.
  • Segment pages by entrance source such as external search engines (Google), pay-per-click campaigns, and links from external pages.  Is the page views rather uniform across the various segments, or to search users prefer certain pages while link users prefer others? Using this, you can begin to understand your audience.  Should you be promoting certain pages?  Search engines are good ad finding deep pages, but your other users may not be able to understand your complicated layout?  Unless you can rationalize discrepancies, assume that something can be improved.

Third is visits, which is provides a glimpse into your reach. How many different people are interacting with you?  With this information, you can determine:

  • Similar to page views, create charts showing total visitor counts for the past few days, weeks, months and years.  These reports show your general trajectory.  Sadly alone, it does not provide much actionable insights, but it puts more advanced metrics in perspective.
  • Segments by source again.  Where are your users coming from?  Search engines, pay-per-click (PPC), social sites, links, direct access?  While you may have alot of users, if they are coming from PPC, then they are somewhat costly/artificial.  If they are all from search engines, great (free!), but perhaps you should expand into social.  This view of your users provides an actionable insight.  Is what is happening what you want to happen?  Percentages for each industry vary considerably, but you should never have all your incoming costomers from one stream.  Diversify.

Common misrepresentations

Hits, page views and visits are seemingly straightforward metrics, but an analytics guru knows to take each of these metrics with a grain of salt.

Hits – Hits are an great way to over estimate your true customer base.  Sine each page view can easily consist of up to 50 hits, saying your “site received 100,000 hits” sure beats “2,000 visitors” to the unknowing listener.  Don’t use them and if your someone else using the metric, stay away from them like the plague.

Page views – Two common mis-representions here are counting unqualified pages as a view (like a pop-up ad or error message).  With today complex webpages, some people confuse a pageview with an event, and there certainly is a grey area.  Another issue is improperly tracked pages (typically not on purpose), where each page registers twice, thus doubling your page view count!  All in all, this is a rather standard metric that you can use reliably.

Visits – Like page views, visits are generally well represented, with one exception.  If customers are viewing without cookies enabled (eg. using private browsing), then most common analytics tools are unable to know when a user is new vs coming from another page within the site.  This means that some visits are recorded as a series of single page visits, and not one visit of multiple page views.  Expect this probem to get worse before it gets better as privacy issues becomes more and more popular.

Summary

In all, ignore hits and focus your efforts on page views and visits.  These latter two metrics form the basis for powerful actionable insights.

Have a questions? Ask it below and I will respond.

Gregory Moore, founder of Uptick Analytics, is a data scientist and analytics guru. With a Ph.D. in the fields of optimization and data mining, he is passionate about finding actionable insights for better performance. Contact Greg at greg@uptickanalytics or on Google Plus

Posted in Uncategorized

Leave a Reply

Gregory Moore, founder of Uptick Analytics, is a data scientist and analytics guru. With a Ph.D. in the fields of optimization and data mining, he is passionate about finding actionable insights for better performance.

Analytics Newsletter

Enter your email address to join our mailing list.