Open Gateway: what's it worth + time for an MVP?

I spent some of my time at MWC 2024 talking with people about three of the industry’s big “open” initiatives – Open Gateway, Open Digital Architecture (ODA), and Open Radio Access Network (ORAN).

To start off, here are my thoughts about Open Gateway, starting to explore the question of (and arguments about) 'what is it worth to the industry?'. In summary:

• My back-of-envelope-o-meter suggests it might potentially enable up to 1% growth from API revenue and a further 5% revenue growth from other plays. BUT you should be careful about these numbers (and all the others you will hear). They are based on estimates with high scope for error, both in the total potential base and the assumptions about how they are achieved.
• That may not feel that impressive or confidence inspiring, but it is still worth doing if it can be done with the right level of investment (i.e. selective and smart) and commitment (collective and resolute). Nobody knows what is really possible at this point, it could be better than that, but it will never be known unless the industry collaborates to achieve scale.
• To achieve this, the industry needs to develop its route to developers, and should now focus on a more focused collaborative MVP around an identity/anti fraud API, such as Number Verification or SIM Swap.

I welcome further input on how others see this. Below, I set out the context, the arguments (there are plenty!), the estimates, and why an MVP makes sense.

Context: the Connected Technology industry needs to be more open

These days, I see Telecoms as being part of a much broader industry that I think of as “Connected Technologies”: connectivity, data, automation, AI, etc. I also see it through the lens that the point of all this technology is to help make better use of stakeholders’ resources (money, time, health, carbon, etc. - see the Coordination Age).

All the “Open” initiatives are important because they are all ways to either help the Connected Technologies industry work better or enable other industries. I see them all as “good ideas” and I like and respect all the people I know who are engaged with them. But how are they doing?

Open Gateway: the big push to open up to other industries (and developers)

Open Gateway was the GSMA’s flagship project at MWC 2024, and is broadly another go at standardising telco APIs (you can see the API definitions here). The idea is to allow 3rd parties (e.g. customers, developers) to access telco assets and information easily, so they can use them to build and deliver new offerings.

Example Open Gateway APIs include:

• Number Verification (is this the phone number it claims to be?)
• SIM Swap (when was the last time the SIM was changed - useful in fraud prevention)
• Device Location Verification (is this device where it says it is?)
• And Mobile Quality on Demand (option to boost connectivity when required).

It was heavily backed (47 operators have signed up) and CEO-fronted at MWC 2024, but there are a lot of questions to answer.

• Do customers want it / are they ready yet?
• Does it serve developers’ needs?
• Do developers know how to find / use it?
• Does it do something better than the competing offerings?
• Do telcos have a strategic plan for rolling it out?

I spent some time with the GSMA’s Henry Calvert who runs Open Gateway. He clearly sees that there’s a way to go yet to resolve these questions, plus others about getting telcos to work together to make it all happen.

Clearly as all the announcements showed, Henry and others have been busy getting operators on board.

So: what's it worth?

When we tried to size the multi-sided telco business model opportunity at STL Partners back in about 2010, we came up with a total potential of about $260bn. We caveated that as very much a 'total possible' number relating to a series of assumptions about the telco assets available and the benefits possible from telecoms platforms models in global industries.

To put this time in context, WhatsApp only had 50m users (now 2.8bn), Blackberry still had 43% of the smarthone market, and Amazon's share price was $6.36 (today $178.75). Kudos to my former colleague Chris Barraclough , now CEO STL, who had both the original vision to develop the multi-sided telecoms business model research and the good judgement to firmly caveat the numbers.

13 years later, it looks like McKinsey has come up with a similar ballpark number of $100-300bn in the next 5-7 yeas for the total potential revenue associated with telco APIs.

Most of this comes from "revenue unlocked in other industries" in "connectivity- and edge-computing-related revenue". I haven't been able to find the source analysis for this, so for now I assume that this is an estimate of potential "unlocked" in other vertical plays in the cited industries. Presumably they mean it is extra money spent on things like turbo-boosted connectivity, edge computing rentals to HSPs, and forays into vertical ICT solutions.

McKinsey also say that telcos "would cede as much as two-thirds of the value creation" to other players, based on current market structures, and estimate the APIs themselves as generating $10-30bn for telcos.

I'm not sure why McKinsey thought the telcos would "cede two-thirds" of the $100-300bn, or that 10% of it would be the API take. Perhaps that is locked away in a deep vault requiring many dollars to view, or they've just made some order of magnitude estimates.

I don't really have a problem if so, but I would like to know what sort of estimates these are. Are they 'we spent an afternoon thinking about this' or based on a genuinely deep analysis? I don't think anybody is going to get their cheque book out because an article says "McKinsey says $300bn" but it will make CEOs look, people will talk, and pretty soon that $300bn will become a conversational assumption.

I'm using it here, for example - but I just want to flag this so you keep that in mind!

A storm in a web-broswer near you

There was then a fair old hoo-hah on the back of a classic “meh” post on Linked-In about all this by disruptive Dean Bubley.

Omdia’s Camille Mendler (excellent on most things, especially Enterprise Telco), Ferry C. Grijpink (Mckinsey, a co-author of the $300bn telco API opportunity report) and Alan Quayle (TADHack – essentially Mr Developer) - all worth listening to – all waded in.

To summarise: opinions were somewhat polarised between “there’s a huge opportunity!” and “telcos will never succeed at it like this!”

Will it "move the needle"?

How much of this 'potential' value you believe is relevant to the situation of your company will depend on a lot of different variables, such as what verticals are most active in our markets, how developed they are, and how well your offering serves them - as well as taking into account your view on the realistic total value pot including Dean's and all the other commentaries.

You might also argue that the APIs will enable you to play higher in the value stack, creating new business models and developing other offerings and applications etc. - as per the McKinsey piece and STL's original hypothesis back in 2010.

But for the sake of this thought-experiment, let's take a mid-point figure for the APIs alone, and put it in context as a 'back-of-the-envelope' excercise:

• Total SP revenue right now* is about $1.75trn - $1,750bn;
• So $20bn (mid-point of $10-30bn for APIs only) is about 1% growth on that - and obviously, SPs and others have to get a lot of currently uncertain things right to get here (NB Dean has also written a critique on this estimate, as you might expect, and Ferry has summarised some of the feedback he's had here);
• Depending on your company's circumstances (timing, markets, forecast beliefs, etc.) you should adjust this up or down for what you think it might possibly be worth to you.

So depending on how Dean-ish you feel today, this estimate means between 0 and 1% growth from direct API revenues, and more if it's particularly pertinent to your markets and you factor in other successful future plays serving the verticals better.

If you go with the MckInsey '$100bn' - perhaps ceding that other two thirds and executing well - that's up to 5% growth overall (but remember this isn't just API revenue - its all the rest of the things you might be able to do).

Is it worth all the hassle?

Now you might think to yourself "is that really worth all the hassle?" - and that is a pertinent question that will doubtless go through a few execs' minds.

I think it is, for the following reasons:

1. This is a move that, done well, opens doors and routes to new markets. It should create new relevance for SPs and a chance to build new things.
2. APIs are one part of fulfilling the needs of a vertical – adding richness to an offer.
3. These numbers are still ball-park estimates. None of us know what the true potential is. Innovation is not a game of certainty.
4. We don't know if we've been wildly optimistic or pessimistic in these sums. It could be a lot better or a lot worse.
5. What they suggest is that a) there is some potential here, but b) the approach needed to develop the potential should be measured.

That doesn't mean 'no investment' or 'go slow'. It means move fast, collaborate to reduce risk and improve scale, and be selective - try to focus on bringing something genuinely useful to market fast, together.

How might telcos actually do this?

• According to the GSMA's published figures** as at Feb 24, 2024, 93 instances of APIs had been launched in 21 countries.
• 67 (72%) of these were of the Anti Fraud API type
• Sim Swap was most popular API, with 25 instances launched and 10 of these 'certified' (see chart at the top)
• Sim Swap or Number Verification had been launched by more than one operator in seven countries

While the noise at MWC24 was all about how many telcos have signed up, the principle challenge seems to be how many are getting usage, and I haven't seen much data on that so far. It's early days, for sure, but a big critique leveled in the hoo-hah described above is that you need to provide APIs that developers want and find easy to use.

Time to focus on a collaborative MVP?

It's interesting that SIM Swap and Number Verification are the APIs most launched and 'certified' (in the case of SIM Swap). Of all the APIs, they are the most telco-specific - they directly relate to telco attributes (number, SIM). plus they should in theory provide immediate additional value in solving a specific and valuable problem: fraud.

They are also not particularly 5G specific, like the much hyped 'Mobile Quality on Demand' API. From the research I have seen, there are sectors of the market that are interested in the more data-heavy APIs. Broadcasters, for example, quite like the idea of being able to call up additional capacity and speed for upload links. However the bulk of market interest seems to be in the more basic identity and fraud management APIs.

Looking at the thoughts of Rakuten's Not Normal CMO Geoff Hollingworth in this thoughtful post on API strategies (and telcos' past listen-fails), this well informed post on API strategies from STL's Darius Singh, and thinking about what Alan Quayle told me recently about past developer needs (they mostly used operator look-up APIs), I am wondering if the answer on how to make Open Gateway work better might start from one these APIs (or something like them).

Now, I don't know for a fact that developers want either of these APIs most, and that should be tested. But let's assume that one of these is the most generally needed. After all, these are the APIs that more operators have lauched first (which presumes their decisions were led at least in part by customer input).

Wouldn't it now be a good idea to get all operators to focus on launching this same API in a 'lean-development' sprint to create global minimum viable proposition (MVP)? At least then, there should in theory be one valuable API that developers can reliably use across all ecosystems and most markets?

Doing this should pilot the way to develop all the necessary processes and agreements to get all the APIs negotiated and delivered, thereby either prove the concept and make it easier to scale - or show that it simply can't work - for the least effort/investment?

As a really radical thought, why not make it free of charge across the board for a guaranteed agreed period of say three years? That way there would be no obstacles to use, and the industry could prove its value unambiguously. It would also give the industry time to work out if and how to price something like number validation, say, that works across markets. Of course, this might make reaching agreement with all the telcos harder, but it would almost certainly make the proof of value stronger.

None of this should mean that individual operators shouldn't do proofs of concept on other APIs, but rather that collaborative efforts might focus on a minimum viable proposition (MVP) in the near future.

This really is an area for collaborative advantage rather than competitive advantage. These APIs become much more valuable when scaled and working everywhere, so a developer only has to find and incorporate the API once. Second best might sometimes be where they work in a specific geographical market (e.g. so you can say "at least we can verify this number in Germany").

Plus - don't forget these APIs will have competition. Below, for example, is a lazy Google for mobile anti fraud APIs.

Don't be misled by the "Prisoner's Dilemma" - it's a distraction

Back in the 2010s, there was a lot of talk around the Prisoner's Dilemma, which is a classic paradox of whether to collaborate or compete. In the dilemma, prisoners can all win a prize of a shorter sentence by collaborating and refusing to betray their fellows, or an individual prize of an early release by betraying all their fellow prisoners, who then get longer sentences. In most cases, someone goes for the early release deal.

This has often been the case in the intervening years in Telecoms since 2010. We have seen many operator initiatives fail (in APIs, payments, advertising, etc.) because each time, one or other operator lost patience and tried to do it by themselves, usually achieving a spurt in unfettered development, followed by collective failure because there was no market at scale.

"To go fast go alone, to to far, go together" as Vodafone's GCEO Margherita Della Valle put it rather well at MWC.

Hence, in this context the Prisoner's Dilemma is an unhelpful and unrealistic paradigm. Maybe there should be an updated and more realistic version, in which the choice is you can go for the early release and get it briefly, but afterwards everybody gets a life sentence.

Will telcos actually collaborate to succeed?

Back to the real world, I think the scrum around Open Gateway will continue for a while. The only thing will carry it through is commitment, listening to stakeholders, and obviously - ultimately solving all the challenges.

Whether the industry is capable of this is a hard question because it’s not just a technology problem but a stakeholder alignment one too, and they’re much trickier to fix across an industry. And the industry has failed before - although there does seem to be a bit more collective resolve this year.

Though at MWC24, I felt that the narrative was “if we all say it loud enough, it will work!” Given that I think success here rests on listening to the right people and I didn’t see much evidence of that, that didn’t feel quite on track to me.

And I did feel there was quite a lot of noise about the fancier APIs - the gaming boosts, etc., and maybe there will be a market for that. But not if the approach is not scaled so developers can use it easily. It's perhaps understandable that folks want to use the fancy stuff in promos at MWC, especially when so many people are keen to prove that 5G really was worth all the investment. But I don't think that is where the strategic emphasis should be for now, based on the reasoning above.

That doesn’t mean that the right things aren’t happening or won't eventually happen, even though that wasn't what could be seen at MWC24. But the longer it takes, the lower the potential to take the prizes theoretically on offer - whatever they turn out to be!

I think Henry and his colleagues see what is needed. I just hope he’ll get the ongoing backing, commitment - and openness - from the industry needed to get there.

PS My thanks to Disruptive Dean and Charlotte Patrick , Telco AI Research Queen, who respectively raged at and politely guided my earlier drafts, and provided some great additional insights.

*High level industry numbers sourced here.

**GSMA numbers for Feb 2024 here.