Harold Wilson, who was Prime Minister of the United Kingdom for parts of the 1960s and 1970s, coined the phrase, “A week is a long time in politics”. I was reminded of that phrase when I read a post just now by David Pogue, the New York Times’ tech columnist about a new book about Microsoft by Mary Jo Foley. Foley posted about the key issue on her own blog as follows:

As I mentioned in the conclusion of Microsoft 2.0, I had just submitted the final version of my book manuscript a week before Microsoft announced its $44 billion bid to buy Yahoo.

Disbelief was followed by utter despair — and not just on Yahoo CEO Jerry Yang’s part. All I could think on February 1 was I was going to have to go back and revise every single one of my 300-plus pages.

I did go back in and update my chapters to reflect the possibility Microsoft might end up buying Yahoo. Then I revised again to say Microsoft did buy Yahoo (given that much of the press in February made it sound like it was pretty much a done deal). Right before my drop-dead go-to-printer date, I revised one last time, saying that Microsoft might or might not buy Yahoo.

Well, as we now know, on May 3, Microsoft withdrew its takeover bid, after being unwilling to meet the higher per-share price that the Yahoo board was demanding.

So this poor woman completed her book on Microsoft, then the Yahoo! bid was announced, she made a bunch of changes to incorporate the seemingly inevitable acquisition and submitted the final version, and then Microsoft called the whole thing off. What a miserable experience, and presumably one which will greatly damage sales of her book.

But all this makes me wonder how much it’s really possible to predict the future in the world of technology. If a week is a long time in politics, it can sometimes be an eon in tech. We’re being asked now for our research publication plans for 2009. The year won’t even start for another seven months, and won’t end for 19 months, and yet we’re supposed to predict the broad outlines of what we’ll publish that far ahead. I just can’t imagine that we’ll be able to do an accurate job of forecasting what’s we’ll publish in late 2009, and yet clients will not doubt want to hold us to at least some of it regardless of whether it’s the most relevant or interesting research to be publishing that far down the road.

Of course, we publish forecasts with a five-year time horizon and generally think we have a good handle on future trends. And in terms of Internet penetration, or wireless subscribers, or MPLS ports, that’s actually fairly straightforward to do. Occasionally, we might buy into the hype around a new product or service too much or underestimate the growth in an unexpectedly hot market. But on the whole those long-term product and service trends are relatively straightforward. They tend to grow in a steady fashion after they cross the famous chasm and so are relatively easy to predict with a reasonable degree of accuracy.

But predicting industry news, actions of specific players and especially mergers and acquisitions is much more art than science, and would take a real crystal ball and not the analyst’s metaphorical one to foretell accurately. This is why it’s important in our job as analysts to separate one from the other. We can still do the forecasting bit, and we can certainly talk about whether it would be wise for a particular company to pursue a certain course. But we shouldn’t really be in the business of predicting decisions or performance by a particular company unless we have real inside information (in which case we probably shouldn’t anyway).

We all have a lot of work to do in providing balance between these various things that we do, but we also need to be honest with our audiences about what we can reasonably do and what we should leave to others more willing to peddle their fortune telling skills.