If the end-users of a business system are not convinced by the merits of a new approach, IT directors should understand that there is usually little point in forcing it on them.
Yet when Lloyd's of London called time on its electronic trading platform last month, many in the insurance industry simply shrugged their shoulders at the inevitability of the failure.
The Kinnect platform was, they argued, an idea before its time that had not been embraced by the brokers and underwriters working within Lloyd's because it was not quite what was needed.
Those same commentators were also bullish in their insistence that a suitable commercial alternative lies just around the corner to help ease the industry's path towards achieving certainty on the details of insurance contracts more quickly, in line with Financial Services Authority demands.
But, at £70m, Kinnect was a costly and time-consuming failure for Lloyd's, and there are clear lessons to be learnt from its downfall.
In common with so many failed major IT projects, the Kinnect programme died a slow and very public death because those charged with implementing it failed to get sufficient stakeholder buy-in from the beginning. This, in turn, failed to bring about the necessary change of culture at Lloyd's to give the platform a fighting chance.
No doubt Lloyd's consulted widely both before and during the development of Kinnect in an effort to ensure it was on the right track, but given the scale of the failure, it must surely have been asking the wrong questions to the wrong people.
Judging by Kinnect's history, the truism that there are no IT projects, only business projects, looks like it may not have found its way into One Lime Street.
Even if the IT directors of the businesses in Lloyd's were happy with the development of Kinnect at various points along the way, very few of the businesses themselves were similarly convinced. When Kinnect's closure was confirmed last week, only about one in 10 of Lloyd's businesses were signed up to the platform, and even fewer were thought to be using it in earnest to drive efficiencies.
This lack of buy-in also affected the platform itself, which was, in the words of one consultant, "a horse designed by committee". In a complex market of competing interests, it is not surprising that it ended up pleasing no one.
IT directors in all areas of business may be increasingly inclined to describe themselves as change agents, but there are still too few paying enough attention to the wider business and user agenda when projects are in development.