Insights Insights

Has the public cloud failed to crack telecom?

Bexprt CEO and Founder Mo Hamdy responds to the recent Light Reading article by Iain Morris, and suggests a pathway for the big beasts in the tech jungle.

“The public cloud has failed to crack telecom” – a provocative article headline, that deserves a more nuanced response! Sharing some thoughts, and I welcome yours!

Having worked with CSPs across 90 countries over the past 20 years, and working now on the other side of the table with hyperscalers on Gen AI and resilience projects, my perception is that reality is some where in between.

CSPs and hyperscalers are two beasts in the same jungle, with different attitudes, tactics and motivations.

For CSPs the challenge is always trying to “differentiate” by being a very smart pipe and a “beyond connectivity provider”. Hyperscalers see the telco industry as a very appealing infrastructure-intense sector, with significant opportunity to move their infrastructure to the public cloud – assuming data solvency is not the issue now. As hyperscalers have their own telco experts, they often assume everything will be migrated smoothly and just work fine!

I firmly believe that adding the hyperscaler’s capabilities to that of the CSPs will bring both optimizing-synergy plus innovation, and will be a great mix that the world has yet to realise the value from. However, there is a journey in the jungle here that both beasts need to talk together, sharing the benefits and risk!

I boldly suggest,

📍CSPs: be open minded when considering cloud technologies! Hyper-scalers are not a passing trend or something for big doom-laden cumbersome projects, or limited to just a hosting platform for workloads. They are not a “project” to be passed to the IT department. For those CSPs taking the“DIY cloud” route: absolute control and security are tempting concepts, but the resources the hyperscalers have in that field will always be far greater, and that’s without mentioning the access to the billions of dollars on innovation all hyperscalers make; DIY is a cul-de-sac to step out of swiftly.

📍Hyperscalers: understand the business models of the CSPs, and appreciate that they are beautiful elephants with technical complexity, multiple-layers of legacy technical and commercial constraints, interwoven decision-making and financial and regulatory responsibility processes, and huge infrastructure costs. Focus on bringing value-driven use cases with a clear value proposition, and think about the quick wins to start small with an experimental mindset to help CSPs realise the value, and to build trust that they will have profitable projects!

I suggest that CSPs should start the journey with hyperscalers in phases: explore, build trust, commit.

  1. Exploration: as with any enterprise with complex supporting functions, start exploring opportunities in the supply chain, finance and accounting software, warehouses, and start migration projects for these domains. DO NOT TO TOUCH THE NETWORK!
  2. Building trust: Internal and external use cases* which assist the CSPs to achieve one of these five strategic goals: revenue generation, cost optimisation, operational efficiency, enhanced customer experience, and fast time to market, whilst again NOT TOUCHING THE NETWORK!
  3. Commitment: now trust has been established between the CSP and hyperscaler, and there is a solid understanding of the hyperscaler’s capabilities, and the two teams talk the same language, now they are ready to start moving some critical workloads like Core, Radio, OSS/BSS with confidence.

So has “the public cloud failed to crack telecom”? I suggest that more accurately, public cloud has yet to reveal its true potential with telecom. This journey, to reach the point of optimising the network together, will take months if not years of alignment and working together, but to be successful, it must start with a solid base.

Lastly, hyperscalers should never start their work with CSPs with shift and lift projects. Instead, focus on use cases, real issues, where data and advanced analytics can reveal nuanced and novel new approaches, for internal and external use. Demonstrate skill, value, impact on customer experience, build trust, and from there, the journey together will have far greater possibilities.

*I will write another blog soon about the practical use cases for the public cloud, internally and externally for CSPs.

This article was first published on LinkedIn, 23rd March 2024.


Bexprt ◦ Your cloud business partner

Specializing in Advanced Analytics, Generative-AI, Resilience & Telco Cloud.

image credit: Robert Bye