Should Pete Tierney, CEO, chairman and president of struggling Case Based Reasoning (CBR) vendor Inference Corp, of Novato, California, quit? Yes, according to some nettled investors in the company, whose stock has been in the doldrums since last November, when it came off its 52 week high of $21 to sink as low as just over $3.50 (it has since climbed up slightly, to $4.625). The company seems to have been outflanked by nimbler competitors in the help desk business and is in the middle of a messy transition to a new strategy, a shift Tierney says he identified and is leading, but the firm has no date for when its fortunes are likely to go into a positive upswing once again. What makes this question so ironic is that Tierney deserves credit for turning the company around at least twice already – once when he arrived seven years ago and two years ago when he helped take Inference public. He joined Inference from a senior sales role at Oracle in 1990, when it was solely a vendor of premium Artificial Intelligence development software, and helped guide it to a more commercial focus for underlying technology in so-called CBR, a way of using a database of past examples to allow a computer to infer a solution to a given problem. Eventually Tierney canned the AI side altogether (it spun off Brightware in May of 1995) and launched Inference as a provider of help desk solution, albeit using CBR as the underlying engine. At first his gamble paid off. For its first year as a public company (its IPO was June 1995) Inference reported a profit of $3.8m on a revenue of $29.4m in the year to January 31. For fiscal 1996 the good news continued; for calendar 1996, announced in mid-February, Inference saw revenues up 22% on 1995 to $36.0m, with net profit of $2.4m, although that was down from the previous year’s $3.8m. Though 1996 as a whole was a good one, it was in the third quarter – when the stock price went stinky – was also the time the challenges to continued success began to make themselves felt. That’s when it lost control of the ball, acknowledges Tierney. Though there was some recovery in the fourth quarter the last six months have seen Inference lose sales and ground. In its first quarter revenues were down 22% to $7.1m from $9.1m same quarter in 1996, with a net loss of $943,000 compared to net income of $753,000 a year before. Bad enough – but the second quarter saw sales plunging 28% from the same period last year, from $9.4m to $6.7m, with net loss of $906,000 compared to $864,000 profit for the same period in 1996. American sales in particular did badly, caving in 39%, from $6.2m to $3.8m. What went wrong? Tierney argues that customers have been preferring to buy what they consider lower cost alternatives which require less technology commitment from more traditional help desk mavens, who are ironically often Inference partners, such as Scopus Technology Inc and Vantive Corp. Inference has decided that its future lies in avoiding a costly battle of attrition with these guys, instead switching to what it dubs self-service help desks on Web sites, such as its one for games player Broderbund Software Inc (CI No 2,910). Inference says it has 70 customers in this space already, including Cisco Systems Inc and Peoplesoft Inc. It has had to strip out most of its US sales force and train newcomers into this new approach, hence the particularly bade second quarter. But what’s to stop the other help desk contenders following Inference’s lead and eating its lunch here too? They will [enter the market], but to do so they will have to have some kind of technology like ours, and we already have a market lead, he bravely states. But given that Inference, at its current stock price and with its current number of shares outstanding at under eight million (the company has been aggressively buying over 505,000 of its own shares back, presumably to raise the price), is worth, according to First Call estimates, a fairly modest $34.4m to buy outright, this begs the question of whether Inference is being played out on a string by one of its rivals who will soon step in to snap it up, saving it the trouble of developing its own technology and also getting a blue chip customer base (Inference customers include British Telecom Plc, GTE and Hewlett-Packard Co). To which speculation Tierney has, natch, no comment. But to avoid thus fate Tierney must really pull a few rabbits out of his hat and build a dependable business model. A third successful turnaround may be stretching even his luck and skills too far.
![](https://www.techmonitor.ai/wp-content/uploads/sites/29/2016/08/UploadsNewsArticle4887467main-300x300.jpg)