Check out this interview given by Peter Mueller, Chief Technology Officer at ATS, to Dan Baker of Black Swan. Mueller gives a refreshing, honest, insightful and succinct analysis of how statistics, data, lateral thinking, commodity computing and the internet are combining to empower assurance practitioners in ways we used to dream about, but struggled to achieve in practice. It is well worth a read.
Moneyball, Hadoop, Cloud, and Assurance
Check out this interview given by Peter Mueller, Chief Technology Officer at ATS, to Dan Baker of Black Swan. Mueller gives a refreshing, honest, insightful and succinct analysis of how statistics, data, lateral thinking, commodity computing and the internet are combining to empower assurance practitioners in ways we used to dream about, but struggled to achieve in practice. It is well worth a read.
2 COMMENTS
Comments are closed.
The Commsrisk Global Fraud Dashboard
Our Global Fraud Dashboard uses AI-powered search to collate, update and visualize data about scams and other network abuses from around the world. New charts are added each month. See it here.
Very interesting interview. I’m surprised to know that dropped calls are not reliable predictor of churn. Literature I have gone through in the past had mentioned that dropped calls is an important attribute for churn but this interview provides a different insight. Maybe the count of dropped calls are negligible today.
Eric, Daniel,
Thanks for reading… I was as surprised as anyone to find some of these non-intuitive findings. And, we started off reading some of the same literature as Daniel mentions.
I think the (unexpected) results on dropped calls and churn boil down to this:
1) The total number of dropped calls has been dropping over the last decade, as networks have improved.
2) The numbers have leveled off to a level where we’re comfortable with them or, at least, we don’t think any competing carrier has anything better to offer.
(Of course, there were outliers. If calls were dropping 90% of the time, then the end user probably had a bad phone… and it behooved the carrier to replace it, fast, or risk losing a customer, but that’s more about the phone than the network, clearly.)
Set against the debates regarding the ILECs and the ‘slow death of the landline’, I guess you could look at #2 as sort of a sad indictment of what we’ve come to expect from wireless: We’re willing to forgo the guaranteed quality and reliability of landline for the mobility (and smartphone-ness) of mobile.
While we’re at it, I should mention that some of the other literature on churn seems flawed, too. Much ink has been spilled on social-network-effects of churn, and it’s been taken as truth that when your friends churn, you churn too. We found that effect to be a lot subtler than the literature suggests. Maybe that should be the subject of another article some time, Mr. Baker!
– Peter Mueller, CTO, ATS