Preference Choice Publication

On Target

David Murphy

DM: So Alon, could you give us a brief introduction to Pontis please?

AW: Sure. We provide a real-time contextual marketing and sales platform for communication service providers, including mobile network operators. We were recently recognised in Gartner’s Cool Vendors report, based on its evaluation of our Marketing Delivery Platform, which allows mobile, TV and fixed line operators to target customers with contextual marketing offers.

We have 120 staff and we have recently changed our business model to more of a service-based model, where we provide the technology, as we always have done, but also ongoing marketing support.

DM: And what does the technology do, exactly?

AW: We provide operators with a new way to manage their prepaid customer base and help them to retain customers and increase their lifetime value.

DM: How do you do this?

AW: Through targeted offers, delivered at the right time, in the right context. So for a prepay customer, it could be sending them a top-up reminder when they get down to €2 credit left, and then if they don’t respond to that, making them a better offer, such as a credit for their next top up. Our system will also raise a big red flag if someone hasn’t spoken on their phone for two days, if that is unusual for that customer.

We are not inventing anything new here. People are used to this sort of stuff with Amazon and Google. We just try to be as accurate as we can be in terms of the segment, the context, the timing, and the offer itself. And if the customer experience is better, it will have a positive financial impact.

DM: And how many operators are you live with?

AW: 22, including four Vodafone opcos, and other Tier 1 operators such as Vimpelcom in Russia and Claro in S. America. We are constantly enriching our knowledge and insights regarding segmentation and effective marketing.  For example, we have learned that if you segment customers only by the average amount spent by value, your targeting is not accurate enough. You have to also address behavioural attributes, such as abnormal silence periods or usage decline, in order to optimize the response and perceived value to the customer.

DM: So how does the technology work exactly? 

AW: The key to it is tracking the customer state and change in behavioural patterns at all times. For example, operators typically measure churn for pre-paid customers after a few months. But a prepay customer might actually churn the week after they signed up, so you need to track consecutive silent days against control segments to see what’s really happening.

We can segment the database by average spend; average single top-up; at what balance the customer typically recharges his account; to what offers he responds. Sub-segments will respond in different ways to the same offer, so different segments need different types of offer. For example, an offer to top up €30 and get 20 per cent off of all your SMS may be good for someone who regularly tops up €20-25, but it’s no good for someone who only ever tops up €10 at a time.

Everything is automated, and that includes the application of the rules in real time to each customer based on his eligibility and state, the messaging and the discount,

DM: And how do you sell the platform?

AW: Apart from an initial set-up fee, it’s typically payment by results; there’s no software licence fee. We are providing a managed service with full IT and marketing support, and we get paid based on the increases in revenues we deliver (vs. a control group)

DM: And what sort of increases do you deliver?

AW: Typically, a 3-7 per cent ARPU increase for voice services; and 8 - 12 per cent reduction in churn: and a 6-9 per cent increase in lifetime value.

DM: And do you only do this for prepaid customers?


AW: We started with prepaid, because if you look at the numbers, Vodafone alone has 240m prepaid customers. In the US, it’s growing, and in the Far East, almost everyone is prepaid. But we have just gone live with our first post-paid implementation, and it’s a different ball game. You need deep domain expertise to succeed; you need to understand the different types of parameters and segmentations between post-paid and prepaid.

With post-paid, it’s all about trying to up-sell and cross-sell the customer to the operator’s other services. So if you have a monthly pack of 200 text messages, if I want you to stay loyal, I might offer you 100 more, but I should only do this when you’ve used up 180 of your 200 text bundle, and it’s only the middle of the month.

There are other opportunities too, if we can connect to the data warehouse and the business intelligence system, so we know what plan you have today, what services you have been offered in the past, what you’ve shown an interest in and what you haven’t, we can use this as the basis for new offers that are relevant to each user.

DM: I do like the idea of the personalisation and the timing of the offer. The only thing I don’t quite understand is why you don’t have every mobile operator in the world beating a path to your door. The way you describe this, I can’t see why anyone would not be interested, especially given the payment-by-results aspect of it. So with the greatest respect, why don’t you have more deployments?

AW: The response we are getting now from operators is completely different to three years ago. At that point, this all looked a bit too intense for them. We did not have enough proof points going back long enough, and telcos are tough to deal with; it’s hard work, and it takes time. But there’s a greater appreciation among operators now that with the prepaid base going up, people using multiple SIMs, they need to find some new value from somewhere, because prepaid customers can go from active to dormant to lost just like that. On the plus side, telcos do have the highest potential for contextual and personalised information if you can get the segmentation, the context and the offer right.