Let Sleeping Dogs Lie

David Murphy

Kieran Kilmartin, Group Marketing Manager at Portrait Software, argues that analytics and careful profiling and segmenting of the customer database can help operators drive down churn and increase ROI, by targeting the right people, and ignoring the wrong ones

Portrait Kieran With a uncertain economy affecting customer buying habits, and increased competition to boot, mobile operators are being forced to take a new look at their customer retention strategies. Companies are striving to trim costs wherever possible, and paying heightened attention to retaining their most valuable customers.
It was the father of the department store, John Wanamaker, who famously said: I know half of my advertising is wasted, I just dont know which half. Nearly every marketer knows this quote, but Wanamaker would turn in his grave if he knew the truth - its worse, much worse.

Red flag
Not only is half the marketing wasted, there is a portion of customers that defect because of marketing a promotional campaign can actually trigger customer churn. For these customers, a renewal reminder is the red flag to leave and a special retention offer opens the defection door. With retention-based direct marketing increasing by 34% globally (while overall direct marketing spending declines by 9%) there is a strong possibility these increasing efforts are counter-productive, eating away at positive return on marketing investment.
The good news is that today, we can predict exactly which half of direct marketing spend  is wasted, and most importantly, we can also predict which segments are likely to defect because of a campaign. It all starts with solid data and accurate measurement of the impact of a campaign using control groups - we call it uplift modeling. Add a focus on relevance and delivering customer value, and the results can be astounding.
By simultaneously modeling customer behaviour from treated and untreated segments (control groups) Portrait Uplift Optimizer is able to segment potential campaign respondents into four segments as follows:
Segment 1: Persuadables. The people who respond to a campaign in just the way you hope. They buy (or renew), but would not have done so, had they not received your marketing campaign.
Segment 2 & 3: Sure Things and Lost Causes. If you do not target Sure Things, they will buy (or renew) anyway. Lost Causes wont buy (or renew) regardless of anything you might do. Including either segment in a campaign simply wastes budget.
Segment 4: Sleeping Dogs. People for whom the campaign triggers a negative response, because of annoyance, being reminded of alternatives or simply being given a chance to overcome inertia. Including them in your campaign not only wastes budget but negatively effects response and revenue.
Lets now consider a couple of real-world examples.


Following a retention mailing that increased churn from 9% to 10%, Telenor, the seventh largest mobile provider in the world, used uplift models to turn things around and reduce churn by 1.2%, gaining $8.8 million (5.5 million) per year.
The uplift model found a most savable sub-segment - just 30% of those previously targeted - who reacted favourably to the campaign and reduced churn across the entire segment from 9 to 7.%.
The combination of increased retention rates and lower marketing campaign costs translated into an eleven-fold increase in uplift campaign ROI, compared to previous programs

T-Mobile, Austria
Following the acquisition of tele.ring, T-Mobile Austria increased its customer base by 50%. In a saturated market, with voice price erosion, customer retention was paramount.
With the use of Portrait Customer Analytics, immediate analytics capabilities were available, including simple cross-tabs, quick predictive models and geographical analysis, allowing faster business decisions and more streamlined operations. 
In particular, a number of diverse predictive modeling applications could be used to determine retention optimization for contract subscribers, to help prevent inactivity for prepay subscribers, and to optimize the roll out of the network.
The solution also enabled the T-Mobile campaign management team to execute complex selections and therefore accurately target the right customers, with the right offer, at the right time and therefore become more effective. As a result, T-Mobile Austria has increased the effectiveness of its retention activities, achieving a 20% reduction in customer churn

Stressed economy
In a stressed economy, companies must protect, retain and grow customer bases, while spending less money. Its a concept that might seem like an impossibility, but analytics and, in particular uplift modeling, are having a profound impact on the way marketers develop campaigns and on the budgets of marketing departments. But we still need to remember to let those sleeping dogs lie!