We’re from the telco, and we’re here to design software

Last week, we came across a Iain Morris article in Light Reading that provoked a lot of conflicting, strong emotions. We consider Morris one of the most knowledgeable networking journalists out there, so we treat his writing with a high degree of respect, but this one really struck a nerve.

The article outlines European telco giant Vodafone’s plan to hire 7,000 software engineers to augment its team of 16,000. It also contains details on how Voda plans to use those engineers, with a plan to focus on Business Support Systems (BSS), the IT systems telcos use to manage their network operations. The focus of the post is on what will remain with 3rd party vendors, like Netcracker who provides Voda’s billing systems, versus what can be done in-house.

At first blush, this post fits neatly with what we wrote a few weeks back – there are now many vendors looking to provide telecom operators with key systems, but the large carriers are still likely to go it alone, avoiding the siren song of the Cloud. So on the one hand, this initiative makes sense, a big company is making the choice to retain key functionality in-house because Cloud economics do not work for everyone.

On the other hand, little fills us with as much dread as the idea of a telco building software. Once upon a time, telcos all built their own billing systems, going back to a time before software even existed, but over the past 100+ years, most telcos have shed those capabilities. For many years, these systems rested with a suite of highly specialized providers – either network equipment makers like Nokia and Ericsson, or telco-focused billing providers like Amdocs. We can think of very few telco software experiences that are pleasant, let alone up to date with modern app design standards. The AT&T app we use to check our bill is functional, but not in a way that inspires much confidence. By contrast, the start-up landscape is littered with the wreckage of companies that tried to bring interesting software to telcos over the past twenty years. The telcos have all tried various forms of incubator labs, venture funds, and “hack-a-thons”, with very little to show for it. Put simply, the best software engineers do not want to work for a highly regulated bureaucracy, and that was true before today’s incredibly tight labor market.

To be fair, the world is changing. There is a lot of activity in the world of telco software today. Analyst Alan Quayle does a lot of interesting work around these changes and this space. And we imagine Voda conducted a thorough analysis of the economic trade-offs of buy vs. build before embarking on this path. They are also trying to hire the bulk of those 7,000 engineers from re-trained internal hires, so theoretically people with a good background in the business logic that needs to get built (we are aware that this argument cuts both ways). So it is perfectly reasonable to assume that Voda can do well with this strategy.

That being said, we think this should not serve as a model for other telcos. Voda is a unique construct, with multiple national carriers all operating in a somewhat loose confederation. Coordinating purchases across these entities is very difficult (which we can say with the irst-hand experience of having tried to sell into all those entities individually). So maybe something about Voda’s structure makes this path more palatable. But that does not really hold for others. Voda is also large enough that they may enjoy some economies of scale by taking this approach. For everyone else, we think the changes brought by 5G allow for the opposite – for telcos to make broader use of 3rd party software solutions, allows them to bring in best of breed, allows them to experiment in ways previously not possible. And that is a better path.

3 responses to “We’re from the telco, and we’re here to design software

  1. Amen. I’m not sure why they would go down that path unless to negotiate a better deal with existing vendors like Amdocs. They know they are not nimble enough to build those solutions, they know there’s a huge lack of resources on the market (and working for SP is everything but sexy/fun). Good luck to those guys but I do hope they have a really, really solid exec/management team to handle this endeavor.

  2. Compare and contrast to CARIAD/Car Software org that VW is trying to build. huge in-soucing activity…effectively from zero

  3. I don’t think they do know this. They have 16,000 software engineers already, and they’re adding 7,000 more. That sounds to me like they think they can build a core competency around it.
    I think a big part of their decision making on this is very specific to the nature of Vodafone’s corporate structure. It may just be easier to produce something in-house than try to get all the holding companies to standardize on one vendor.

Leave a Reply