I try very hard to dig out the truth for any technology blog I do, and one way to do that is to constantly check what I’m finding against other reports that target the same technology. I’ve blogged on telco attitudes toward open networks, based on what I hear from 81 operators. SDxCentral reports the result of an Analysis Mason study (commissioned by Dell) on that same topic, so I want to look at the article, as well as another that cites Omdia and Technology Business Research (TBR) comments and is related (it seems to me) to the first piece.
The Analysis Mason study included 50 Tier One operators, and they say that only 20% of them had an open network strategy in place. There may be a semantic point here, but my own interactions found that 59 of 81 operators, and all the Tier One operators, had a de facto strategy, meaning they’d looked at open network technology and made a decision on how it will impact their infrastructure plans. Did they all have a formal strategy? Without defining just what would be involved in that, I don’t know the answer. In Tier Ones, I found that pretty much all the CxOs knew what the company’s position on open networking was.
The report does say that “We are seeing that operators are really interested in building this more disaggregated and more open interfaces and more horizontal kind of network clouds for their 5G and potentially 6G networks. So interest and understanding is there.” That, to me, means they were requiring a lot of formalism in the open network strategy, but agreed the majority of operators had in fact assessed an open approach. However, I didn’t find operators talked about abstract things like open interfaces and horizontal clouds.
The Omdia, report was on operators’ digital transformation, and the article on that topic didn’t even mention open networks as a factor in picking the leading operators in that space. Instead, it talked about AI, ML, and automation. Just as we have to ask what an open network strategy would look like, we have to ask what “digital transformation” is, because that term was used only three times in over a hundred and fifty interactions I had.
Rather than try to uncover meanings in someone else’s work, let me instead introduce what I heard from operators. With regard to open networks, their objectives were to control costs (especially capex) and minimize the risk of vendor lock-in. With regard to transformation, their goal was to control opex.
My previous blogs on telco open initiatives (HERE and HERE, and my Network World article) found that because operators were not broadly excited about getting into the higher-layer services space, they didn’t see open networks as a pathway to new revenues. Given that, business sanity and the need to make a business case dictates that it be aimed at cost management. The same would be true for digital transformation (or whatever we call it).
What then unites the two studies, and also my own telco analysis, is the telcos’ almost-irrational and possibly lemming-like-suicidal rejection of any radical service changes. That’s a problem because there’s no theoretical ceiling on how high you could push revenues, but you can’t reduce costs to zero, or even near-zero. As a strategy for transforming a telco business model, revenues would win, but telcos have simply not been willing to take realistic steps in that direction.
In my view, open networking in the Analysis Mason sense of open interfaces and horizontal clouds is justified if you presume a vibrant community of developers building to those interfaces and deploying on those horizontal clouds. Can anyone with a straight face envision a vibrant developer community tweaking connection services at starvation profit levels? I can’t, nor can at least three quarters of telco planners who offered me their views. But even willing developers couldn’t overcome operator ossification with respect to service mix.
If you pull this happy developer community vision out of open networking, you’re left with the fragile cost minimization mission to justify it. The problem with that is that an open network model that drives down equipment costs does that by generating competition, and to realize the benefit operators would need to select price leaders in multiple categories. They generally don’t like that idea because it increases integration costs, finger-pointing, and vendor stability risks. One planner told me “If you assume any vendor has a ten percent chance of withdrawing a product you’re using, and you have one vendor, that’s a ten percent risk. If you have five vendors, the chances that at least one will fail is almost fifty-fifty” (I get 41% in the math). This mitigates any benefit achieved by price-shopping across vendors, which means that open networks don’t actually save money, said the planner.
How about the opex side, the “AI” and “automation” story? Another analyst (TBR) cited in the second SDxCentral story said that the two would produce “more efficient network architectures”, but I wonder whether they can do anything to make infrastructure less expensive by improving network efficiency. Opex, perhaps, but opex has been the focus of cost management for decades, so it’s not clear how much can be wrung out of it now.
Is open network progress possible? I think the first citation above supposes it is, because it asks what’s preventing open network adoption by telcos. I’m not convinced that telcos will really accept it as a goal as long as they see it only as insurance against vendor lock-in, which is what it is today. But that isn’t necessarily bad in itself; what is probably bad is that telcos have forsworn the real benefit of openness, which is the ability support a vibrant development community. If you can’t or won’t offer those developers a great opportunity, you’ve killed that benefit.
Is telco digital transformation possible? That’s actually a much more difficult question because of the general softness of the term and the specific challenge of applying the enterprise definition to telcos. I’ve written about the former, and my view is that “digital transformation” for an enterprise is a large-scale activity to rebuild business processes to optimize the technology available to support them. IT and network technology, for enterprises, are tools in business operations. For telcos, IT and networking are arguably the product. There is an enormous sunk cost in telco IT and networking, and enormous inertia in business practices related to them.
