Khoros Atlas Logo

Highlighted
Advisor
Advisor

measuring calls deflection vs bounce rate

Jump to solution

We are a new support community (telco). As you can imagine our main pillar of business case assumption is a call deflection.

Recently we have started adwords promotion and the level of bounce rate has increased to ~60%

 

I wonder how to deal with bounce rate in formula that we use (below). Should we simply exclude the bounce rate traffic?

 (Deflected calls = Unique visitors (LSI) * %seeking support (survey) * %resolution rate (survey)*%deflection rate (survey)

 

It would be nice to see your point of view

 

 

 

 

 

6 Replies 6

Re: measuring calls deflection vs bounce rate

Jump to solution

Bounce Rate is pretty difficult to suss out in this scenario. On one hand, you could look at it as a user hitting the page, deciding it isn't what they want, and leaving. On the other hand, you could look at it as a user landing on a content page, getting exactly what they need, and clicking away to their next task. I'm not sure that making the assumption that either of those scenarios accounts for 100% of bounced visits is realistic.

 

This is where you'll likely have to make a judgement call and be ready to explain why you think that's the case when someone asks about how you calculate your deflection metrics.

Reply
Loading...
CCO Emeritus

Re: measuring calls deflection vs bounce rate

Jump to solution

I don't recommend removing bounces from your ROI calculation. In fact, I can't think of one community that does this. (If you do and you are reading this, speak up!)

 

@BrianOblinger hints at why this is never done -- it's because single-page visits (aka bounces) are often successful visits in a support community. If you're a typical public-facing community, and even with your AdWords traffic, you're probably generating more than half your visits from search, which means they are landing on content pages with answers. They may not need to go further.

 

In the data I've seen, bounce rates for content sites of all kinds tend to range from 40-60%. The conventional wisdom says that that forum-based sites skew to the high end of that range, but most Lithium communities fall closer to the middle. 

 

It's not surprising that visitors from a campaign are more likely to bounce. Just because of the nature of campaigns, the resulting visits tend to be less goal-oriented. But it doesn't seem to be driving your bounce rates outside a normal range. 

 

Bottom line, most companies count them, and you should too. 

 

 

Reply
Loading...
Advisor
Advisor

Re: measuring calls deflection vs bounce rate

Jump to solution

@JoeC, @BrianOblinger thanks ! Great advices, that helps me a lot. 

 

Reply
Loading...
Valued Contributor
Valued Contributor

Re: measuring calls deflection vs bounce rate

Jump to solution

This makes sense, thanks @JoeC - given what you've said and the formula suggested by @drolcak, what is the formula you'd recommend?

Pete @ Bankwest
http://forum.bankwest.com.au/
Reply
Loading...
CCO Emeritus

Re: measuring calls deflection vs bounce rate

Jump to solution

@drolcak is exactly right re the formula, Pete: Deflected calls = Unique visitors (LSI) * %seeking support (survey) * %resolution rate (survey)*%deflection rate (survey).  If you are using our Value Analytics feature, it gives you the last three data points for this formula. 

Reply
Loading...
CCO Emeritus

Re: measuring calls deflection vs bounce rate

Jump to solution
Just one additional nuance here - I think our Business Value Engineering team prefers Visits here instead of Unique Visitors. It's a slightly less conservative metric (since each visitor might visit more than once in the period), but it's closer to how support organizations typically measure (not in terms of people, but in terms of calls).
Reply
Loading...