Corporate catastrophe follows botched beta

Early in 2000 I signed on as an engineering associate with a West Coast software vendor I’ll call “Port Portal.” The guy who hired me, “Alan,” was CTO and VP of Engineering. As it happens, he had just been hired himself; before taking the job he’d been a divisional CIO at a local telecom company.

At this point, Port was riding high on the surging wave of interest in enterprise information portals. In fact, the company had just executed a successful IPO and moved into a spanking new set of corporate offices. The new CTO had convinced the rest of the management team that we were positioned to dominate the market within two years. We were going to crush the competition with innovative new products and brilliant business strategies.

Because the company was now public, we had to address Wall Street’s concerns about hitting our revenue numbers, and Alan drafted me to help prepare a presentation for some deep-pocketed financiers. To my surprise, halfway through the presentation, he boldly announced that we would be releasing the next major version of our software in June of that year — less than six months away.

This was not only the first time I’d heard anything about new software, it was apparently some pretty powerful code since, as our CTO explained, it was based on advanced architecture that would enable customers to plug virtually any information source they wanted into the portal. News to me.

I joined a meeting in Alan’s office the next morning, looking forward to hearing how we were gonna pull off this engineering coup in less than half a year. Alan explained that we’d save months of development time because the software would be released without a beta test. Yikes! When I expressed my concern, he assured me that the engineering team he’d brought with him from his old telecom had a robust alpha testing process, and the application wouldn’t be released until it was fully checked out.

OK, I said, but why not run a beta test too, just to be sure? Alan looked at me like I was a dim bulb. If we took time for a beta, he pointed out, we’d miss the June release date he’d promised Wall St. Duh.

We shipped the software on time, and to my relief, initial reviews were positive. But within a month, users were reporting serious reliability and integrity issues. One customer, with an enterprise portal intended to support thousands of concurrent users, reported that the system crashed if more than 100 users were on it at the same time. Two weeks later one of our biggest customers, a multi-billion dollar bank, demanded its money back.

Our competitors got wind of our problems, analysts refused to recommend us, and customers were furious. Worse yet, sales leads dried up and cash flow went negative. The company wound up taking in less than half the revenue it had projected for the quarter and was forced to lay off 30 percent of the staff. Despite our desperate efforts to retrofit the code, sales and stock prices plunged, and Port was eventually sold for pennies per share.

The only good news was that Alan was fired in the first wave of layoffs. Apparently somebody had told upper management about the alpha-beta business. It might even have been me.