Is Enterprise IT starving its operational process by reducing headcount and overcapitalising on assets ? If so, what arguments can we make for or against this idea ?
CapEx Rich but Living in OpEx Poverty
Is it a misconception that heavy capital purchases can reduce operational costs ? The last 20 years has been defined by ever increasing capital spending on fixed assets while reducing the cost of staffing and operations. The “spend-more capital” strategy of the last 20 years hasn’t produced the returns or operational benefits that CIO’s are expecting and disenchantment with the current mode of working is widespread. Gartner has reported that 30% of projects over $1MM fail to come in on budget and on time, or to deliver the original business promise and the failure is largely due to lack of organisational resources.
Cloud companies have shown that deep investments in people greatly reduces CapEx and improves service agility, density and visibility. Human infrastructure costs money but often orders of magnitude less than a new network or storage array.
- IT is over-spending on capital purchases and failing to leverage the value of existing IT assets
- The success of DevOps shows that over-investment in human infrastructure produces real increases in productivity through transformation
- Cloud companies use large investments in OpEx, specifically people, to increase efficiency and utilisation of infrastructure.
- Enterprise IT should plan to grow head count instead of capital spending to increase utilisation and financial efficiency.
Capital Theory Disproved By Cloud
The purchase promise offered by existing IT vendors is that increasing Capex will reduce the cost of ownership through faster performance, better features, better software and usability. Yet the public cloud uses conventionally ‘over staffed’ IT teams creating automation and orchestration that reduces capital spend up to 70%. Modest investments in human infrastructure instead of physical or software has built an entire product category that is growing in excess of 40% per annum.
In simple terms, public clouds must go to extreme lengths to reduce capital spending to improve the operating cash flow. Capital reduction practices such as whitebox switches and servers, distributed software and new designs on power and cooling have reduced the up front spend to the point that scaling compute is possible and practical business.
Operating clouds at scale has driven an industry wide focus on operational efficiency in capital. The fast pace of growth combined with investment ahead of customer revenues (build it before they come) created the business need and then simple investment in human infrastructure created a solution. This solution has manifested in two key areas – reducing capex through smarter purchasing and increasing human infrastructure costs to increase utilisation and chargeable loads in plant.
This increase in OpEx is already having measurable market impact by slowing revenue and profit growth of large-cap companies such EMC, Cisco and IBM.
Business Efficiency from OpEx
The success of this approach may be measured in the business efficiency of cloud companies. Over-investing in human opex has led to transformational change in sustaining costs for an infrastructure through programmatic/software operation of existing assets and technology.
Sustaining costs such as power, space and compute capacity are controlled (at best) by increasing the fundamental utilisation of the infrastructure through smarter use of software increasing the utilisation of existing assets. Where an Enterprise data centre has less than 20% total utilisation today, the private cloud will drive this towards 50%.
Investment Protection as Insurance on Over-Capitalisation
There are a number of canaries on over-spending on capital but perhaps the clearest is the perceived need for “investment protection”. The intent is to assure customers that large capital purchases will offer value in the following areas:
- support for many upgrades over time
- support for in-place, non-destructive, non-disruptive upgrades
- large service organisation to warehouse and distribute custom/proprietary parts in addition to specific skills to install said custom parts.
All of these factors are intended to assure the customers that they will earn a return on a large capital purchase but ignores human factor in device operations. New platforms and hardware with more human intervention and operation could achieve higher utilisation.
Self Interest and False Messaging
Incumbent vendors strongly promote messages of high capital spend with strong self-interest to their own business models. Boosting vendor cash flows while customers bear the risk and costs of achieving ROI through effective operations. Why are customers required to invest heavily before earning return on capital deployed ?
Two factors indicate customer discontent. First, ‘false messaging’ around high cost assets worked until cloud companies become successful and showed that low capital/high opex infrastructures are viable. Cloud technology is now proven as a low capital business model.
The second factor is that customers are communicating better. CIOs, designers and architects are able to gather more information from Internet, social media and conferences to discuss and review the new ways.
DevOps is Opex Heavy
The rapid success of DevOps movement is a signal that plentiful supply of human infrastructure can create huge value by applying head count to reducing operational inefficiencies.
The DevOps movement is the strongest proof point that investments in Human Infrastructure produce huge dividends. Many call this a “movement” and laud it as some “revolution” while ignoring the underlying purpose of extracting value from over-capitalisation in fixed assets.
The top priority for 2014 according to Harvey Nash CIO Survey 2014 is staff retention.
At a simpler level, the current levels of understaffing in IT teams lead to constant firefighting that leaves no energy, time or motivation to conduct fire prevention.
When technical staff gather the stories of operational exhaustion with low levels of motivation, reducing competency and reducing skill development are all too common. IT staff are consistently discussing burnout due to lack of time, training and simple rest between firefights.
The Information Technology Burnout Project shows just how bad this issue can be.
This article is continued in a Human Infrastructure Poverty In the Enterprise – Part 2 where I consider Open Source, Outsourcing design to vendors, Skills Creation & Expertise and more.