Why is understanding engagement so hard?
We have more data than ever and you would think defining what good engagement looks like would be easy to understand.
Just because we can get the data easily it doesn’t mean the answer immediately jumps out at us.
There was a time when we had to think about our questions rather than just extracting the data and saying “there you go, see the data shows . . .”
In this post we look at “what is engagement?” and “how do we measure it?”
Here’s an example…
In retail car dealerships, as they used to be, the sales people understood which customers were browsing and which ones were in buying mode.
How? They were able to look at customers from past experience and know which ones were in buying mode vs those who were wasting their time. Customers who had made a purchase previously gave off signals. The sales person would be able to use these signals to know when it was right to help the customer and provide time and effort to sell a car.
In digital something goes missing. The word I want to use to demonstrate this is “engagement”.
People often ask us “Can you create an engagement score or tell us what an engaged customer looks like?”
The answer we give back is always the same.
“What are you trying to achieve?”…“What’s the goal?”…“Where’s the win?”
then and only then can you measure and understand engagement.
“Our definition of engagement is a strong measure of demonstrable behaviour from which we can identify if the visitor has achieved our business goal ($).”
Example goals vary, ideally the ones which directly link to revenue are best, since we can then identify the value of the engagement, but they don’t have to be financial ones.
They can include:
- Adding to cart
- Newsletter sign up
- Competition entry
- Purchase online
- Test drive booking
- Whitepaper download
- Online tool use
- Login to customer area
- Raise a customer service ticket
So why do people still use dwell time and page views as a measure of engagement?
It is nearly as insane as using impressions as a measure of success. Neither gives you a true picture of a positive outcome. These metrics only provide information to build additional information into the outcome but should not be used as the only source of measurement.
Real world scenarios:
Engagement as we highlighted above is a signal not the answer.
It has to be taken in context, in relation to the business and the outcome we want the visitor to take.
High dwell time for an advertiser is good, BUT ONLY if they are reading more than one article. Looking at one article for 40 minutes for instance might suggest someone is out to lunch and not actually engaged at all.
The one click purchase funnel seen in ecommerce (created by Amazon back in September 1999 – which interestingly also owns the “1-Click” trademark.) wants a maximum of 3 pages view. Any more suggests a lack of engagement.
For automotive the ultimate engagement for new car sales might be 7 clicks. Why? That is the number of clicks it takes someone who has reviewed a car to complete the test drive booking process. However this is only for repeat visitors. 1st time visitors need more time and a strong indicator could be the car configurator which needs 15 page clicks to complete and shows a good level of commitment.
How should we measure it?
- Identify the goal you are trying to measure customers/visitors against
- Look at customers who have achieved this goal previously and review their dwell time and page views.
- Break this down further by source and device as this can massively affect things
- Monitor the overall engagement of your audience to understand if they are heading in the right direction
The Crank view
Don’t do stupid things by using inaccurate (vanity metrics) measures of engagement.
Use engagement to measure visitors in relation to the goal you are trying to focus on.
Then and only then can you provide an engagement metric for converted and non converted visitors.
At Crank we prefer to contextualise this by source of traffic and device so our clients know what good looks like and are able to measure if a specific campaign or message is close to delivering the right behaviour beyond a click or non bouncing visitors.
What’s you viewpoint?