Understanding Web Analytics and Key Performance Indicators

This article is an exclusive pre-publication extract from the 3rd Edition of the bestselling digital marketing book, Understanding Digital Marketing by Damian Ryan.

It covers:

What are KPIs?

The concept of KPIs is nothing new, and has been common in the world of business analysis for many years. KPIs are used to distill key trends from complex, often disparate pools of data, and to present them as a series of clear, unequivocal indices – a snapshot of how your organization (or website, in our case) is performing at any given time. KPIs do ‘exactly what it says on the tin’. They indicate progress (or lack of it) in areas that are key to your website’s performance.

Why KPIs are important

The real value of KPIs is that they let you extract meaning from your data at a glance. Without them, it is all too easy to drown in the proliferation of data that your web analytics solution churns out. It is a classic case of not seeing the wood for the trees. By defining and measuring your KPIs you are creating a regular snapshot that allows you to monitor the performance of your marketing over time.

You know that if this KPI is going up it means one thing, if that one’s going down it means another, and so on. Your KPIs not only give you an immediate sense of the overall health of your marketing, but also help to highlight potential problems, and point you in the right direction before you delve deeper into your data looking for solutions

Choosing effective KPIs

In the document ‘Web Analytics Key Metrics and KPIs’ (Creese & Burby, 2005) the Web Analytics Association (WAA) defines a KPI in the context of web analytics as:

KPI (Key Performance Indicator): while a KPI can be either a count or a ratio, it is frequently a ratio. While basic counts and ratios can be used by all Web site types, a KPI is infused with business strategy – hence the term ‘Key’ – and therefore the set of appropriate KPIs typically differs between site and process types.

Another thing to note is that the term KPI and metric are often used interchangeably. This is misleading, because although a KPI is always a metric, a metric is not necessarily a KPI. So how do you tell the difference?

From a digital marketing perspective, choosing the right KPIs is crucial to monitoring your marketing’s performance effectively, and allowing you to make informed decisions for continuous improvement. But with a bewildering array of different metrics to choose from, it is also notoriously difficult to pin down exactly what represents a KPI for your site.

If you find yourself struggling with this, it is an area where a session or two with a professional web analytics consultant could be money well spent. Don’t let the consultant take over – you know your own business better than they ever will; rather, leverage their expertise with web metrics to help you define your own KPIs. The important thing is that you end up with a manageable suite of KPIs (usually numbering in the single figures) that together encapsulate the performance of your website.

Some generic web-based KPIs you may find useful:

1. Conversion rate
This is the proportion of visitors to your site who go on to perform a predefined action – such as complete a purchase, subscribe to your online newsletter, register on the forum, fill in an enquiry form or any other conversion factor you have defined. Naturally, the higher your conversion rate, the more of your visitors are carrying out the actions you want them to perform on the site, and the better your site’s performance (to get an idea of some average conversion rates across a variety of online business categories see http://index.fireclick.com).

2. Page views
Simple and straightforward, this is the number of pages viewed by your visitors over a given period, providing you filter out robot and spider traffic and manage measurement of viewability of course!

3. Absolute unique visitors
The number of individuals who visited your site over a given period (as opposed to visits, where each returning visitor is counted again).

4. New v returning visitors
The proportion of your visitors who have been to your site before, assuming the analytics package can recognize them through reconciliation with other data (for example they accept and haven’t deleted cookies).

5. Bounce rate
The bounce rate is the number of people who arrive on your site, and then leave again having only looked at that single landing page. This can be an important metric, potentially highlighting that your traffic perhaps isn’t targeted enough (your keyword choices might be too generic) or your landing page isn’t delivering what visitors expect when they arrive. Bear in mind, though, that some sites will have a naturally high bounce rate (think of a dictionary site, for example: a visitor arrives at the definition page for the word they were searching for, reads the definition and leaves).

6. Abandonment rate
Abandonment rate comes in a variety of flavours – it basically highlights the proportion of your visitors who start down a predefi ned conversion funnel (a series of pages leading to a target action or conversion), but bail out before committing to the desired action. The classic example is visitors dumping an e-commerce shopping cart before checking out, or abandoning the checkout process.

7. Cost per conversion (CPC)
This is basically a calculation of the total cost of advertising (or of a particular advertising campaign where you have tagged the ads so that your analytics software can differentiate resulting traffic) divided by the total number of conversions generated as a result.

There are plenty more. A look at the dashboard or overview page of your web analytics package of choice will offer plenty more, and you’ll find literally hundreds of suggested KPIs online. In the end, picking the metrics that are relevant as KPIs for your website is down to you.

More pertinent data, in addition to monthly unique visitors, would be:

8. Daily unique visitors
An indicator of the frequency of visits and by extension the value placed on that property by the user; three quarters of Facebook users visit daily, one half of LinkedIn users visit monthly – do the maths.

9. Average time spent per daily unique within five user quintiles from most to least time spent
An indicator of the depth of the relationship between property and user.

10. A frequency distribution with accompanying geodemographic and device data by user quintile from the heaviest users to the lightest users
indicating the characteristics of the most and least committed users, such as the relationship between the relatively small cohort of active Tweeters as opposed to the larger cohort of passive followers and the relationship between YouTube devotees and those who view occasionally.

11. Volume of content shared to Facebook, Twitter, LinkedIn and YouTube per unique visitor
Showing the likelihood of that property being a source of influence. The majority of the 10 most shared sources on Facebook and Twitter are news organizations with their roots in television or newspapers. Two of the other three are Buzzfeed and the Huffington Post. What might this imply? At another level, two recipe sites with a vastly different sharing profile may indicate a difference in value to advertisers.

12. The application of unified IDs
showing consumption of content by device type and the nature of cross-device contiguous consumption. These five data points, if made available as standard measures, will paint a far more textured view of the web’s leading properties than exists today, and infuse audience data with real meaning.

Other than confirming that there are many very large sources of internet audiences, this information is uniquely useless because it tells us ridiculously little about real user engagement with the properties in question other than the raw reach of any one site; if we buy a page in Vogue we buy the whole reach, when we buy Yahoo, for the most part we don’t. Of course we can de-duplicate audiences in the pursuit of optimizing reach and frequency, and apply behavioural and other data but it would be helpful to know more.

Even in this data-driven age, buyers of media, creators of advertising and owners of brands have an interest in the composition and characteristics of the environments in which their advertising and brands appear. Knowing why someone does something, and how often, is every bit as interesting as knowing how many do it. This is particularly relevant when the pursuit of long-term marketing effect and brand health are priorities supported by the need to tell stories rather than a simple focus on immediate actions.

The most likely beneficiaries of these data sets are publishers, platforms and aggregators that have deep and frequent purpose-driven interactions with their audiences. That might imply that the data could hold significant advantages for the creators of original content who often don’t top the unique user charts, as well as high-utility destinations such as Google. Frequently a perceived lack of scale disguises the value inherent in strong relationships and the influence of those relationships on both the formation of opinions, decision making and the creation and transmission of influence.

Ultimately, efficiency and effectiveness in advertising lies in the content of the message, its context, its relevance to the recipient and the price and timing of its delivery. If it is true that context and relevance are elevated by user engagement with adjacent content, the data sets proposed above are likely to be contributors to success, or at least a valuable price modifier to available inventory. If this is not true, the data will tell us soon enough.