Customer Engagement Is Measurable

In a recent post, Avinash claims that engagement is not a metric, and writes:

Engagement is not a metric that anyone understands and even when used it rarely drives the action/improvement on the website….It is nearly impossible to define engagement in a standard way that can be applied across the board.”

My take: I think Avinash is taking an uncharacteristically narrow view of the term engagement, and the ability to measure it.

The biggest issue with the way the term engagement is used in the marketing community is its narrow connection to websites and the online channel. When marketers think of “customer engagement”, they should be thinking about how engaged the customer is with the company, product, or brand. The level of involvement with the website — or with a particular ad (online or offline) — is just one dimension of a customer’s engagement.

Customer engagement encompasses a number of dimensions:

  1. Product involvement. A customer who doesn’t care about the product, is likely to be less committed or emotionally attached to the firm providing the product.
  2. Frequency of purchase. A customer who purchases more frequently may be more engaged than other customers.
  3. Frequency of service interactions. Branding experts like to say that repeated, positive interactions lead to brand affinity. And they’re right to a certain extent, but….
  4. Types of interactions. …not all types of interactions are created equally. Checking account balances is a very different type of interaction than a request to help choose between product or service options.
  5. Online behavior. Time spent on a site might be very important. But, like types of interactions, not all web pages are created equally.
  6. Referral behavior/intention. Customer who are likely to refer a firm to friends/family might be more engaged — a customer who actually does refer the firm, even more engaged.
  7. Velocity. The rate of change in the indicators listed above may be a signal of engagement.

Avinash is on the right track, however when he says that it is nearly impossible to define engagement in a standard way. I would suggest, though, that a standard definition is feasible — but that measuring it in a universally standard way is what’s impossible.

And that’s good.

Who said we need a standard way of measuring engagement? This insistence on a standard definition and approach is to measurement is silly. You don’t hear anyone getting all worked up about the fact that market share can be calculated any number of ways, and that the denominator in that metric is hardly consistent or easily measured.

Measuring engagement needs to be done in the context of a firm’s strategy and it’s own theory of the customer — that is, what behaviors the firm believes constitutes an engaged customer.

Measured correctly, engagement meets one of Avinash’s golden rules — to me instantly useful. Using market research data, I measured customer engagement with their banks using the attributes described above.

I then segmented the respondents into four categories, based on their level of engagement, and the breadth of their relationship with their banks (based on the number of products owned). The result: A metric that is immediately useful in helping marketers address some strategic questions about their marketing and customer strategy.


Marketers need to stop getting their knickers in a knot trying to boil engagement down to a single metric that relates to a web site or the online channel. It’s a descriptor of a customer’s attitudes, not a channel’s performance.

A metric, when used appropriately, can help execs make decisions and manage. But considering the way engagement is being defined and measured today, it’s no wonder Avinash has come to the conclusions that he has.

Technorati Tags: , ,


15 thoughts on “Customer Engagement Is Measurable

  1. Yessir, the issue goes way beyond web analytics, though I would argue some of the most precise and interesting data on engagement is available through web analytics.

    If folks would only look at the right data. I’m with you, engagement = attitudes, not actions. To go further, I’d say in a lot of cases you can predict those actions and so actually rank customers by their level of engagement, which I believe is where you are headed with the above chart.

    It’s fine to navel-gaze about all the complexity you create around this idea (especially true in web analytics), but in the end, if the analysis does not lead to action (as suggested by your chart above) what good is it?

  2. Thanks for commenting, Jim. You know, of course, that we’ve discussed this before, but Avinash’s post gave me the license to post some of these thoughts in a different context.

    I couldn’t agree more that Web analytics could be a great source of data to drive engagement measurement. But until (or unless) a large percentage of customers are conducting the vast majority of transactions and interactions online, then taking the Web-driven data out of the context of the cross-channel may lead to warped views on who’s engaged and who isn’t.

    I like to joke that the customer who checks his bank account balance online 17 times a week isn’t necessarily an engaged customer — he’s more a likely a neurotic customer.

  3. Or they’re pushing their budget to the limit…is that the same?

    But this example really highlights the need for a two-dimensional view of the engagement idea, using the same “matrix” approach we have been tossing around.

    In this case, there’s one thing we know about someone who checks their bank account 17 times a week – they are very highly likely to check it a lot of times next week. Then the question becomes, of what value is this information?

    Is there any value created for the company because of this activity? Probably not. Can we create any value for the customer because of this activity? Perhaps; maybe we could text the bank balance every day to their cell phone; would they like that?

    The customer-centric question to ask is not “how much activity”, it is:

    1. How important will this activity be to the customer in the future? There’s not a lot you can do “today” for the customer who was checking his balance 17 times a day but stopped 3 weeks ago…

    2. What does this activity tell us about how to better serve the customer, if anything?

  4. Ron: Thanks so much for extending the conversation. Each of the six suggestions you have on top of the post is excellent at measuring various relevant nuances of “engagement”.

    The extension to online – offline was particularly helpful, it poses its own challenges, your example of the 2 x 2 matrix from your banking experience is excellent.

    In the end I think we are more close in our perspectives than apart.

    PS: I always enjoy the JimN RonS witty banter. 🙂

  5. Jim — Actually, there is something of potential value to do w/ the guy checking his balance 17 times a day — ask him if he’d like an alert pushed to his email address or cell phone when his balance hits (or falls to) a certain amount.

  6. Ron,

    Great post! Your response to Kaushik reminds me of a fairly extensive series I wrote in my blog about measuring visitor engagement. Your list above is an excellent online/offline counterpart to the components (or indices) I propose for measuring engagement using strictly online and quantitative data. If you have time, have a look at this post:

    Or, if you’re really bored, you could read all of the posts on the subject, working backwards from here:

    Personally I think that working out strategies for individual organizations to measure engagement using both qualitative and quantitative and a “Web Analytics 2.0” model is some of the most exciting work that the web analytics department gets to do in this day and age. But, like web analytics, sometimes this work is hard so you really have to be up for it.

    Anyway, thanks for highlighting the value of measuring engagement to your audience!

    Eric T. Peterson

  7. Pingback: Tomorrows Brands - Today’s Top Blog Posts on Advertising - Powered by SocialRank

  8. Ron, This is a excellent post. When I read Avinash’s post, I had some mixed feelings and I was finding hard to agree with him. Now, I think I am with you in this post.


  9. Ron,
    Regarding the the guy checking his balance 17 times a day — this can happen with many, but not on a consistent basis. Many a times, this increase in frequency is associated with an event (expecting a deposit, spike in expenditures and cutting it close as Jim said, checking to see if a particular check cleared etc.). I like your alert idea — but I would say we give the customers the option to create/execute/stop these alerts (than us pushing).

    Coming back to the post 🙂 — definitely a great post. I think you are right on the money regarding the dimensions — but coming from a research design background, I couldn’t help but expand your ‘model’ with a couple of moderators like Industry and Size of the customer:-) from my experience these play a role in how much you interact and what type of interactions you are having.What I totally agree with you is that Engagement CANNOT be measured just by looking at their online visit in silo. Anyone who is doing that is not really measuring engagement but like Avinash says, another metric just relabeled as engagement.

    Thanks for the great read and the matrix.

  10. Ned — Thanks for your comments. I couldn’t agree more that an “engagement analysis” must be done in the context of the product category, as well as the company’s individual strategy. Marketing has got to sit down and decide “what does a good customer relationship look like from a behavioral perspective?”

  11. Pingback: More Thoughts On Customer Engagement « Marketing ROI: Whims from Ron Shevlin

  12. It amazes me when marketers try to invent “metrics”. Metrics for marketing should be the same metrics that the CFO uses. Online, especially, marketers should be held accountable to generating high quality clicks, signups and leads that result in sales. It is very possible now for any sized businesses to track exactly what’s working for them: to measure exactly what’s generating leads and sales. With with blogs, email, wikis, ppc, seo, webinars, it’s entirely possible to advertise and market directly to the appropriate audience online, whether it’s a specific industry, location, type of position, etc. Marketers should be tracking which of their activities result in revenue and do more of what works and less of what doesn’t.

  13. Pingback: Windows Live Search Releases Custom 404 Error Page Toolkit | The Golden Hat

  14. Pingback: Customer Engagement RIP « Ron Shevlin’s Marketing Whims

Comments are closed.