September 8th, 2008
Building A Synchronized Set of Prepaid Offers
This is the third and final post in a series of three where we revealed how we went about impacting prepaid behavior in our work with one of the largest prepaid operators in the world.
The following are the main behavior indicators we take into account when building a synchronized set of offers to different segments. To see how we impacted each please see the before and after graphs in How Do We Impact Prepaid Behavior?
> The time period that goes by between top ups – the shorter the period is means the subscriber is more active.
> The time that goes by from reaching a zero balance until the next top up is even more important, because each day that goes by without talking, means the subscriber could be churning. This resonating silence is referred to as a ‘dormant subscriber’.
> The third factor is the date of the last call, or days from the last call. This is extremely important because since there is no contract, there is no proactive activity involved in case the subscriber wants to churn like calling the operator to terminate service. That is the main reason why so many prepaid subscribers become dormant, and why this phenomenon is often referred to as ‘passive churn’.
The above factors help us define and ultimately achieve our three main business goals:
A - The need to reduce churn or in other words retain customers.
B - The need to increase ARPU.
C- The need to increase top up amounts which is the key to keeping customers active at all times.
How do we go about achieving this?
- By encouraging subscribers to recharge more frequently – For example: In this top up sample, the average recharge amount is between €10-15 per recharge so we offered, “if you recharge more than €20 you get 10% immediate bonus. Or we offer a recurring benefit such as, “recharge more than €20 and get 10% bonus on all recharges during the next month. All this encourages subscribers to recharge more and for higher amounts, and gives the subscribers an incentive to remain active over time.
- By making sure they don’t reach zero balance – For example: We implement a proactive approach using event/trigger based activities through a series of special marketing offers to prevent subscribers form reaching a zero balance. So the moment the subscriber reaches a low balance threshold of €3 for example, he receives an SMS: “recharge now and get an immediate bonus”. This is also accompanied by reminders, where if they don’t recharge they receive the offer again after 3 days.
- By identifying high-churn risk subscribers and proactively approaching them with the right retention offer – For example: We don’t want to reach a point where subscribers have not talked for several consecutive weeks and become dormant. There are two options for combating this:
Using A General Churn Indicator - We identified that over 80% of subscribers that don’t talk for 14 days will become dormant users. Therefore the moment 10 days has gone by from the last call we proactively suggest a conditional benefit such as, “recharge or talk and receive a benefit” –giving the subscriber the motivation to be more active. Subscribers that responded were not offered any follow up offer but their usage level was monitored , and were eventually communicated with a follow up offer after one week.
Using A Personal Churn Indicator – Here we use the same offer but the indicator is different. Since we have learnt the usage patterns of many subscribers groups, we can build a personal profile for each subscriber. We know for example that the maximum period a certain subscriber won’t talk at any given time is 3 days, so if we see he hasn’t spoken for a week that’s a red flag. But for other subscribers’ segment (lower users) this indicator is not relevant and will not be flagged as a churn indicator. This churn risk segmentation method is also known as the RFM model: Recency, Frequency, Monetary Value.
- By generating ongoing interaction with the subscriber across his life stage, or leading them up a “value path” where we track the user’s behavior over time and offer relevant and more targeted offers at each point of interaction - For example: If we have a subscriber with an average ARPU that recharges a low to average amount, our goal is to a) increase the average top up amount and b) increase ARPU. So how do we engage subscribers in ongoing interaction? We discovered a direct correlation between the balance of the subscriber and their activity, namely the larger their account balance is - the faster they spend it.. For this type of segment we created a top up benefit offer where whoever recharges more receives a usage benefit.
The bottom line is that we experienced great success in achieving our main business goals:> an annual churn rate reduction from 30% to 27% (which is a 3% difference of 30% and therefore 10% shift overall)
> ARPU grew by 20% during the Special Offers period
> The average top up amount grew by 15%
Other directions for bringing about an improvement include:
1) Adding another dimension of segmentation: the value of the customer. A customer that talks for €100 is worth retaining much more than one that talks for €20 a month.
2) Measuring the responsiveness of subscribers to be more efficient at specific cycles of the campaign, for example a customer that responds to offers really well. Or we differentiate customers by their response to a certain type of offer, for example, responds well to top up bonuses and less to usage benefits. Or another segmentation can be done by communication channel, for example responds more to SMS than voice messages.
Even though our tactics mainly focus on dividing customers into segments and dealing with each segment in a certain way, the cube diagram below attempts to demonstrates the value of analyzing different usage parameters and building a comprehensive plan to address these sub segments. In a nutshell, by crossing three behavioral dimensions you can map your customer base into micro segments, by profiling or ranking the churn propensity of each micro segment and matching the most effective offer for achieving the appropriate business goal. For example: Mapping your prepaid customer base through the following three top up dimensions: frequency, average amount per top up and days in zero balance.
Having this ‘big picture’ perspective at all times allows us to be very effective in our marketing activities and achieve great business results for our customers.
We sincerely hope you gained valuable insight from this three part series on how to go about impacting prepaid behavior.
Comments
2 Responses to “Building A Synchronized Set of Prepaid Offers”
Leave a Reply
-
Subscribe
Dear Nir, I am just able to see your interesting article and I’d love to ask you some a few questions.
In your article, you said Using A General Churn Indicator - We identified that over 80% of subscribers that don’t talk for 14 days will become dormant users.
Could you please explain how you identifed this 14 days ?
Hi Meryem,
In order to track customers who became dormant, we monitor on a daily basis every cutomer’s usage (voice calls, SMS, content services) and balance recharges.
Then, with analytic tool we have embedded in Pontis platform, we count how many consecutive days with no activity each customer has and proactively contact customers after 14 “silent” day with a retention offer.
Working by this method allow to run follow-up campaigns as well,
for example: customer who did not repond to the first retention offer within 1 week, will be address with another, more attractive retention offer.