Corel plus Borland plus Linux – is this the big one?
The competition to MS could just be reaching critical mass
Analysis Microsoft must be regretting that Corel's home page is able to quote MSNBC saying last month that "Corel Linux is terrific". Feelings at Fort Redmond must be less generous this week following news of the Corel merger with Inprise/Borland. Corel CEO Michael Cowpland sees his destiny as establishing Linux on the desktop on a big scale, but triumphalism about a "Linux powerhouse" is dangerous in an ethos where power sharing is preferred. The deal was negotiated in less than a month, and is being termed a "merger of equals". This characterisation seems fair given the fact that there is no significant premium on Borland shares (each share will be exchanged for 0.747 of a Corel share), nor is there a protective collar in the event that the value of the shares of either company changes before completion, which is expected in 60 to 90 days. No regulatory problems are anticipated. The companies are valued on the basis of their market capitalisations last Friday evening, making Corel worth $1.366 billion and Borland $1.071 billion, for a total of $2.437 billion. This gives a 56:44 split in favour of Corel, with Cowpland having personally something over 6 per cent of the merged company. It's pretty unusual for a Canadian company to be top dog in a merger with an American company, and some wags have been wondering why it's not the other way around. Questions have also been raised as to whether Cowpland would be better as a Gatesian chairman and chief architect, with Fuller as the CEO - and of course there's no reason why this couldn't happen in the future. No further acquisitions are on the horizon, we were told, and it has been decided to call in McKinsey to advise about the reorganisation of the companies, which is expected to lead to administrative savings yet to be fully quantified, but $30 million in FY 2000 have been suggested. The team will be led by former Corel CFO Michael O'Reilly, who now acts as a consultant and who has agreed to stay until June. These moves are likely keep the Street calm. Although they are not ruled out, there appears to be no immediate plan for further staff reductions, which is hardly surprising in view of the slimming down undertaken by both companies over recent years. Fuller denies that his intention had been to sell Borland when he took over, but repeated that his task had been to increase shareholder value. This deal does not directly do that because the premium is insignificant, but he has created better upside potential. Cowpland stresses that there will be a good two-way flow of technology. Borland will stay in California, so there won't be a repeat of the cost-cutting decampment of WP to Ottawa. Fuller pointed out that despite some claims, "not everything is invented in the Valley", and that the Canadian talent pool was significant. Certainly salaries are much less in Canada. Both Corel and Borland have suffered from serious financial difficulties as a result of acquisitions. In Borland's case, the company never really recovered from the Ashton Tate acquisition, especially as dBase quickly became a spent force in the absence of a Windows version. When Borland founder Philippe Kahn was still at the helm, he nearly torpedoed Novell's acquisition of WordPerfect, since a deal with Borland to acquire Quattro Pro was essential if a suite was to be created. Those present at the negotiations confirm that Kahn wanted too much, and was only very reluctantly persuaded to do the deal. But Novell found itself with an office suite that was fast losing momentum against MS Office, and Bob Frankenberg, who took over after Ray Noorda left, had to divest the WordPerfect suite to Corel (and Unix to SCO) in what amounted to a fire sale. Corel therefore inherited Borland's Quattro Pro spreadsheet, as well as around a million Paradox database licences that Novell held. Later Corel also acquired Paradox from Borland. Only now is the Corel WordPerfect Suite beginning to regain momentum, largely because it is being made available on Linux as well as Windows. Finance Corel has moved into profit, albeit thanks to a windfall from the Canadian tax man at a very welcome moment. Borland had expected to move into profit in Q3 this year, and now thinks this could be earlier because of savings resulting from the deal. Since the acquisition is being done with paper, Corel will gain cash as a result, which should make possible a better marketing infrastructure and more advertising. Borland itself has been producing Linux products for more than two years, although it has not broken out its Linux revenue in the past. Cowpland noted that this would change, and that the combined company - using the Corel name, but keeping the Borland product names - would break out Linux revenue. So far as Corel's Linux revenue this year is concerned, he estimated it at $20 to $30 million, before that from Borland is added in. Cowpland denied that Corel was totally gambling on Linux, because "Windows products were the cash cow" and he expected the demand for these to remain strong. Derek Burney, Corel's CTO, said this week that Corel only devoted 10 percent of its resources to Linux directly, although indirect support effectively resulted in half of his staff being involved. Consequently, the Linux bet is actually quite small. There is some truth in Cowpland's claim that there is no similar company in the Linux space, but with combined 1999 revenues of only $418 million, there is a very long way to go before Corel approaches the Suns or Oracles. Cowpland rightly points to the tremendous upside potential, but that does not mean that Corel can easily increase its revenue very quickly because of the low price of Linux products. Products In terms of products, there is no overlap at all, which is quite remarkable, but the two businesses are culturally very different. WordPerfect for Linux requires no learning curve for previous WP users, and little for MS Word users. There have been some 2 million downloads of WP for Linux, Cowpland said. Corel also says it is on target to start shipping the Corel Office Suite for Linux in April, with European availability in May. This is expected to be followed by Draw and PhotoPaint in September or October. Corel is currently deciding what it intends to do with Ventura, with one possible outcome being to keep it as a publishing package suitable for long documents, with up-to-date XML developments. So far as Corel's relationship to Cygnus is concerned, GCC (the widely-used Linux compiler) is still likely to be funded in part by Corel. Corel will not offer an open source version of WordPerfect or Draw, since that's where it makes its money. WINE development will also continue. Fuller claims that some 70 per cent of web servers use some Borland technology. Despite the Red Hat-Cygnus deal, Red Hat and other Linux distribution vendors are not seen as competitors by the new Corel, more as partners, although this view may not be universally held. It now seems probable that Linux versions of C++ Builder, Delphi and Object Pascal will be developed. This is the third Linux merger recently, following Red Hat/Cygnus and VA Linux/Andover. Both Cowpland and Fuller were a little blasé about their competitors. Corel is clearly aware that despite a 90 percent difference between the cost of Solaris and Linux, this will not automatically drive Sun users to abandon Solaris. Sun must view the merger with some concern, since it strengthens the competition to both Solaris and the StarOffice suite. Whatever Corel's desires, its relationships with other Linux distributors are likely to be delicate, but if the market continues to expand, there will be less friction. Corel has made some of the right moves towards open-source developers, and recently started opensource.corel.com for developers. Borland meanwhile has made Interbase open source. Microsoft Relationship Microsoft is probably thinking that with Borland's tools, the new Corel product line-up poses not just increased competition for MS Office and Windows, but for BackOffice as well, and that's unwelcome. In addition, with Delphi as a scripting language, plus J-builder and Java Beans, there's the ingredients of enterprise-level Web development. A reliable assessment of how Corel would shape up against Microsoft is premature in that the outcome of the Microsoft trial may change any prognosis significantly. Both companies have "been bloodied by Microsoft", but the Linux landscape was large, Cowpland said. Several key Borland staff were enticed away when Microsoft offered enormous inducements in an attempt to bring about the crippling or demise of Borland. This resulted in CEO Del Yocam suing Microsoft - a case that Microsoft settled on secret terms, disguised as a technology sharing agreement, with a subsequent acquisition of shares by Microsoft, just as had happened with Apple. Yocam had previously sued Microsoft when he was COO of Apple, although the case was eventually lost because certain key evidence was not produced in court. Fuller said of Borland's Microsoft relationship that it was "good and strong" and "Microsoft will continue to be a player in this marketplace for a few more years", but of course it's what Microsoft does that will count. Whether Microsoft maintains its 10 per cent stake in Borland, as it says it will, or sells it as it did recently with its SCO holding, also remains to be seen. The more probable reaction is no reaction for some considerable time, lest by making any move, Microsoft admits that Corel is a serious if small competitor. Future The merger does seem to be a good thing for the open software movement, but Corel bears the added responsibility of needing to act as a model community citizen if it is to gain and maintain the respect of the Linux community. Corel now has around 500 developers working with Linux to some extent (out of a combined staff of 2,200) and may well grow significantly, but in terms of its impact on the Linux market, it is just a big fish in a very large ocean. Corel will need to work on its corporate image, since it has yet to make significant progress in being accepted as an enterprise player. Cowpland has found himself on the defensive about Corel's lack of large enterprise customers, but points out that it was always the case that smaller companies moved faster, and that he was confident that enterprises would follow the trend in due course. Borland has generally had a loyal developer following, and brings some 5 million of them to join Corel's 50 million customers. To make money, Corel will need to make significant headway in OEM preloads. Certainly its attractiveness is increasing, but whether major OEMs will break ranks with Microsoft may well depend on the outcome of the Microsoft trial, and whether one of the remedies will be a published OEM price list that stops Microsoft exerting pricing pressure on OEMs. ®