The double bubble: Why refactoring after migrating to the cloud is the best approach
There are two common approaches telcos use when moving workloads to the public cloud. Which one is best?
Dear Telco CFO,
5G investments got you down? Were you promised big new revenue monetization with 5G roll outs that have yet to materialize? Are you left holding the bag on the ROIC calculations that aren’t looking too good right now?
I don’t know if I can solve all of your troubles, but I can give you four great ideas to start battling your VPs with. Here is your cheat sheet for how to save some coins and help fund the next G, which I’m sure is around the corner.
CSPs already working on going fully public cloud will potentially reduce their IT costs by 55%, according to Gartner. Imagine being the telco that finds your competitors are saving millions of dollars, and have limitless capacity on demand, access to phenomenal resources, rapid time to market for new applications and services, and unparalleled security – and realizing that you could have been there first.
As Chris Nims from Capital One said at AWS Re:invent 2020, the public cloud “… offers nearly infinite computing power in real time and nearly unlimited storage at a cost approaching zero.” Start pushing your CIO/CTO to get moving on the public cloud journey.
Since everything will be running in the public cloud, you can eliminate most or all of your data centers. Capital One recently closed all of its data centers (down from 8) to move its applications to AWS. Do a quick back of the envelope calculation of the savings on real estate, electricity, hardware, maintenance, headcount and just the overall headache that goes away with the elimination of your data centers.
Want more data? Read what Deutsche Telekom did with the reduction of 76 (!) data centers. Triple-digit million-euro savings annually. WOW.
While you’re at it: kill the office.
In 2020, your organization went through a real fire drill to move everything and everyone to remote work. And guess what? Business managed to continue, teams were productive, while all of your office space was left idle (and you continue to pay rent). Don’t go back to the old way! Take this opportunity to change. Just look at Pinterest, which reportedly paid $89.5m to terminate its San Francisco office lease when they figured out they no longer needed a big fancy office. Better yet: do it for the people. It’s how people want to work. According to Slack, only 12% of workers want to return to full-time office work, and 72% want a hybrid remote-office model moving forward.
2021 is the year in which OpenRAN is going to take off. You know it’s a great idea when network vendors start writing blogs about how crappy it is; that’s because OpenRAN is a serious threat to their future. Pioneers like Rakuten claim operators can reduce capex by 40% with its telco-in-a-box network. My advice to you: get out there and start experimenting! You may have setbacks along the way, but you need to start experimenting now so you can roll it out in earnest at your next earliest opportunity. Learn what the objections are to the new technology and what issues exist and start to calculate if the savings are real. Take the leap and use that money to pay for …
Ah, so here’s the bad news. With all of this movement and change, you’re going to have to find some extra cash to pay for the move one way or the other. I call these costs the double bubble.
What is the double bubble? It’s the cost of continuing to pay for something to run in production or be used (like a data center or office space) while you transition to the new way of working.
Here are some examples:
These costs are inevitable. I see too many organizations shy away from these double bubble costs because they don’t want to pay for them, can’t afford to, or it just seems wasteful. The problem with this approach is that it results in never making the transition to the new way of doing things and, therefore, never realizing the savings. You never change because you’re paralyzed in fear or penny pinching. You have to take the plunge!
With the exception of OpenRAN, I’ve personally done all of these transformations:
At one of my former companies, a massive $15m a year was being spent on ten data centers covering 56,000 sq ft of space. With the help of Pythian, I migrated 3,500 machines to GCP in just 3-4 months, becoming more agile and reducing costs. We took IT spend from $15m to $1.5 and the project cost $650k. The project paid for itself in eight months.
At another former company, $5m a year was being spent on five data centers. By moving to AWS and optimizing heavily, we reduced spend by 90% to $500k. The double bubble cost was ~$300k and the project paid for itself in six months. Phenomenal!
Perhaps the biggest challenge and greatest achievement – taking 28 offices with massive overheads and reducing them to just four in only 12 months. We had a double bubble of $1m USD as we wound down the office and exited leases. The double bubble paid for itself in six months.
Yes, there was a double bubble of $2m spend, but we never would have achieved close to $20m in savings had we not embarked on these projects. The art to managing the double bubble is:
Depending on the move, you should see your double bubble costs paid back within just 6-18 months!
The bigger the project, the bigger the savings, but also – the more things can go wrong. Your teams are going to push back: “we can’t” … “it’s too hard” … ”it won’t work.”
Your response? Yes, it is hard. But if we plan this right, it will work.
The double bubble cost is not a good reason to not do something. Create a business case, push hard, and don’t take no for an answer. It will pay for itself in time and you will be rewarded with all the savings and simplicity the new world brings.
If you need help with these projects, give us a call. We have teams that can help you with items 1 – 3. And OpenRAN? I’m sure I could find someone to provide the help you need. Just give me a call at 925-TELCO-DR.
Later, nerds!
Recent Posts
"*" indicates required fields