A simple yet effective manner in which to use analytics to prevent competitors from poaching customers in the telecommunications sector, campaign sensitivity analysis is sure to become a common practice in the field of business intelligence / data mining over the coming years…
You can download PDF version of this whitepaper here.
The number of short-term campaigns run by telecoms has proliferated in recent years. With enhancements in campaign management tools as well as in decision-making processes, companies are now able to launch campaigns at the drop of a hat. Combined with advanced business intelligence utilization, these ATL & BTL campaigns are more customized than ever, targeted towards sensitive & important customer segments that often jump at the promotional offer.
This is a particular issue for mobile operators operating in markets where prepaid dual-SIM usage rate is high – all the subscriber has to do to avail from the offer is pull one SIM card out of their mobile phone and replace it with the competitors. Such subscribers essentially churn for the short-term when they do so, with most coming back when the promotional offer is over.
This short-term churn can have a significant effect on the bottom line, however. A recent analysis around international calling campaigns found approximately 1/4th of a mobile operator’s clients switched their international calling to the operator’s competitor when the competitor launched a campaign with better rates. Further, around 10% of those subscribers did not come back once the campaign was over.
So what’s a telecom to do to prevent competitors from poaching its customers? The answer, we believe, is conduct campaign sensitivity analysis then operationalize around the findings.
The conducting of campaign sensitivity analysis will provide a wealth of information to operators who undertake such an effort. Once conducted, the analysis will show which of an operator’s current customers are particularly sensitive to campaigns launched by its competitors (essentially those that are particularly price sensitive), and in particular, to which types of campaigns. By identifying and properly tagging such customers as campaign sensitive, an operator that acts swiftly will be able to prevent a significant amount of the short-term churn that competitor campaigns cause.
To benefit from such an analysis, the following five steps need to be undertaken:
1. Past Campaign Analysis – The first step is to go back historically (one year should suffice) and identify all of the campaigns that competitor operators launched. The types of campaigns cataloged should be any that is relatively short-term in nature (1 week to 3 months, for example), and offers some type of discount / incentive to the marketplace (that is or is not better than your offerings – examples include discounted local minutes, international minutes, SMS, data, roaming, etc.). The objective is to catalog all competitor campaigns that may have impacted your customer base – the analysis will be conducted around these offers.
In regards to each competitor campaign, when cataloging, the following details need to be obtained:
- Competitor Name
- Offer Incentive (i.e. 30% discount on international minutes to a given country)
- Offer Incentive Price (i.e. 25 cents per minute)
- Offer Product Type (i.e. prepaid)
- Offer Stipulations / Rules (i.e. discount applicable on minutes after the first two)
- Offer Date Start Date
- Offer Date End Rate
2. Past Self Analysis – Once the competitor campaign catalog has been developed, a self-analysis will need to be conducted, around self offerings in the market at the time competitors conducted their campaigns. This will require an operator to examine each historical competitor campaign, identify the attributes of the offer (i.e. product type, price, etc.), and determine what itself was offering its own subscribers at that time. So, as an example, if the competitor campaign was offering all-direction local calls after 8pm at 10 cents a minute for postpaid customers, then the operator would need to note what the rate they were offering at the same time was to the same customer type, during those same hours.
3. Macro-Level Usage / Behavior Analysis – The next step is to dig in to the data, to analyze the behavior of existing subscribers before, during, and after the period competitors ran their campaigns in the past. The objective is to identify which subscribers changed their behaviors when competitors ran specific campaigns, then, to tag them as campaign sensitive based on type of campaign. The analysis should seek to identify two general facts:
- The overall financial and churn-related impact the competitor campaign had on the operator.
- The overall financial and churn-related impact the competitor campaign had on any single given subscriber.
By analyzing the change in behavior of the existing subscriber base triggered by the competitor campaign, the above two facts can be obtained.
In regards to a macro-level usage / behavior change, an example would best illustrate the analysis that is needed here, around a scenario where a competitor reduced its international calling rates to all European countries to 1 USD per minute for prepaid customers, for a period of one month…
What analyzing usage changes shows in this case, for example, is that the campaign had a 1.5 million USD short-term impact on the bottom line. This did not end with the finalization of the campaign, however, as revenues around the specific behavior dropped by 400K USD per month afterwards.
In terms of the impact it had on the subscriber base, the analysis shows the number of unique subscribers calling Europe dipped by over 50,000, with around 30,000 appearing to stay away permanently (likely staying at the competitor).
A further specific analysis on the impact of the campaign on churn shows 23,193 unique customers appear to have taken away at least this portion of their ARPU to the competitor.
There are numerous additional analyses that can be conducted around each campaign and its impact, but the above is enough to allow an operator to understand on a macro-level the impact it has / had / will have.
When conducting this analysis, seasonality should naturally be examined and factored in, if necessary, to isolate and understand the true impact the campaigns have had on the bottom line.
4. Micro-Level Usage / Behavior Analysis – With the above complete, the analysis can shift down to a micro-level, down to a specific subscriber. The goal here is to identify which subscribers change their usage and behaviors to what degree when a competitor conducts a certain type of campaign. For each subscriber, the following would be the output of such an analysis:
What the above shows is an example of a subscriber that is extremely campaign sensitive but loyal – while it appears he has jumped over to the competitor during the campaign (and resulted in a 20 USD one month ARPU dip), he has come back after the campaign and resumed his normal behavior.
Essentially each subscriber needs to be tagged (segmented and tagged) based on the behavior they have exhibited around each type of campaign, taking into consideration their behavior and change in usage, as well as churn pattern. Once all competitor campaigns are cataloged and each subscriber behavior analyzed around each campaign period, then a subscriber specific profile around campaign sensitivity can be created. A simplified example of what this could look like:
What the above shows is basically the impact on the bottom line any given type of campaign by any competitor has on any given customer – it directly indicates which types of campaigns have an impact on which customer – with this step complete, each customer is now “tagged.”
5. Strategies Design – With the above complete, the next step is to design strategies around the findings. At the core of the strategy is determining at which price point to drop one’s own offerings to, for which customer, in which type of campaign, when which competitor launches it. If, for example, a subscriber has been tagged as campaign sensitive at a certain price point around international calling, then the internal price point for the same service may need to be brought to the level of the competitors – in which case the price sensitive subscriber will have no reason for switching to the competitor.
An example – a customer is particularly campaign sensitive around international calls – his average ARPU has dropped from 50 USD to 10 USD during such campaigns in the past (a net loss of 40 ARPU). The competitor launches another similar campaign (rates 10% lower than what is currently being offered by the operator) – the operator, acting swiftly, changes it pricing around international calls for to match the competitors and communicates this immediately (via SMS or phone call) to the subscriber – rather than losing 40 ARPU, the operator only loses 4 ARPU (from the effect of the rate discount).
When the strategies are being designed, profitability by customer type, market regulations, channels to use, etc., all need to be considered and appropriately selected. Ultimately, specific strategies and tactics than can be utilized and acted on very swiftly need to be designed and in place to get the full benefits from such a practice. Naturally, pilots should be conducted when operationalizing, to ensure the right mix (offer, channel, message, etc.) is at play to maximize the retention of subscribers.
We believe this unique approach is one that can go a long way towards offsetting the effect competitor campaigns have on an operator’s bottom line. While it’s not the easiest practice to get up and running, once done so, the benefits can be tremendous. To learn more about campaign sensitivity analyis, please contact us at email@example.com.