The pioneers of push technology are trying to re-position their companies after a stuttering start. Over the past two years, Kim Polese, chief executive of small internet start-up company Marimba, has become a media superstar. She has made keynote speeches at some of the industry’s most prestigious conferences, appeared on the covers of countless publications, and has been described variously as ‘the Madonna of Silicon Valley’, ‘the diva of push’, and ‘the poster child of the internet generation’. Cynics might say the reason for Polese’s rise to stardom is that IT magazines do not often have an excuse to feature an attractive, young, female chief executive on their covers. Others put it down to the exciting potential of the technology she represents. Polese has become the human face of ‘push technology’, an internet technique that automatically delivers software and content to a user’s desktop over the internet, without the user having to request the information. In early 1996, analysts predicted that the potential impact of push technology on Internet applications was colossal, and Marimba, along with push competitors such as BackWeb and PointCast, were likely to experience exponential growth. High profile venture capital firms such as Goldman Sachs and Kleiner Perkins seemed to agree. Both organizations invested significant funds in push-related start-ups. But now, two years on, the push tag has become rather tarnished. Indeed, even the companies that pioneered push are now seeking to distance themselves from the term that made them famous. We’re not a push company – we’re actually an application distribution and management vendor, insists Tom Banahan, vice president of market development at Marimba. We’re not using the word push anymore, concurs John Laing, president and CEO of rival Wayfarer. It’s gotten a bad name. We’re selling information management. And PointCast, one of the earliest and biggest names in push technology, now markets itself as ‘the internet’s number one news and information service.’

Marshmallow

But why have these companies so deliberately turned their backs on the concept that originally carried them to the forefront of public attention? The answer, they say, is that the term ‘push technology’ is too vague. Push was a very big marshmallow, says Banahan of Marimba. It was a category without definition and many companies were grouped under this umbrella term when in actual fact, they had very disparate technologies, very different value propositions, and were aimed at very different markets, he says. Analysts agree. Push technology, says Stan Lepeak, vice president of market analysts the Meta Group, was never a discrete process, but simply a technology that could be used to enhance an application. But the problems associated with push are not merely confined to points of definition. For users, the technology’s fall from grace has been precipitated by implementation difficulties. For example, the rate at which early push technology applications ate up bandwidth and choked servers with unsolicited information soon dampened its appeal among network managers. Network overloading was a particular problem for PointCast, says Lepeak. PointCast, launched in February 1996, was designed to scour the internet for user-specified news and information, and then deliver it to the desktop on a rolling basis. But users found the constant stream of data clogged up networks and that the data was often irrelevant. As a result, says Lepeak, many customers stopped using it. PointCast has since sought to correct this weakness by pitching a caching manager to IT departments and dividing content into 20 channels, each aimed at different vertical industries, such as telecommunications, law and financial services. But there were other problems with the push concept – many applications proved too cumbersome over early dial-up internet connections. Also damaging, says Lepeak, was the feeling among some reluctant users that they were being subjected to a sophisticated form of ‘spam’ – unsolicit

ed, multimedia mail. User confidence in push technology has been further dented by bitter wrangling within the industry. Microsoft and Netscape, both keen to corner the consumer Internet market, have publicly disagreed over push standards, while Novadigm, a push-based application management company, and Marimba have been locked in a fierce legal battle since March 1997. Novadigm claims Marimba has infringed its push- related patents, an accusation Marimba fiercely contests. Moreover, a number of the early push players have disappeared, creating instability in the market. According to analysts, the reason for the fallout is that there were simply too many players chasing too little revenue, leading to a spate of acquisitions and bankruptcies. During 1997, Washington, DC-based Torso was acquired by webcasting software specialist Wayfarer; San Mateo, California-based inCommon was bought by a Silicon Valley neighbor, Reuters subsidiary Tibco; San Jose-based BackWeb snapped up Canadian push specialist Lanacom; and New York-based iFusion went bankrupt. The companies that have survived this enforced period of consolidation say that new variants of push technology are now ready for widespread commercial application, and that business prospects are looking up. But beyond the accolades and front cover articles, it remains hard to quantify the success of the vendors that remain in this sector. The majority are still privately-owned and have not revealed details of their financial results. Banahan of Marimba points to the company’s burgeoning headcount as evidence of its success. In 1997, he says, Marimba, set up a number of pilot systems and this year, those pilots are beginning to turn into deployments. Eli Barkat, CEO of BackWeb, is a little more specific. Over the past five quarters, he says, the company has closed 200 deals with Fortune 1,000 companies, and the average deal size has risen from $10,000 to $60,000. He adds that BackWeb will achieve profitability this year, possibly. In March 1996, meanwhile, media baron Rupert Murdoch’s News Corporation reportedly offered between $350m and $450m to buy PointCast, but was turned down. The company has since filed for an initial public offering (IPO) which is expected to raise about $37.5m. It will be listed on Nasdaq, and the offering has been underwritten by three prestigious merchant banks, Lehman Brothers, BT Alex Brown, and BankAmerica Robertson Stephens. According to Lepeak of the Meta Group, the eventual fate of these companies will depend primarily on the quality of the goods and services that vendors can develop around push technology. But, he says, the leading vendors were right to eschew ‘push’ as a marketing concept in favor of emphasizing applications. And, following on from that change of focus, a number of viable applications of push technology are becoming apparent (see below, Push technology reborn). With the development of applications such as these, the Yankee Group has forecast that the IT industry will generate more than $5.7bn in push-related revenue by the year 2000, from less than $10m in 1996. Although many technology issues still need to be resolved, such as how to make it easier to plug push technology into an existing environment, Lepeak also believes that push will become an integral part of a wide range of products. It’s not the panacea that was promised, but it could certainly be extremely valuable as a facet of broader applications, he says. Barkat of BackWeb is characteristically bullish about his company’s prospects, saying that applications based on push technology will become mainstream in the second half of 1998. There is, however, a clear motive for this forthright optimism. BackWeb is aiming to follow PointCast’s lead and go public. We’re absolutely, 100% on an IPO path. I believe life begins the day after you go public, says Barkat. An IPO is also firmly in Marimba’s company plans, says Banahan, although he will not confirm that such a move is scheduled for this year. Will these companies – PointCast, Marimba and BackWeb – make it? Lepea

k of the Meta Group believes so. All will [go for an] IPO this year – if they don’t get bought first, he says. Other analysts also confidently name these vendors as the three companies out of a handful of players with a reasonable chance of success. The prospects for others, including Infobeat, Wayfarer Communications and Datachannel, they say, remain in the balance.

Computer Business Review

á