IT: – Still the Black Hole of the Balance Sheet?” was the title of an interesting media roundtable organised by Managed Objects.
It discussed measuring the cost/benefit of IT in the business, although a research survey (of almost 300 IT and business managers) commissioned by Managed Objects as a basis for the discussion, appeared to concentrate mostly on the cost side of the equation and the granularity with which costs could be measured.
Jim White, Business Technologist at Managed Objects, summarised its research findings as saying that just over half of the IT managers surveyed were happy with the way that IT controlled costs - but a similar proportion of business managers felt that cost control by IT was only “somewhat effective”. Since the business pays IT salaries, “could do better” is a worrying rating.
The survey also indicated some concern with the granularity of IT costings and a need to cost individual IT services, not just measure the overall cost of technology.
But, does the measured “cost” of an IT service have any real meaning, except in relation to a measured “value” delivered by the IT service?
The first UK application I was involved with enabled a Bank to run its back office centrally in London, instead of computerising and integrating semi-automated back offices throughout Europe. The value delivered by this at the time was pretty obvious at the macro level – and it rendered the cost of the mainframe involved almost irrelevant. On the other hand, few dotcom initiatives in the 1990s made any money, so subsidising them heavily from parts of the business that actually made a profit made no sense - unless the predicted value of these dotcom systems could be measured and monitored; and costs kept in line with the business value from the dotcom services when (or if) it was delivered.
Value of IT
The “value of IT” is something that should concern all of us as developers – because if our paymasters can’t easily measure the value delivered by the IT systems or services we develop, they may well underestimate it (nearly half of the respondents to the Managed Objects survey thought that their companies would spend too little on IT in 2006). The hard costs of IT seem obvious to the people signing the cheques and will tend to weigh heavier in the balance than any nebulous value delivered, when boardroom discussions turn to cost-cutting and outsourcing.
On the other hand, IT must have a value, even today when the “low hanging fruit” (like the bank back office) have mostly been plucked, since the easiest way of reducing the cost of IT to zero is to stop using it; and I don’t see many organisations replacing serried rows of PCs with serried rows of clerks… It is just hard to put a unarguable monetary figure on IT’s value to the organisation, although most of us accept that, after a certain point, cutting costs can reduce the value delivered to nothing. .
Back at the roundtable, Will Cappelli, Research VP at Gartner, thought that IT executives, at least, usually appreciated the problems of costing IT; but that allocating costs to individual services was non-trivial, even now.
However, when I asked him why Gartner, following on from its success with its somewhat simplistic but apparently useful TCO (Total Cost of Ownership) models hadn’t followed up with a Total Value of Ownership (TVO) model, he was less than reassuring. Apparently, the problem is too hard and, although there are value models which could be used, they are academically controversial and complicated to use. “We know,” he said, “where value is [qualitatively] but not what it is [quantitatively]….”
[Nevertheless, Gartner research does work on the general business value of IT, see here, for example; and even has a “Total Value of Opportunity” methodology, although this seems to be more to do with investment planning than with the value being delivered by operational systems. Forrester has something similar, in its Total Economic Impact (TEI) framework, originally from the Giga Information Group. Neither Gartner’s TVO nor Forrester’s TEI seem to feature as much as TCO in IT investment discussions today, although you will see them quoted on occasion and they do bring value into the discussion]
Cappelli did suggest that Real Options Pricing might be our best bet for assessing the hard value delivered by IT but didn’t go into the details of this. An academic paper making “A case for using real options pricing analysis to evaluate information technology project investments” (Information Systems Research, Vol. 10, No. 1, pp. 70-86, 1999), which you can find here, will give readers some idea of what we’re talking about, but it certainly ain’t easy going, and seems to relate more to predicting the future value of investment than to monitoring the value actually being delivered by operational systems.
Even in the simple case where the IT investment directly relates to a discrete business opportunity, the meaning of “the value of IT” isn’t exactly obvious. If you invest £100k in a computer, with its own staff and premises, to run a new Widget processing service that delivers £100k per annum profit, to come up with a “value of IT” you presumably have to subtract the value that that £100k investment could deliver simply by being invested in a safe security, discount the value of future money, subtract some value due to the increased business risks from being involved in widget processing - and decide just how much of the “value” apparently delivered by your new computer really comes from the business acumen of the staff using it anyway. [Caution: although I have worked for a Bureau of Transport Economics in a previous life, IANAE – I Am Not An Economist – so please don’t bother writing in to tell me this; unless you can provide a proper, and plausible, model for value delivery from operational IT services at the same time <g>]. If your new IT system shares resources with other systems and delivers value to a range of services/stakeholders, the problem of assessing the value it delivers is that much more complicated.
However, even a simple metric for value would be better than no metric at all. If I put a number on the value of IT to a service, stakeholders will form an opinion of whether my metric is placed too high, too low or about right. From that, we can eventually agree on a value metric – and if this not commensurate with the TCO of the technology involved, we can consider investing further (to exploit the value available from the service) or reducing costs (if value isn’t being delivered). But we must have a hard value metric (even if only one at the same level as Gartner’s simple TCO metric), which can be refined and improved; not just a soft opinion. And this implies some sort of automated algorithm, embedded in a tool; I just don’t see the average IT shop employing trained economists to undertake complex value assessments.
Personally, I suspect that the problem with routine “value of operational IT” metrics may be partly that, while TCO is a useful weapon in the vendor marketroid’s armoury, a “TVO” metric for value might be a two-edged sword. After all, it might show that upgrading to the next release of a technology with no extra functionality that you actually need doesn’t deliver any real value.
I remember when Microsoft started using TCO as a selling point for Microsoft Enterprise technologies, asking about the corresponding “value of ownership” for these technologies and being told that it was undoubtedly very high indeed – although Microsoft couldn’t actually put a figure on it. It was waiting for someone like Gartner to come up with a TVO model… Nevertheless, IBM Global Services can apparently offer to help IT executives “quantify and communicate the value of IT”, although some might feel it could have a vested interest here, and this doesn’t seem to involve a simple value metric corresponding to TCO anyway.
It seems strange to me that IT Governance initiatives (which encompass IT cost/benefit and value analysis) often seem to be driven by the vendor community as an opportunity to sell products rather than by the customer community that really needs them. [Perhaps ITIL is the honourable exception.] Some of the IT Governance products are very fine, no doubt; and some vendors almost altruistic, perhaps; but relying on this seems somewhat risky, from a business perspective. A Value of Ownership metric would be a useful weapon in the armoury of a customer – I wonder whether customers will eventually get together to sponsor a simple industry-wide TVO metric?
At the roundtable, Simon Clark, Partner at Fidelity Ventures (a Venture capitalist) provided an interesting sidelight on all this, since these days you can’t assess a company as an investment possibility purely on its real-estate, inventory, premises and investments etc. You need to value its “intellectual property”; and some of that is represented as metadata in repositories, program logic, configuration management databases etc – the value of IT again (although a lot of Intellectual Property will remain in people’s heads, at least until ice-skating in Hades becomes popular). And, again, there seems to be no straightforward value metric. The venture capitalists simply recruit retiring IT directors to help them make an assessment of whether the IT in a potential investment target is up to scratch – or even the major part of the investment opportunity.
There was considerable discussion at the roundtable around these and other points, but this rather confirmed the general view resulting from Managed Objects’ research: that people are happy with macro-level costing of IT, but that there is a demand for micro-level costing by service and department, which is being developed now; while usable value metrics, are both needed and some way off. Jim Knight summed up the situation neatly by saying that although we are close to delivering on service-based costing, we are just starting on service-based value. ®
David Norfolk is the author of IT Governance, published by Thorogood. More details here.