Microsperience

“BSSOSS market to grow 200% in next 12 months.” Teresa Cottam looks at market forecasting.

BSSOSS market to grow 200% in next 12 months. Got your attention didn’t I? And you see that’s the problem with forecasts – we all want to see a big number and a graph rising steeply to the right.

I don’t do forecasting, so I’ll confess that before I begin. My market prediction was just designed to get your attention. And that, of course, is the problem.

I do understand the need for forecasts. Before a firm invests in a new opportunity, it’s useful to know if there’s any potential there. They’re needed for business cases, for valuations and so on.

That’s all well and good. But I think sometimes we get carried away with it all, as though the numbers were real rather than indicative, and not just a good (and sometimes not so good) guess about what might happen.

Some forecasts can be so wildly inaccurate that I remember one incident – oh a good 10+ years ago – where the sub-editor had mistakenly written billions rather than millions and no-one questioned the figures – not ever. The mistake was only picked up years later when an update of the report was prepared and it was realised the market potential had been exponentially over-egged. Nowadays, some forecasts are like yo-yos, constantly being revised up and down to the point where you start to feel sea sick.

2 + 2 + 2 does not equal 100 billion

It’s not forecasting itself, per se, that’s the problem though: it’s the way we interpret and use forecasts. It’s also our emotional reaction to them. I’ve often heard vendors tell me that they don’t read the report just the charts. But we have to become more educated about how we view them, and learn to interpret them rather than just mindlessly repeat them. We also have to question them more.

My reason for writing this is that I am seeing some very bullish forecasts for new areas of software tech – Cloud being one of these that springs to mind. I’ve been making the point a lot recently when people talk about “robust growth”, that this is just the opinion of a small team of analysts – usually one or two – who might very well be wrong. I’ve asked where the money is going to come from to build these multi-billion dollar markets, and pointed out that there must be a substitution effect going on because all indicators suggest there’s no new money (at least in telecoms IT). That’s before we get into the nitty gritty of the detail of what’s actually being counted.

We hear a lot of talk about running leaner. About getting our IT costs down. Then we all get so very excited when someone runs out a forecast predicting some software domain is going to grow rapidly. Look at that the other way round: it means somewhere costs are going up. Economics suggests these new costs will have to be paid for.

I can’t see many CFOs opening their purses and throwing money at software like it’s 1999, so if these forecasts are right then we first need to free up some cash from somewhere else. If the budget and market is level or declining overall, then the logical conclusion is that if one sector sees money flowing into it, another will see money flowing out. If you spend on Cloud, then you’re not spending on hardware. If you spend on social CRM then you need to pay for that by spending less on other parts of CRM or by creating efficiencies somewhere else.

This is where one of the big errors creeps in – valuations for vendors are based on their presence in a growing market. Often financial analysts miss the subtle fact they’re also a player in a sector that’s declining or will be substituted. On that point, I sometimes read financial analysts’ interpretation of markets I know well and I simply don’t recognise them. Vendors need forecasts to show a steep gradient to the right, so the problem is no-one’s interested in forecasts that show decline or stagnation – there’s no market for them. I’ve seen the sales figures: bullish forecasts sell. This encourages a lack of honesty, or at the very least a tendency to be particularly optimistic.

Substitution effects can be real uplifts to IT, but are often less than anticipated

Of course there’s real growth. For example, some CSPs are replacing bespoke or home-grown solutions with COTS or SaaS. So the market, depending on how you describe it, may appear to be growing – in fact, there’s a substitution effect going on. Usually there’s also an expectation that money will be saved. The same is true when we “automate” – one key motivation is to reduce people costs, so we’re substituting software spend for wages. However, there’s also an expectation from the business of reducing spending when you do this – I’m yet to see a case where savings from say wages are pumped fully back into IT. The business needs to announce that it’s saved money.

Likewise the money to fund new initiatives is often “trapped” in existing commitments. Releasing this cash takes longer than many analysts seem to understand and is not a trivial undertaking. Many CSPs still have some way to go before they take the title “slimmer of the year” and claim a “cellulite free” back-end.

However, while there’s still scope to drive more cost out of IT in order to finance lower prices while maintaining margins, and to finance new initiatives, there’s a natural software and hardware renewal cycle, and I haven’t yet seen much evidence that CSPs are prematurely chucking away software or hardware to finance the much-desired modern infrastructure. Isn’t the money for renewal still largely locked up in legacy? Isn’t the lifespan of solutions actually expanding rather than contracting? Isn’t the evidence that while you can modernise it’s not easy, and requires investment – ie cash – in order to do so. Aren’t many CSPs actually frustrated by their inability to switch off legacy? Even those that have managed to do this have spent an awful lot of money, time and effort doing so that belies that flick to the right we see in a typical forecast.

These are the reasons why actual adoption patterns often don’t fit the steep curve of the forecast.

I haven’t yet found the money tree

Please let me know if there’s a secret BSSOSS money tree that telecoms CFOs are using to finance some great big software investments. I’m still hearing that budgets are essentially flat and certainly down on what they were at the peak of the market. The credit crunch means finance is still hard to come by.

Constrained IT budgets are nothing new either – all I’ve ever heard in my career is that IT budgets are tight and CSPs need to save money. Living in the UK I’d originally assumed that this was a factor of our incumbent moving from being a public-owned utility to a privately-owned competitive telco, coupled with a regulator who’s driven down prices (and therefore margins). As prices have decreased for basic services, so the need to run leaner has increased. The credit crunch has magnified this effect because prices are now also being constrained by disposable income, which is down.

One of the very real challenges we face is having to do more – help the business reduce costs elsewhere, drive through operational efficiency, innovate and so on – with less (ie without any corresponding increase in IT budget).

I’m not saying that there won’t be growth in some IT areas, I’m just trying to point out that forecasting often produces a distorted picture of the market. The reality is you don’t always need to spend more to achieve your objectives – you can spend smarter. Of course the bad news is that although I keep looking, to date I still haven’t found the secret BSSOSS money tree.

What do you think about bullish forecasts? Do you think we’re going to see budget inflation any time soon? Do you know where the secret BSSOSS money tree is?