Telcos planning to offer edge compute services should draw wider lessons from public cloud computing.
For that, we need to go back to 2006 when Amazon launched S3 and EC2. At the time, Amazon’s cloud services were seen more a curiosity than a serious alternative for enterprises to entrust their mainstream compute workloads and data. Indeed many viewed it as unreliable, insecure and more expensive than owned infrastructure (particularly once enterprises had extracted benefits of adopting virtualisation for their private clouds). And what’s more, the services came from a online retailer – hardly a credible IT service provider. This negative view persisted for several years.
Lack of agility holds telcos back
Start-ups and developers in enterprises were the leading adopters of public cloud as they embraced the flexibility it offered. This flexibility meant that, with a credit card, they could be up and running for a fraction of the cost and none of the hassle or delays associated with securing “physical” IT resources. They had fewer restrictions – no making a business case, securing budget, satisfying compliance, selecting over-specified hardware and software from approved vendors, ordering, shipping, installing, on-boarding support, testing, maintenance. You know the drill.
Enterprises tolerated the unsanctioned use of public cloud for “grey” computing as short-term expediency. If these temporary “skunkworks” proved useful, they could be moved back under the formal IT umbrella for production.
But it didn’t work out that way.
Multi-cloud is the new norm
Enterprises adopted SaaS and then embraced hybrid cloud. Multi-cloud is the new norm and public IaaS a $50bn market. Public cloud is now cheaper, although arguably still more expensive than owned-compute infrastructure. Particularly if telcos don’t factor in high levels of utilisation for the life of the hardware and all costs (including opportunity costs). This is a key point for telcos planning their edge compute businesses – success depends on lowering adoption barriers.
What opportunities are there for operators in offering edge compute?
I speak to strategists and innovators at telecoms operators about their plans for edge compute (XaaS to enterprises and third-party application providers). Although most see potential opportunity for operators in offering edge compute services, none claim to have a precise idea where this opportunity lies: the enterprise edge? the mobile edge? the IoT edge? the city/sovereign cloud? A few are doubtful that there’s any opportunity at all.
They argue that, since none of their customers have asked for edge compute, there’s no potential demand. Others point to operators’ track record in public cloud. Here, they state that although there could be potential demand for edge compute, telcos are genetically incapable of capitalising on this.
Five ideas for operators to build edge compute
In simple terms, operators looking to build edge compute business should:
- Make it easy, fast and simple for developers and application providers to start using their telco edge services.
- Plan and design for using telco edge with on-premise edge and core (private and public) cloud.
- If you’re following a strategy focused on enterprise edge, develop “hybrid edge” propositions, including in partnership with on-premise edge compute vendors.
- For operators pursuing “city cloud” strategies, identify and develop approaches to minimise upfront investment from local authorities while allowing them to maintain control.
- And if you’re targeting IoT edge, acknowledge the importance of hyperscale players (Azure, AWS, Google) as a channel. Co-opetition applies here.
Telcos hoping to bridge the gap between on-premise/on-device edge and hyperscale need to focus on building bridges. This will be more productive than building walls around empty gardens.
This article is from a piece first published on LinkedIn by Philip Laidler, Consulting Director. You can follow our LinkedIn page for news and information.