Outsourcers are a bit like private equity. They make money by taking something over, ruthlessly and rigorously standardising everything about, and making money on the difference between the price they agreed in advance and the costs they actually incur post-cleanup.
Nothing wrong with that. It’s a business, like any other, providing something the customer wants, at a price the customer is prepared to pay, and usually one that is profitable for the provider as well.
It’s the same sort of thinking that pervaded firms when they had to decide what to own and what to buy in. Why firms used to focus on deep vertical integration.
It keeps the costs down.
Which is a good thing, isn’t it, keeping costs down?
Firms couldn’t be wrong. Private equity operators couldn’t be wrong. Outsourcers couldn’t be wrong.
Vertical integration coupled with rigorous standardisation was the way to go.
So the thinking was applied to systems as well. Integrate everything in sight, as deeply as possible. Standardise everything within the resultant monolith. Reduce transaction costs as a result.
Yup, makes sense.
Provided.
And it’s a big “provided”.
Provided we’re talking about steady-state. No changes.
In all these models, transaction costs are kept low because nothing is allowed to change. The firms of the early 20th century could mandate it. PE companies could enforce it. Outsourcers could demand it. Usually by deploying a simple mechanism: high tariffs for any changes.
And so it is with monolithic vertically integrated systems. Good for steady state.
But very expensive to change.
The world we live in is a world of change. Demanded change. Mandated change. Customer-driven change.
Steady-state transaction costs can be controlled, of course. But only in a steady-state world.
What is now needed is a means of managing transaction costs in an ever-changing world.
And that’s where the cloud comes in. Real cloud, not the zero-sum-game “private” porcine make-up.
There was probably a king on the throne of the United Kingdom the first time the phrase “keep fixed costs as low as possible, and let variable costs be driven by the market” was used. Keeping fixed costs low makes sense. Because the cost of change is usually the cost of changing something in the fixed cost bucket. Variable costs, by their very nature, vary according to volume.
Imagine you bought a fixed-price ticket to Paris. And, sometime after you bought it, the demand drivers and environment had changed, and you had to go to New York instead. Would you say to yourself “Well, I bought this ticket to go to Paris, and that’s what I will do, regardless of where I’m meant to be”? Or would you berate yourself for buying a fixed-fare no-changes low-cost ticket while knowing that things might change?
Sometimes I’ve heard IT people talk like that. “I have to try and get value from what I bought, even if it is of no use to me”. I guess this is part of what gets called Sunk Cost Fallacy.
I’m going to continue to think about this, just wanted to air some of the thoughts, especially since some of them are thirty years old.
But before I end this post, I wanted to bring up two more things: meeting scheduling; innovation. Because I think this fixed cost/variable cost dichotomy plays out well in those areas as well.
Take scheduling meetings. How many times have you been in a situation where it’s really important to get a particular bunch of people together, and the “system” spurts out the 12th of Never as the next available date? A strategy that relies on agent negotiations between Outlook calendars….
I’ve always considered this a strange phenomenon. If it’s important to keep fixed costs low, and if human time/salary is one of the biggest costs, then why pre-commit a high proportion of the time? Isn’t that creating a new fixed cost? What I tend to do is to hold on jealously to white space in my schedule for as long as possible, so that I can respond to changes in demand quickly and effectively. Of course there are some regular meetings, these have to be scheduled in advance. When someone plans a visit, meetings have to be scheduled in advance, to “lock” the event in…..otherwise the cost of travel should not be undertaken. But for most of the time, it’s best to keep the swing room. Knowledge work is lumpy. Peaks and troughs.
Similarly, let’s think of innovation, particularly innovation with a technological component. One possible barrier to innovation is the corporate belief in Sunk Cost Fallacy. The space to innovate is taken up by the long-term commitments within the fixed cost bucket. An inspection of the ratio between fixed and variable costs for each part of the estate may provide a proxy for the innovation capability of that part of the estate. Obviously it would depend on the cycle of investment the firm is in, the market conditions (steady-state or fast-changing), the market itself (a manufacturer of oil tankers is going to have somewhat different characteristics from a retail bank), a whole slew of exogenous and endogenous variables.
But despite all that, I am intrigued by the possibility that something within the fixed/variable ratio may be a simple proxy for the cost of change, and therefore for some part of the capacity to innovate.
Views?
Spotting the proxy may not be so simple: Facebook – shifting fixed/variable ratios from costly Oracle licences to FOSS. Salesforce.com – shifting application building from IT dept’s to business domain experts.
Cloud gives an inkling by shifting fixed costs to variable allows change to adaptive business models and Growth.
Isn’t cloud at it’s heart just the ruthless standardisation of IT?
Don’t forget the tie that we get between sunk cost and political capital – the ‘I invested my career in this so we have to make it work’ syndrome.
Some very interesting things here.
(1) Change states: I think you might be interested in the topic of Dynamic Capabilities (i.e. strategic capabilities as being in a state of flux, and evolving from state to state. Thus, the ability to identify, and obtain capabilities is the important thing? – This being in contrast to the Porterian view of Value Chains)
(2) Standardisation – Differentiation and Environmental Uncertainty (Complexity, Uncertainty, and Ambiguity all being different) and are highly correlated with “person based characteristics” such as tolerance for ambiguity and history of personal experience combined with other areas such as “ability to sense” those environmental elements ( I think this ties to your thoughts on filtering by the way).
(3) Labour as Allocation: I think that various Entrepreneurs have advised start up CEO’s to not “over book” their time as it decreases their ability to respond to day to day changes. It might also tie to actual time management competency differences between start up’s and blue chips. In blue chip the Managers efficiency is based on their ability to manage the access of others to their time, in start up land, it’s largely the opposite. This might indicate that Start up’s and blue chip employees operate in different types of networks, and need different network competencies.
JP,
Insightful post!. In my experience, any new technology investment is driven by the budget rituals of the organization. The problem with Cloud is that it doesn’t subject itself merely as an investment candidate. Rather, it challenges the budgeting principles themselves. That’s not easy! :-)
Coming to Fixed Vs Variable costs, my view is that, the objective of general business management is to promise and deliver stability despite dynamics in the ecosystem. And, especially markets (wall street) demand predictability. Volatility is something shareholders don’t like to hear!. If a business needs to commit on certainity and meet its projected quarter-on-quarter forecasts, it needs to operate in fixed costs paradigm!. It needs to fight out the volatality to deliver the static/promised results!.
For IT management, its even more tougher – CFOs/CEOs write a static cheque (“x” amount of dollars) and they want CIOs to make x% of savings every year. CIOs can be adventurous and employ Cloud to manage variability and eventually they may make some savings as well. But, the ceiling is fixed by CEO/CFOs. The IT spending cannot exceed beyond that static/fixed budget.
Another reason for not going to Cloud or even Private Cloud for that matter is, having it on-premises gives a sense of control to CIOs and gives them the ability to manage the commitments and expenses by making drastic measures without depending on any external entity!.