Share this content
Formulae for customer engagement measurement: Myth or reality?by
27th Jan 2010
Share this content
There are many issues your company will have to contend with if you want to measure customer engagement. But while sophisticated engagement measurement may be elusive, the truth is out there, says Hugh Gage.
Engagement is a bit like brand awareness - it’s a highly desirable but rather nebulous concept to measure. In the online world, engagement is very popular and the more sophisticated marketers and site proprietors are often keen, for a myriad of different reasons, to know how it can be measured.
In the beginning, engagement on a website was considered roughly in terms of 'page views per visit' and 'average time spent on site' but for obvious reasons these metrics by themselves or even in concert became obsolete - after all if Joe Bloggs comes onto your site and spends a long time on it viewing a great many pages does it mean he’s highly engaged with the content or just completely and frustratingly lost? Having said that, it is clear that when trying to measure customer or visitor engagement online, the starting point must be the data.
At this point it might be worth considering the difference between customer and visitor engagement because they amount to different things from an analytics measurement perspective; visitor engagement largely revolves around the behaviour of a visitor on your website, customer engagement is more complex and revolves around the customer, wherever they may be, and this of course can take in many different touch points from your site, to your CRM strategy, to your social media communications strategy, to your after-sales service. But for the purposes of this article we will think of customer engagement at least in the context of businesses for which the website is a central pillar if not the central pillar.
Now add to that the concept of visitor/customer engagement on two such different sites as an ecommerce site and a news publisher site. Firstly, one might argue whether you could even consider an online news publisher to have a customer - as opposed to a reader - but in order to do that you would need to identify the business objectives of the site, this then brings us back to web analytics 101. Before passing Go and collecting £200 you must always, whatever your business, first identify the business objectives of your. After all, if you managed to find a satisfactory way of measuring either visitor or customer engagement but didn’t actually manage to increase revenue or take any useful action as a result of it then really, what have you achieved? It would be like having a whole load of friends on Facebook that you never actually met in the real world.
Of course, because engagement can be defined in more ways than one it stands to reason that there is no set formula for measuring it. However, in the world of web analytics some have indeed produced formulae for measuring engagement. The irony of this of course is that engagement is surely all about communication, yet to reduce the measurement of that to a formula almost feels like the antitheses of engagement. To be fair, in most cases engagement formulae will be specific to the site and the business, again taking us back to the business objectives.
Thinking again about visitor engagement, where average session length is long and average page views per visit are high, if the site in question were an ecommerce site with a high conversion rate then one might assume that a long visit length and high page views per visit do indeed relate to a higher level of visitor engagement. This would be even more the case if session length and page views per visit dropped off at the same time as overall conversion declined. In other words, to measure visitor engagement you’ll need an anchor metric against which to define success or not. This anchor metric will depend on the business objectives for the site.
So it is possible to produce a formula to measure visitor engagement but it will need to involve a scoring method which will be the result of a number of different factors/metrics combined and these may not just be data-centric. What if shoppers on your ecommerce site have to go through a registration process before making a purchase and what if half the fields in that registration process were voluntary but if filled in would be of genuine benefit to the new customer? People that fill in the bare minimum could be seen to be less engaged, however those that fill in more of the voluntary fields could be said to be more engaged.
Quality and quantity
Now consider this in the context of customer engagement. It would be 'simple' to measure customer engagement if it only applied to activity on a website - but it doesn’t. A customer by definition has already purchased and therefore already been through the purchase process and no doubt formed an opinion based on ease of purchase together with the pre and post sale experience. As a result, a customer will subsequently engage with a brand in other areas, possibly by word of mouth in conversation with others or possibly through social networking or product review sites. These are just two of many areas in which a customer may be considered to re-engage with a brand without necessarily accessing the website and because of that measurement using web analytics data, as with visitor engagement, becomes only part of the wider process.
At this point, assessing engagement becomes more qualitative. Web analytics doesn’t just have to be a quantitative discipline and in fact in order to get a more rounded picture of behaviour it helps to draw from as many sources as possible. And so it is when measuring customer engagement, additional sources such as buzz monitoring, on-site and after-sale customer surveys, after-sale customer support feedback, lifetime purchasing history and so on all become relevant to the assessment of customer engagement. The difficulty really comes in tying them all together and in some cases it is nearly impossible to do. But in other cases the more sophisticated marketers will find ways of joining up the various data points so that, for example, raw quantitative customer engagement data from a web analytics package can be wedded to qualitative survey data so that answers can at least be matched back to on site visitor behaviour.
Finally, as alluded to earlier, sophisticated customer engagement measurement may be nice to have but unless it leads to insight which makes a positive contribution it becomes redundant. One of the main issues facing web analysts is that of presenting insight to key stakeholders within the business who do not normally have much exposure to his kind of measurement. These people are then required to make business decisions which involve budgets so it’s important that they have a clear line of sight in terms of their objectives when assessing customer engagement.
Hugh Gage is an independent web analytics and usability consultant. He works with businesses to help improve online marketing and website performance. His background is in we analytics, online advertising communications and strategy development. He also writes the Web Pro Analytics column in .Net magazine.
Share this content