The MEA DNS Wars (1998 - Present) - Episode II - Part 2

The State of The MEA Internet (The Wholesale View)

If you recall the discussion in Part 1 of this article, a few international carriers had already established IP Transit POPs on a revenue-sharing basis in a few MEA markets, but those did not manage to bring down the cost of wholesale IP Transit across the region. Those IP Transit POPs also did very little to improve the quality of service (measured by Latency/RTD) offered to customers, as network/service requests still needed to be long-hauled to - mostly - Europe to fetch the required content and intra-region IP traffic was really intercontinental traffic (the notorious "traffic tromboning effect", explained here What is the trombone effect?).

To improve quality of service to major Internet networks/content (Google's "walled garden" of YouTube, Gmail, Docs, and other goodies was the favorite target), carriers simply resorted to hosting edge caches in-country, so the most popular content pieces were served locally rather than internationally (saving the carrier a decent amount of costs for international bandwidth & IP Transit ports that otherwise would have been needed to deliver that content to customers in-country).

Neither of those network components is a replacement for a "bona fide" Internet Exchange (IX). Carriers really did NOT have a commercial interest in establishing such IXs, as that was the role expected of regulators or other independent industry governing bodies. To see the results of this reluctance, all you need is a quick glance at a map showing all Internet Exchanges globally (a very good one can be found at Internet Exchange Map), and then the true picture of MEA's Internet landscape hits you. There are very few IXs in MEA, and most of the ones that show up on a search do not strictly meet the definition of an IX, as they are EITHER maintained by a single carrier and are set up as an IP Transit facility, OR they are no longer maintained/supported at all (there is no shortage of white elephants in MEA!)

With probably one exception, the global IX operators have only very recently found it feasible to deploy POPs in MEA. Regional carriers - who are traditionally very protective of their turf and historically viewed everything Internet-related with skeptical eyes - eventually capitulated, and so a few interesting partnerships developed that would see global IXs offer Peering (as opposed to IP Transit) to large customers/networks regionally. You can find a very nice (and concise!) discussion of the success factors of an IX at IXP Models.

I should probably close this point by stating - again! - that IXs should have been at the top of the Telecom policymaking agenda, and not just left to the operational & commercial convenience of Telcos/Carriers. Regulators across the region were generally lax on that front and were almost obsessed with regulating consumer mobile services or - in a few cases - international facilities services, and so IXs were never seen as a true priority to develop the Internet economies of MEA countries. There is a nice discussion for the IX necessity in the African context in this Building Local Exchange of Content in Africa article.

The State of The MEA Internet (The National/Macro Market View)

So, after all is said and done, and all those wholesale IPT Transit deals made, IXPs established, websites & domains registered, and all the regulatory policies were put in place, how does the MEA Internet landscape look?

One of the most insightful studies about the state of the Internet across MEA was the 2015 MEAC DNS Study commissioned by ICANN and done by EURid. I am not aware wheather a newer study has been conducted since then, and so this remains a valuable reference point. You can find the 130-pages study online here EURid MEAC DNS Study.

The study provides many valuable nuggets of intelligence, but I will just pick out 2 for this discussion.

A. Only 5% of websites in the MEA region are hosted locally!! Even if we assume that ratio has doubled between 2015 and 2019 to reach - hypothetically speaking - 10% local hosting, this remains an appalling figure compared with other regions of the world, and it painfully shows BOTH the deficiency in local infrastructure (data centers, colo facilities, etc.) AND the complete absence of regulatory oversight aimed at developing that market (through incentives, licensing regimes, etc.)

B. More illuminating - and super fitting for discussion given the title of this episode - is the situation of DNS registration across the MEA region. There are considerable difficulties & administrative red tape (including extensive security constraints!) that individuals/businesses who seek to register domains within their own country's ccTLD (like .ae, .eg, .om, and so on) face, so they end up resorting to international registries.

The chart below should be a real eye-opener, to say the least. It shows - again, as of 2015 - the relative positions of MEA markets within the registered domain names race, if you will. The report states that the annual growth percentage across the study region was a whopping 20%, which is not that impressive really given the low base the region is moving off of (the law of small numbers).

Still, you can see which economies are ahead of others in terms of embracing the Internet and establishing a presence online, either for marketing/advertising or for hardcore business transaction purposes (E-commerce). Interestingly, and beyond all the hype surrounding a few MEA markets as regional leaders in Internet services and all that good stuff, it is actually 2 countries that are generally considered OUTSIDE of MEA (Turkey & Iran) that are ahead of everyone else!!

One of the unintended consequences of the challenges of registering DNS locally across MEA (caused by regulatory red tape, lack of local registries, and virtually non-existent hosting facilities) was a spike in the adoption of social media services, as businesses resorted to those to establish an online presence, interact with customers, and market their wares/services. While this was also way cheaper to operate (no domain registration fees, no web site development or hosting charges, etc.), unfortunately, this trend had some unforeseen ramifications for a few countries.

Winners & Losers

MEA as a whole is clearly the loser!

While there are great differences between the Internet access speeds and the maturity of online services offered across the MEA region (both government & business services), it is clear that the entire region is lagging behind in the global context. Beyond statistics about Internet penetration and broadband coverage per market, the national economies of the biggest MEA markets are still unable to take full advantage of the Internet as the engine of the 4th industrial revolution (with data as its fuel). Business use of Internet services & technologies is largely in its infancy (meaning less transactional and merely informational), and the amount of content that is produced inside MEA is statistically insignificant. Qualitatively speaking, even for the content that is created within MEA, it is dominated (my anecdotal observation) by User Generated Content (YouTube Videos, etc.) and so has little sustainable economic value.

The moral? MEA Internet services - across the board - represents a great growth opportunity for Telcos (or more generally, for "infrastructure services providers"). I say this to everyone who argues that Telecoms is dead and it is a commodity business with no real growth outlook. Maybe Telecoms has no future someplace in the world, but in MEA, Telecoms (as far as those Internet services covered here are concerned) has the brightest of futures!

Let the - right, for a change - games begin, please :-)

End of Episode II

Episode III (next week): The MEA Cloud Wars (the juicy stuff, finally!)