Virtualization is not just a new technology for telecom: virtualization revolutionizes the basic economics, and changes many long held assumptions about costs, what’s feasible, and how networks operate and should be operated. Or, more accurately, it holds that promise – if we collectively take the right actions.

At Appledore Research Group, we often like to return to “first principles” rather than accept “common wisdom”. So, as virtualization emerges as the technology for tomorrow, and in some cases today, we elected to go back to the “why?” and figure out what virtualization can do, financially, for a business.

Very briefly, our approach was to benchmark today’s networks and operations as quantitatively as is practical, and then look into the areas where virtualization had significant and even radical impact on costs. This allowed us to create a qualitative, but valid, model of the how a business could differ in a virtualized world, and what actions CSPs and suppliers must take to realize that transformation.

pie-chart-for-econ-blog

“Follow the Money” is always good advice: A breakdown of Expenses

For details, we refer readers to a pair of framework reports – “Market Forecast for Virtualized Network Infrastructure (Cloud CAPEX)”, ARG 2015 and “The Economics of Virtualized Networks”, ARG 2016. Below I’ll discuss a few of the most germane and occasionally counter-intuitive findings.

Its important to realize that well under half of the capital employed in networks can be virtualized or benefits from virtualization; therefore we must continue to plan for assets and management tasks where capital is allocated or regionalized and cannot be easily “instantiated”, “moved” or “scaled”. Yet at the same time this is an asset: CSPs have what hyper-scale datacenters do not: low latency proximity to customers, vehicles, and “things” of all types. This can be critical to low-cost, low-latency services.

Much has been made of the move from dedicated and specialized hardware to generic, uniform X86 based infrastructure. No doubt, it is cheaper on a “MIPS per unit spend” basis. Yet our analysis finds that if capex savings are an important goal (and they should be), the far larger prize comes from better utilization of embedded capacity – a ratio that is very poor today for myriad reasons outlined in the documents. This implies that operational automation and grooming may be “trillion dollar issues” – not merely technical playthings for leading-edge engineers.

Similarly, we find that virtualized technology may completely change the assumptions that underlie product innovation and packaging – driving down costs and complexity, and making new revenue sources and market segments profitable. Let’s say that differently: cost cutting can be used to cleverly grow revenues and margins. The charts are in the reports.

revenue-curves-diagram-for-econ-blog

Yep, Technology Can Help Capture More Revenue and Margin….

In the end, what we found was a very compelling case for the transformation of business assumptions, management processes, and the supporting OSS, BSS, MANO domain controllers and most importantly – the architecture they are designed to. Automation becomes critical to profitability; micro-services and modularity become central to flexibility and agility.

In summary, by returning to first principles, what appeared to be a very interesting, and certainly cheaper, technology suddenly transformed into an engine of innovation, revenue growth, capex utilization and, yes, cost reduction.

We encourage CSPs and software suppliers alike to think about tomorrow’s operations as potentially very different from today’s, and think about “how will we yield truly transformative change?”. And, of course, we recommend reading the reports below as a great starting place….

https://appledoreresearch.com/report/virtual-network-capex-global-forecast-2015-2025/

https://appledoreresearch.com/report/economics-virtualized-networks/

 

Thanks for reading,

Grant