My advice is to start simple and to talk small steps from there.  Don’t worry initially about actionable insights, they will come with time.  Instead, just get your company used to looking at web metrics.  Create a one page report containing basic metrics and start sending it out to key internal stakeholders.  Include some comments but again, starting simple, these need only be what has happened that week.

Based on experience, once departments and managers start receiving data, they will start asking questions.  Set yourself a target of answering 1 to 3 of these questions each week, focusing initially on those questions that you can answer (including the answers in the weekly report).

Over time, as experience grows, the questions will get more complicated and the answer will become more useful to the business, they will turn into actionable insights.  The report that is sent out will evolve as well, as people understand which metric they need to see and which are only nice to know.  KPIs will be defined and used to measure the success of the business.  The weekly report will contain recommendations on how the business can improve and this information will be used.

Just keep in mind that this won’t all happen overnight so don’t expect it to.  if you try to do it all at once, it likely won’t work.  Take small steps, educate the people around you and you can make a real difference to any organisation that you work for.

For suggestions on which metrics to start with, try reading Your First Metrics – Part 1 and Your First Metrics – Part 2.

• the number of people, use unique visitors
• traffic levels, use visits
• quantity of content consumed, use events

Caveat: there are certain times when you want to group people on a certain criteria.  Examples of this include wanting to know the number of people who accessed a certain site section or who arrived via a particular traffic source.  However people can view multiple site section and access a site via multiple traffic sources.  Therefore be aware that the sum of people via traffic sources, site sections and other similar situations will be more than the actual number of unique visitors to a website. 

If you want to evenly attribute a total between different options, you might need to choose a lower level metric.  The important thing in this case is to choose the metric which is independent for what you are looking at.  For example, a unique visitor may have multiple entry types but each visit only has a single entry type.  Unique visitors and visits can view multiple site sections but events are each from only a single site section.

I generally have an excel sheet set up that contains all the historical data for these metrics. Additional metrics can be created by applying calculations to the original metrics with simple Excel formulae.The new data can be added as a new line in this spreadsheet with charts updated accordingly.

The report items that are required for this report is as follows (in my preferred viewing order):

• Unique visitors
• Total visits
• Total page views
• Average duration per visit
• Page views per visit
• New vs returning visitors
• Unique visitors per entry type

These report items have different default time intervals. As we are only looking here at single time periods, the time interval for each report item should be set to “Full period”.

If the data is going to be transferred into Excel, there is no need for charts, additional statistics or descriptions and as such, these options can be deselected. There is no harm in leaving them in the report but I generally like having the report simpler and cleaner.

The report item “Page views per visit” is used to calculate the bounce rate for the website. It is a frequency table containing the number of visits during that time period per each number of page views e.g. number of visits with 1 page view, number of visits with 2 page views and so on. As a bounce can be defined as a visit with only a single page view, the bounce rate is the percentage of visits where only a single page is viewed. Therefore only the top line of data is required and the “Maximum number of lines” can be reduced to 1.

This report can then be set up for either a weekly or monthly report (or daily although I would question how valuable this is) using the time periods “Last week” and “Last month”. At this stage, you should now have a nice report containing all the relevant report items to get a quick overview of the performance of your website.

If this data is being captured and trended in an Excel file, some basic calculations need to be performed to get all my basic site performance metrics. Frequency (visits per unique visitor) and page views per visit can be calculated by dividing the appropriate metrics. The bounce rate calculation is single page visits divided by total visits.

I would recommend calculating the proportions of new, returning and unknown visitors and the proportions of the different traffic sources (entry types). An understanding of the reason for a change in the traffic levels for a website can be gained from looking at both the absolute numbers and the proportions for each of these metrics.

So that’s the basics, I feel everyone and every website can benefit from looking at this data. For real insights, more detailed reports will be required but this should tell you if you need to look further and where you should look first.

As a quick summary, here are the metrics you should have if you have transferred the data to an Excel spreadsheet:

• Unique visitors
• Visits
• Page views
• Frequency
• Page views per visit
• Duration per visit
• Bounce rate
• New, returning and unknown visitors (the number of visitors and proportions of each)
• Entry type – Clickin, Search engine, External referrers, Direct Entry (the number of unique visitors and the proportions of each)

So what does everyone else compare against?

For weekly or monthly data, the commonly used comparison periods are the previous period and the same period in the previous year. By comparing against the previous period, you can quickly see if things got better or worse or even just stayed mostly the same.

I am less sure what a comparison against the same period in the previous year is meant to show. This may have been good in traditional industries but the internet changes too fast for this to be useful in my mind. Your website or traffic mix will change and suddenly you are constantly around 20% off last year every single week, making that comparison just additional numbers on the report.

So what should I compare against then?

The point of a comparison is to aid in understanding if the performance for that time period was good or bad. Therefore, logically, the best comparison to be made is against the expected or forecast performance for that period.

This forecast should have been created using the data from recent periods and using the data from last year, taking into account known or expected website and traffic mix changes. Once you start setting goals through forecasts and targets, you start becoming accountable.

Now this is slightly scary, it is so much easier to say ‘hey, we are up 20% on last year, we must be going well’ (ignoring how you were 40% up two months ago). But when you are admit that ‘hey, we are down 1.5% against where I thought we would be’, you can start to investigate the reasons why and see what is and isn’t working.

Anything else?

While I don’t like comparing against the same period last year, the trend or % change for the same period last year can be useful. This can show you if a change this year is merely a seasonal effect or something that should be investigated further.

Targets are slightly different from forecasts and not as useful for short term comparisons. They are where you want to be longer term, something to keep an eye on but not a call for action if not at that level immediately. And like everything else, they should be one of the factors that are used in developing the forecasts.

Summary

Give your data company. In order to provide it with a meaningful existence, it should be compared against other numbers. The most useful comparison is against a forecast as this is the true measure of whether performance was above or below expectations. And the forecast should draw on the knowledge of all other numbers that could be compared against.

Note: I will be writing some posts about forecasting in the future. If you need a forecast to understand performance, you need to know how to forecast (ideally with some degree of accuracy).

The set of metrics that I listed in the post last week (visits, visitors and interactions) could be described as the basic traffic metrics. Each are absolute numbers as opposed to a percentage or ratio and each can be used to describe the popularity of the website during the previous period.

The following set of metrics are what I describe as (for want of a better term) as engagement metrics. I do not claim that they describe completely that mystical experience known as a user’s engagement with a website but I believe they give an indication of the level to which people are engaging. All are ratios calculating by combining two other metrics.

Frequency – Visits per Visitor

This is the average number of visits each visitor to the website made during the specified period calculated by dividing the number of visits by the number of unique visitors for that time period. It reflects whether the users of a website are interacting with it on a regular basis or only ever visiting once. The minimum number for this metric is 1.0 – every visitor must have made at least one visit.

Interactions per Visit

This is the average number of interactions made with the website during each visit during the specified time period – calculated by the dividing the total number of interactions by the number of visits for that time period. Again, the usual type of interaction that is used here is page views.

It is an interesting metric in that there is no good or bad result, it all depends on the context. For example, a higher number of interactions per visit may suggest a greater level of engagement but it could also be that the user was having to search through the site in an attempt to find the information they were after.

Duration per visit

Similar to the previous metric, this is the average time a user spends on a site in each visit. It can provide additional information about the average visit regarding the length of time people spend on the website.

A key point to remember though is that web analytics packages can only work off known measurements and there is generally no measurement for when a user leaves a website. Therefore the duration for each visit is typically under reported and cannot be measured at all when only one measurement is received.

Bounce Rate

This is a measure of the number of time a website is visited and the user leaves immediately without engaging with it at all. The simple method of judging whether there was any engagement is to assume that the user has to have more than one interaction. Therefore bounce rate is commonly calculated as the percentage of visits with only one page view.

A website that is used as the user’s home page (portals, google, news sites) would have a high bounce rate for all visits where the user opened their browser but then immediately went to an alternative site. Also, it possible for a user to get all the information they need from a website from a single page – high engagement but only a single interaction. Basically every site has a certain level of bounce rate, it is how it changes over time that is important.

What Next

So these 7 metrics would be your basic set of traffic and engagement metrics, and the basis for any investigation into the performance of a website. From here I am going to go into some methods of how to look at the numbers and how to understand better what they are telling you.

Note: we are dealing with the average in all the engagement metrics and this could produce some misleading results unless you think through what the change actually means. For example, a influx of traffic from a new source may lead to a drop in the overall site frequency as the new visitors have a lower average frequency than the current average average, not that people are less satisfied with the site.

While I will cover this specific concept further in a future post I think the key point I wanted to make is that numbers should not be just taken as black and white, you need to do some thinking in order to understand a change. And you need to understand a change in order to do something about it.

The common theme that runs through any discussion of web analytics is that the industry needs to move beyond a focus on metrics and simply providing reports to instead focusing on the provision of insights and recommendations. I totally agree, in order to add value to a business and justify our salaries, that this is what we need to be doing. But the other point that is often raised is that web analytics is really hard – it is the identification of these insights and recommendations that is the really hard part. I believe that everyone needs a starting point, some simple metrics that may not provide any real insights but give the reader a quick overview of performance. While simple and possibly not insightful in themselves, these metrics do indicate if some aspect of performance has changed and they can also give some hints about the areas that should be examined in more detail in order to do that adding of value.

Traffic Metrics

The starting point for any review of the performance of a website must be the level of traffic that it has. That is the number of visits to the site, the number of people making these visits and the level of interaction they had with the site. These really are simple metrics but they instantly tell you if something has happened to your site performance. An unexpected change to any of these metrics would usually be the first indicator that something has broken or that the new marketing campaign is having an impact.

Visits

Even I find it weird that I am starting with Visits ahead of Visitors but my gut instinct feel that this is an under rated metric is constantly growing stronger. I am starting to think that the visit, that is a sequence of one or more interactions with a website ending with a period of 30 min of inactivity, is actually the foundation metric for web analytics. From a user point of view, it covers all interactions they have with a website during the period they have engaged with it. Understanding user behaviour during this period is a key element of developing websites and/or marketing that meet customer requirements and expectations.

Traditional offline marketing has a big focus on the importance and value from segmenting customers. The online world can take this a step further by segmenting at visit level. For me, the best type of segmentation is based on customer needs/thoughts/actions. Now a person may have many reasons for visiting a website and to assign them to a single bucket reduced the understanding that can be taken from segementation and an investigation into user behaviour. This is improved by segmenting at visit level, placing into different buckets based on method used to get to the site and actions taken once there.

Visitors

Quite simply the number of people who accessed the website during the period specified. I do not agree with the concept of there being two metrics, visitors and unique visitors, where visitors is the number of people accessing the website per day totalled for the period specified while unique visitors is the actual number of people for that period. Adding visitors up over a number of days and calling it a total is not mathematically or logically sound nor does it provide useful information. Instead, for me, it is unique visitors at daily, weekly, monthly or campaign level all the way.

This number is useful in understanding if more or less people are accessing the site and how many people it actually is who are looking at all these pages, placing these orders, clicking out to other sites. There are issues with the accuracy of this metric due to cookie blocking, cookie deletion and multiple computers but the solution is to accept this fact and move on.

Interactions

Traditionally this metric was the number of page views and it indicated how much visitors were navigating through the site. This concept is changing due to the introduction of so many new ways of presenting information on a website, many interactive and not requiring a page to be viewed. Hence my use of the term ‘interactions’, suggesting this could measure the level to which visitors have clicked on or viewed multiple elements of a website, whether this is a page or something else.

For simplicity, I think many sites can still start with using page views as a measure here. In my previous role at Ask.com, we used the number of queries made as our measure of interaction as that was more relevant than the number of pages viewed. Another alternative could be clicks or screen views viewed – the key point is that it should always be an action that every visitor with even minimal engagement with the website performs at least once. More detailed (and insightful) metrics will cover areas that only a proportion of visitors perform, such as placing an order or viewing a certain type of page but the number of interactions gives the brief overview that is a good starting point.

In my next post, I will go on to describe the other simple metrics that I believe are a good starting point for examing the performance of a website.