Mobile World Congress always prompts plenty of press releases but an announcement made by assurance and analytics vendor MDS deserves some attention. They claim to be able to use stats to predict customer behavior in ways that will lead to significant savings for some telcos. The MDS announcement states that their wholesale cost optimization solution can reduce the costs of digital service providers by…
…using MDS’ predictive analytics and optimisation. It utilises complex statistical modelling to forecast subscriber usage based on behaviour, and furthermore, automatically places each SIM on the most cost-effective wholesale bundle, thereby creating costs savings of 10-15%. This solution is already delivering benefits today for one of the premier international MVNOs.
Saving 15 percent for one telco is not the same as saving 15 percent for all of them, and as a former auditor I prefer to see numbers backed up by evidence. Another potential issue is that customer behavior need not be consistent all over the world, for every telco. However, this is an intriguing boast by MDS. Better statistical analysis should lead to better predictions. The scale of the claim is believable also. A 15 percent saving is not that different to claims of savings that customers have enjoyed by being transferred to retail plans more suitable for them.
One telcos’ wholesale cost is another telcos’ wholesale revenue but which business would ignore the opportunity to cut direct costs by over 10 percent? Success depends on the predictive power of the stats model, which begs a question of whether different vendors might offer competing predictive models, and how their efficiency might be compared in practice. This is definitely a topic worthy of further investigation. Sadly, MDS has not named the ‘premier international MVNO’ that uses their wholesale cost optimization solution. If the claims are true, that MVNO probably does not want competitors to emulate their success.