InformationWeek

Slips and Misses: How Much Do Product Delays Really Matter?

By Stuart J. Johnston
May 5, 1997

A few weeks ago, the world was treated to the "shocking" n ews that the "Memphis" release of Windows 95, sometimes referred to as Windows 97, is not going to ship this year.

Despite the hoopla, users have indicated that a slipped ship date, even for an update of Windows, is of less importance than many other pressing issues they face. Corporate IT managers are just too darned busy deploying Windows 95 and doing other aspects of their jobs to want to stop midstream and deploy a major update to their desktop operating system.

A recent InformationWeek survey of 151 IT managers underscores how truly disinterested IS managers are in missed ship dates. Of the IT managers interviewed, only three of them (or about 2%) said not getting Memphis this year would have a major impact on their plans. And while only 27% are already finished deploying Windows 95, fully 76% will deploy Windows 95 on the bulk of their desktops before they do anything else. Just nine respondents, or 8%, said they're going to stay with Windows 3.x for now. Finally, 39% will skip Memphis altogether and go to Windows NT Workstation.

Given the issues facing IS executives, the results are not surprising. Software updates these days come like rain; and like rain, it takes a while for them to soak into the user base. Corporations have a large investment in desktop machines and an even larger investment in administration and management of those PCs. In fact, despite Microsoft's big push to get corporations to adopt Windows 95 even before it shipped, the majority of them are just in the process of deploying it now. No surprise there. Slow, cautious adoption following many months of careful evaluation is a hallmark of "best practices" in corporate IT departments.

It was the slowness of that corporate process that prompted many pundits a year ago to pronounce Windows 95 dead on arrival. However, the software has an installed base in excess of 60 million and a ship rate of more than 4 million new copies per month.

An additional factor, this time around, is that Windows NT 4.0, which gives NT Workstation and Server the same basic user interface as Win95, shipped just last summer. That finally enabled companies to have either of two 32-bit desktop operating systems with the same user interface, and some commonality of support tasks, throughout the enterprise.

Therefore, it should be no surprise that only 10% of the IT managers in the InformationWeek survey found the addition of the Active Desktop -- which changes the user interface once again -- the most interesting new element of Memphis. The last thing many of them want to do is to standardize on a single user interface and then suddenly go back to supporting two.

A top IT manager at a company that is well along in moving its entire enterprise of 30,000 PCs to Win95 recently told me he is concerned about Memphis slipping but went on to say that initially the company would not be deploying the Active Desktop so as to avoid supporting two interfaces.

So, although Memphis is indeed most likely delayed until 1998 -- for most users a nyway -- the news is not as simple as that. What Microsoft said is that it had notified PC manufacturers that it would not be able to get them "golden" code in time for them to pre-install Memphis on machines intended for the retail market at Christmas. The Christmas deadline works out to be about the end of July. Five months would still remain in the year after that missed date.

Now, if Microsoft chooses, the company can pretty easily see that new PCs come with a coupon for a free or low-cost upgrade. But notice that we are talking about the consumer marketplace here. Microsoft clearly demonstrated last fall that it can ship final software to its critical corporate customers late in the year, even if it can't have the product in the retail channel in time for Christmas. Office 97 went to manufacturing in late November, and it didn't officially ship for retail availability until mid-January.

But those corporations that needed Office 97 got copies of the CD-ROMs in December so they could begin deploy ing immediately. If Microsoft finishes beta testing Memphis before the year is out, it can always send CDs to those critical customers. That doesn't mean that the company will finish in time, but if Microsoft does, it certainly won't take five months for the company to get the operating system to users.

There are also other historical data points we can refer back to that bear this out. Two and a half years ago there was a real hullabaloo over the fact that final shipment of Windows 95 had slipped into mid-1995. By the time it did finally ship, the conventional wisdom was that it was more than a year late. Perceptions were different than reality, however.

The facts are that when it shipped on Aug. 24, 1995, Win95 was actually only nine months later than what Microsoft had originally promised in April of 1993, when Bill Gates first spoke publicly about what he referred to then as Windows 4.0.

Were some corporate customers hurt by that nine-month schedule slide? Some, surely -- especially thos e that like being on the bleeding edge. But as you can see from the Win95 adoption pattern among corporations to date and the latest survey results, corporate IT departments are, by and large, a careful lot that tend to move deliberately and aim to avoid irreversible errors.

In fact, as we start to live more and more in Web years and the pace of change accelerates geometrically, it also appears that the number of PCs out there has risen to the point that, from both a logistical and budgetary standpoint, the deployment speed for operating systems has reached some kind of coefficient of friction that impedes it from going any faster.

Only time will tell if network computers or NetPCs will reduce that friction and speed the process further. And it certainly won't be boring as we wait to find out. But I suspect that there is some kind of human intellectual barrier that limits how quickly we can accept ever-faster rates of change and that will be the ultimate determinant of when the slippage of an opera ting system delivery schedule is important. That is, we may yet live to heave a sigh of relief when a slip occurs.

This latest slippage itself should not be too surprising given that it happens with most major products from most major vendors, unless the vendor is cagey about never promising a release date in advance. And slippages certainly have happened frequently enough with Microsoft operating systems that we should not be surprised at all when they do occur.

Return to AuthorITies or view AuthorITies archives

Comments?

http://www.informationweek.com