SaaS Web Analytics - Part 1

Web analytics has been around for a very long time in the form of java applets, php scripts or through the use of standalone applications that trawl through server logs. Historically if you needed more insight than the basic number of pageviews or unique visitors you had to write your own tools or buy into huge service contracts with IBM for example.

The industry has now evolved with the likes of Google Analytics economising access to hosted tracking tools. Web data is now being leveraged to offer valuable insights into customer behaviour, drive website design, influence product decisions and measure the impact of marketing.

Recognising the value in tracking data, many companies have sprung-up to compete with Google:

Do they all do the same thing and why should you pick one over the other?

Things to Consider

The way you interpret data can result in very different conclusions. Often quoted as evidence of the popularity of a site, the number of 'active users' is a very important metric. An 'active user' is not a standardised SI unit and can therefore give misleading results. Understanding what this means and how you define its relation to your own website is very important when trying to draw reliable conclusions.

An active user could be somebody who visits the site once a week, another definition might be that the user has to post a comment or at least be on the site for 5 minutes to count as active. The definition of active will vary on what your website is trying to offer. Is connecting to a webpage but minimising the window and forgetting about it still an active session, what if you're listening to a podcast on a website in the background?

With more sophisticated tracking (combined with the added information a client browser can provide) you have more data and more ways of creating an accurate decision threshold. You could track specific events, mouse movements, detect scrolling, trace how the user found your site and a plethora of other options. If you could do all this in real time you could react faster to live events, is your website trending on twitter, has a TV advert going out on air caused a surge in traffic, how successful was sponsoring Glastonbury over the weekend? What if a bot network is spamming your website? Can I accurately update a 'most read' or 'currently trending' scoreboard of content?

The quality, quantity and speed at which this data is available are all key factors in deciding which analytics platform to choose.

Other features to consider are:

  • How hard is it to set up and configure the tools.
  • How reliable is the information that the tools provide.
  • How easily can the data be communicated to relevant team members.
  • How quickly can you react to key metrics changing.
  • How expensive is the service.

Google Analytics

By far the most popular analytics service, Google Analytics evolved as a platform to help support its AdWords pay per click product. Launched in 2005 it offers a completely free tier of service but is still very focussed around advertising, e-commerce and user clicks. Tracking user engagement can be more tricky and lead to rise of competitors specifically focussed around this such as Chartbeat.

It uses javascript and pixel tracking to offer historical analytics and semi-real time tracking (inferred to within 5 minute windows). A limitation of this is that if a user is reading lots of content or watching a stream etc then that the user session may be inaccurately represented if they haven't triggered a new event during for a long time.

The data can be presented in dashboards but they're up to the user to implement for themselves and customise how they work. However saying this you can implement custom dashboards as plugins however its still far more fiddly than some of the other solutions available. Therefore getting more useful data out of Google can often take some trial and error and it isn't well suited for real time applications.

10 million hits can be tracked each month for free however if you want more advance features there is a flat rate annual fee of £90,000. However offering so many features for free there is no risk to trying it out.


They utilise real time continuous pinging between the client and tracking server to accurately measure session times. It can also track users across different devices and platforms to accurately count the number of unique visitors.

We calculate the "active" visitors number by combining a number of factors including whether or not a visitor has your website open in their frontmost tab or window, mouse movements, scroll position, and key strokes.

In contrast, Google Analytics defines an active user as active if he or she has triggered an event or pageview within the last 5 minutes. GoSquared prides itself on its extremely easy to use interface and simple set up procedure. Its dashboard is beautifully animated, responsively designed and fully interactive and requires zero input from the user to configure.

It offers all of your key metrics at a glance in real time whilst also doing trends analysis and e-commerce reporting. In a world where everybody has multiple devices, their newest feature, People Analytics can accurately profile people from the numbers. Tying together multiple sessions from multiple devices and identifying unique user ID's from the tracked website, GoSquared can accurately count individual new 'people' instead of just misleadingly inflating the number of hits.

This is a very powerful feature that can also take that person's unique identifier (usually an email address) and draw connections to their social media accounts. This allows you to search/filter through your user base by somebody's Twitter/Facebook/Linkedin etc account and reach out to them on their preferred platform. In addition it has a very well documented API for integrating the analytics reporting easily into other services.

GoSquared offers a limited usage free tier and a 2 week free trial however more serious usage requires a monthly or annual subscription. The price you pay is a tiered model that offers significant volume discounts and works out substantially cheaper than Google's premium pricing. For its simple set up and instant usability, the speed at which you can react to changes makes it a very compelling choice for many.

I will discuss the other popular platforms in part 2 of this post.

*indicates required