Can/Should MVNOs offer AI products in their suites?

The answer is definitely yes! And the reasons for this are many:

• The host MNOs world over are in the process of setting up AI datacenters to enhance their existing fleet. Part of this is driven by solid business imperatives and some of that is pure peer pressure
• Unlike traditional Mobile telecommunications products, the MNOs do not understand (yet) how to package and offer the products to their customers. They would be looking for ideas from MVNOs who tend to be more nimble. In this regard, AI products have the ability to change the traditional balance of power between the MVNOs and their host MNO, which hitherto has been heavily tilted towards the MNO.
• A large majority of MVNO are actually owned and operated by highly experienced and skilled techies. They are thus in a better position to quickly understand and react to a rapidly changing market conditions, and
• It gives the MVNOs an opportunity to break away from the traditional cost price market model that severely restricts the ability of the MVNOs to improve their margins outside of radical cost reduction.

In the last 3 years, whenever my colleagues and I have spoken at MVNO conferences, we have emphasised on the need to reform the financial model for the MVNOs. There are really only 2 (both bad) options available to the MVNOs to improve their margins. They can either cut costs or cut prices. The former has physical limitations and the latter only encourages the race to the bottom. This needs to change.

Thus, the only other options available are use of thick MVNOs (expensive) or higher flexibility in product configurations afforded to the MVNOs by their hosts. This is not about to happen either because it most likely cannibalizes their basic products base.

On the other hand AI products are less structured and quite possibly the MVNOs could use a percentage of the value product as a charge rather than fixed prices. Let me explain how this potentially works. So, a host MVNO puts up an AI datacenter and offers GPU as a service to the MVNO. The MVNO could then set up an optimisation package on the back of the GPUaaS where they offer their customers ability to reduce their costs and in turn, get paid a percentage of the savings. The MVNOs will be in this case combining 2 of their well known capabilities:

• Technical capabilities of the founders and management, and
• Close relationship with the market segment to whom they sell.

Moreover, these MVNOs already have existing relationships with the host MNO, hence likely to be onboarded less painfully, and in most instances I believe happily.

Incidentally this business model from MNO perspective has been on display by operators in the global south who have successfully incorporated fintechtech capabilities in their offering. Unlike selling minutes of talk and MBs of data, fintech products are sold as a percentage of the quantum being transacted. To the best of my knowledge this is the only telecommunications product whose incremental revenue does not lead to an increase in cost of delivery. And in one particular case, Kenya’s Safaricom has fintech contributing to 42% of the revenue with all the attendant implications on the margins. As a matter of fact, rumor had it that when Safaricom expanded into neighboring Ethiopia, they were building a telecommunications network on which to run their fintech products.

Maybe there is no real opportunity for MVNOs to barge into the Fintech revenues by the operators (and certainly not in the OECD countries where regulations make it almost impossible), but that success can be replicated in selling AI products. And MVNOs can be part of that conversation.

I’m always willing to listen on how we can help you achieve success in that journey.