The MEA Cloud Wars (~2017 - Present)- Episode III

How did we get here?

This is the 3rd & final episode of the MEA regional Telecoms & Internet services competitive analysis series. Similar to the previous two episodes, the region’s key Telco players found themselves trying to win this regional battle against their neighboring rivals, although I should point out that not every Telco engaged in this battle had fully prepared for it internally, and in some cases, Telcos are not even aware there is an ongoing battle!

Unlike the previous 2 rounds of battle, however, the MEA Telcos couldn’t win relying solely on their own resources/capital. They needed considerable support from specific partners in the new Telecoms/Connectivity ecosystem. Additionally, and as will become clear shortly, this ongoing war continues to put considerable pressure on the regional Telcos, almost tearing their outdated organizational structures apart in some cases (can you say “silver lining”?).

Setting the scene

Globally speaking, the top 5-6 Cloud Service Providers (CSPs) have been engaged in a series of regional Cloud Wars for the past few years. Each one of the top 3 players in this space (AWS, Microsoft Azure, and Google Cloud), in addition to Oracle, IBM, HP, and Alibaba have been ramping up its global coverage to reach as many customers as possible, as efficiently as possible; offering state-of-the-art IaaS solutions. The MEA Cloud Wars are set against that global backdrop, and it adds another layer to the global competitive narrative through the coupling of regional Telcos/Carriers with those global CSPs.

This means that the Cloud Wars around MEA at this initial phase are primarily “inbound” in nature; focusing on attracting key global CSPs to specific locations in collaboration/partnership with the Telco there. In a subsequent phase, an “outbound” campaign mode will inevitably emerge wherein those in-region CSP facilities will be leveraged to attract cross-border customers. Customers in this context can be both businesses AND government/semi-government entities, hence the gravity of the decisions associated with partner selection.

Critical Success Factors

Unlike other regions of the world, where only commercial & operational considerations determine the chances of success of a specific CSP against another, the MEA region (due to its complicated regulatory environment) has a couple of strategic decisive factors that will “frame” to a large extent the Telco’s tactical maneuvering as they engage those CSPs in this regional battle.

The biggest enabler for true success in this regional war is – of course - Big G ("Governments"). The presence of a favorable government policy on cloud services in a specific country/market goes a long way in terms of “signaling” to the Telco players in that market that their efforts will be rewarded. In cases where the government has not yet figured out how to “deal with” this new cloud “thingy”, the chances are slim of achieving anything material, even with the most intrepid Telco. No matter how large a specific market is, a “Pro Cloud” government policy – rooted in a vision for a digital economy – is a must for that market to be considered a contender in this regional race.

The second biggest enabler for success is Big B (the Telco's "Board of Directors"). It is incumbent on the BoDs of those Telcos who aspire to play a leading regional role to resolve some really tough conflicts that their executive management teams face as they build these alliances and try to attract credible partners. While those Telcos have invested heavily in super-important network facilities without which a CSP would be dead in the waters (international submarine cables, landing stations, data centers/Colo space, and national backhaul rings that reach key business hubs), those strategic assets need to be “untangled” somehow from a number of internal legacy strings attached to them, before management can action their battle plans. It is incumbent on the BoD members – therefore – to realize ahead of time that in order for their Telcos to fully capitalize on the cloud services opportunity, they need to re-examine their organization structure and ensure it is matched to the opportunity at hand.

The New Business Connectivity Landscape

As the enterprise architecture shifted over the past few years towards increased adoption of cloud-hosted business applications, the Telcos found themselves not so much catering to “site-to-site” connectivity needs for their enterprise accounts, but rather to “cloud interconnection” needs. Large enterprises effectively stopped asking carriers the question “Can you connect these 20/30/50 sites through a secure private network over your infrastructure?” and started asking the more vexing question “Can you securely connect my site(s) to Cloud A, B, or C?”

In the new architecture, enterprises now have their own “PoPs” inside large downtown data centers (preferably carrier-neutral, but that’s still a rarity in MEA) and those PoPs want to interconnect with other PoPs (those belonging to the CSPs) to access specific business applications. Telcos who want to compete in this new landscape need to be able to not only facilitate this interconnection (a bona fide carrier concept) but to somehow attract the right kinds of services providers that those customers want to interconnect with (namely the CSPs).

There are 2 levels of infrastructure deployments that CSPs can utilize to serve those large in-country enterprise customers:

• Cloud On-ramp: This is essentially an edge node for the CSP’s global network that is deployed in-country, and mostly hosted within a particular Telco’s data center. The actual services/servers the customers would be accessing are hosted at an international location, so the access speed/latency is about the same as before deploying the edge node locally, but what customers gain is more security of their data since they are connecting to the CSP network locally, and the connection from the edge node to the CSP core is efficiently engineered & secured by the CSP.
• A full-fledged cloud data center where the CSP locally hosts equipment that offers compute, storage, and other cloud services. This is of course the best scenario for customers and the one where MEA as a region is lagging behind other parts of the world.

The MEA CSP/IaaS Landscape

TeleGeography does a decent job tracking deployments of cloud infrastructure across the globe, including MEA. A snapshot of infrastructure deployments as of Q1'2019 speaks volumes about where MEA stands (which in itself - and for the Telcos with the right strategy in place - is truly a unique growth opportunity).

We can also peek into the 2019 planned global cloud date center deployments (pipeline of infrastructure development) and see how MEA compares here (this snapshot was in Q1'2019):

OK, so what about the regional MEA CSP data center (CSPs actually call them Availability Zones or Regions) deployment picture for each individual CSP?

Publicly available information from AWS, Microsoft Azure, and Google Cloud paint a very clear picture, which - as you would expect - is not materially different from the On-Ramp situation. If you consider AWS you'll find the following global infrastructure map:

While for Microsoft Azure, the picture looks as follows:

And finally, the Google Cloud Global Coverage is shown below:

Of course, the world is not restricted to those 3 CSPs, but the point here is that the MEA region - regardless of CSP - has a lot of potential as the current level of deployment is not in line with the other regional indicators. The regional race is heating up, though, and in the past 7 months (January'2019 - July'2019) 2 of those 3 CSPs established their own regions/zones in MEA (those dots on the map), so there are encouraging signs of coverage improvement.

What will be super interesting to watch unfold is how some of the biggest markets in the region (which have not seen any CSP infrastructure deployment yet) will competitively position themselves in this race.

What's at stake?

It is difficult to imagine the major CSPs deploying infrastructure (of either flavor) in every MEA market. This means that several markets (with all different classes of customers in each of those markets) will inevitably be served from infrastructure that is hosted in a neighboring country.

A. For the CSPs this forces a certain selectivity in terms of market analysis & ranking, in order to ensure they get a sufficient level of infrastructure deployed to cater to each market's specific opportunity and avoid a situation where a competing CSP gets a first-mover advantage into a key market. There is - however - a very short list of markets that can sustain several CSP data center deployments simultaneously.

B. For the Telcos this dynamic means they only have a few really big/worthy partnerships that they can focus on developing, but that's for the first mode of deployment (the CSP edge node) as the full-fledged zone/region deployment is done independently from a Telco (to maintain the CSP's global value proposition). Still, prudent Telcos' executive management can extract a few interesting synergies (generating a nice sustainable revenue windfall) from the presence of an availability zone/infrastructure region in their service area.

C. For diversified MEA Telco Groups with multiple OpCos across the region, there are - at least on the face of it - clear advantages in this race, but such a portfolio needs to be carefully aligned in a way that strategically leverages this multi-market operating footprint in a competitively meaningful way (against other regional rivals), as well as create more value for their target CSP partner.

D. Finally for Governments, the stakes are the highest. If a specific government misses the opportunity to develop a set of incentives that would be attractive for a credible global CSP to deploy infrastructure in its country (thereby achieving a number of benefits for its local economy, not only for its digital services market), then business users of cloud IaaS solutions will have to resort to accessing infrastructure hosted next doors. For a commercial business (a bank, a manufacturing company, etc.) the penalty of such a scenario can be manageable (a few extra milliseconds of data access RTD, and maybe some international access charges for a private connection), but for government business users (a ministry, the Central Bank, a Stock Exchange, etc.) there might be bigger issues to contend with if the infrastructure is not inside the country.

The Conflict of Interest Dilemma

A very peculiar problem arises within the typical MEA Telco organization as it goes to battle in this regional conflict. Because of the vertically integrated nature of the incumbent Telco in almost all of MEA’s markets (where the same company provides mobile & fixed services to the consumer, enterprise, and wholesale segments), providing market access to a global CSP can actually put that Telco’s enterprise revenues at risk (what enterprise would procure connectivity or application services from its local Telco when it can get world-class cloud services right across the street?)

This dynamic is the most vexing one that almost all Telcos currently struggle with as they monitor the regional race and try to determine how to play. The "monolithic" Telco (with its business model firmly built around locking customers into long business services contracts) is too rigid to accommodate an "open-for-all, take what you want, switch when you want" model of the CSPs. Protecting the local enterprise revenue (holding the CSPs at the gate & blocking their market entry by all means) may seem a logical option, and it may very well be tactically speaking, but it would be strategically fatal given the regional implications (and unhindered cross-border market reach) of these Capex-intensive infrastructure assets. This conflict is one of those circles that only the BoD of each Telco can square, as the executive management team would not risk it without a mandate that is long-term and rooted in a holistic vision of the market.

Tip: Before thinking about inviting the proverbial fox into the proverbial henhouse, "Old MacDonald" Telco should think about what kind of proverbial farm it wants to run!

Closing Remarks

This war just started a few short years ago, so it is way too early to call winners & losers. There are some early leadership positions that have emerged both on the CSP side and the Telco side, but these may not necessarily last. The key point here is there are very valuable "pertinent" lessons from the earlier rounds of the MEA competitive race for infrastructure services dominance, and players - especially governments & regulators - who aspire to win this round would do well to heed those lessons.

End of the Digital Warfare series...