From Software Futures, a sister publication
Select Software has cut an enviable swathe through the muddied waters of object modeling in the US – Can it be stopped? Select started off as a tiny software company based out of Cheltenham, UK and began a policy for winning hearts, minds and market share. Four years on, and we find Select no longer so closely aligned with its then hit record product, diagramming tools for supporting the development methodology, SSADM (Structured Systems Analysis and Design Methodology), which was itself developed around the time Man moved to bronze tools by our old friends LBMS, then a consulting operation pure and simple.
By Gary Flood
Instead, we now find Select speaking American, objects, Forte and Microsoft, in something roughly in that order. Now its chosen established methods and diagrams to support are those associated with Rumbaugh OMT/Jacobson Use Cases, not dusty old SSADM, and its target development environments that of second generation high-end client/server 4GL Forte (as in, C++ territory) and Microsoft’s Visual Basic Enterprise Edition, which is claimed by Microsoft to have sold an impressive 100,000 copies since last September. Not that it’s all been totally plain sailing; there was a brief and curious interregnum involving what turned out to be an ill-starred choice for an executive to lead its push into the North American market. Let’s just say that we’ll spare everyone (and the person concerned) their blushes by concentrating on the positive elements in what seems to be one of those genuine cases of a software company doing it right, looking after their customers, and proving a pleasure to work with. These pages are so seldom blessed with such a genuinely positive story that we’d consider it well-advised by other vendors (and you know who you are!) to listen and learn. The first part of the story must be an update on who or what Select is, in terms of facts and figures. Central to that aspect is the importance of 33 year old Stuart Frost, its founder and managing director, to the whole Select philosophy. Frost founded the company way back in 1988, and has remained its driving force ever since. Its first product was an affordable PC tool to help developers using the Yourdon structured method, which sold like hot cakes at its unveiling at a trade show that year. Its next flagship tool, SSADM Professional (retailing for less than $1,500!), debuted in 1993. Select’s financial year used to end in September, and for the year ending in that month for financial 1992 it recorded turnover of $916,500, with profit of $82,500 (note that Select is still private and all these figures have to basically be taken on trust). For financial 1993 it claims to have made an extremely healthy $600,000 profit on turnover of $2.4m. After that milestone year it’s stopped releasing profit figures as such, but insists it’s still doing very well. In growth terms, it’s claiming 1994 ended on a figure of $3.2m. Then in 1995 it switched its financial year to March, so its revenue figure of $5.5m has been extended to ‘about’ $9m for the year just now ending (i.e. an 18 month year). It’s confidently predicting 1996 will take the company into the $13m runrate. To put it another way – in four years Select’s headcount has gone from less than 10 to over around 100.
Established methods
Select has always been a pragmatic company. We can see the first aspect of the reason for its happy vibes in its flexible approach to supporting methods that look as if they’re about to become established. A CASE (pun intended) in point: the move to OO. For steady as Select’s growth has been, it wasn’t until it decided to shift out of the rapidly maturing structured methods space into that of the new hot market of object-oriented methods support that it really began to push the pedal down. (Its first foray was a Select OMT Professional product, which we covered in January 1995.) In August 1994 it announced what some at least saw as an unholy alliance with everyone’s favorite 80
0 pound gorilla, the Redmond Rampager, the Demon of the Desktop itself, Microsoft. The idea was that by integrating Select software engineering tools with OLE/Visual Basic, users and developers would benefit from a client/server-based development approach contrasting with both Select’s existing Yourdon/SSADM heritage and other 4GLs’ adherence to some of the principles around Rapid Application Development (RAD). This new wrinkle to the age-old how to build software better conundrum involved support for Rumbaugh/OMT and Jacobson, as noted earlier. Select believed this approach added value through a concentration on the specific activities that comprise business processes, viewing such processes in terms of two levels, local and corporate. From this it built a four-layer architecture – a user interface level, local and corporate levels and a physical level; each level containing objects defining its functionality. The first level, that of the user interface, contains objects such as screens, menus and forms and is therefore highly dependent upon the implementation technology used. The second and third levels covered local (data needed for specific projects) and corporate (holding data relevant to the whole company) objects. The fourth level is defined as the physical level, and encapsulates the storage of data in a relational or object database. This whole approach has been subsumed under a global name, Enterprise, which is the target for specific deployment environments – hence Select Enterprise, an object-oriented modeling toolset for multi-tier client/server applications allowing code generation for VB or Forte.
Pleasing your partner
The link – sacred or profane we’ll leave you to decide – brings in the next aspect of Select’s success: good, solid partnerships that mean more than just marketing blah, and where both companies actually get something. The two most critical so far have been with Microsoft and Forte. The Microsoft link is obviously important to Select and Frost in particular, but not so much from the bigger player’s point of view. Last month Microsoft was happy to include Select in its Enterprise Development Partners Program, whereby 22 charter members promised to work together to provide customers with robust, integrated custom client/server tool suites. Partners including Select will receive access to early alpha releases of Microsoft software and invitations to technical summits focusing on joint integration of Microsoft and partner development tools. The program will also offer co- marketing opportunities for vendors in activities such as joint advertising, enterprise partner catalogues, case studies and seminars. (How far is this from FlexiInternational’s Stefan Bothe’s wry remark that he was honored to pay to become one of Microsoft’s 17,500 closest partners? Well, 22 is a smaller number…) But Microsoft demurs. The program provides the foundation to expand large corporate development using Microsoft Visual Tools, says Jon Roskill, director of marketing for Visual Basic at Microsoft in Redmond. He went on to tell us: Select is one of the early leaders in data modeling; it understands the Microsoft developer extremely well, and its mindset is fairly aggressive. It’s shown a serious commitment to the Visual Basic platform. As even Frost agrees, it’s still early days with Microsoft, but it sounds promising from Select’s viewpoint that it’s in the inner circle for VB already. On the Forte side, we talked with Ed Horst, director of product marketing. He stressed that Forte has links with many vendors (including Cadre, Rational and Ptech) in what he prefers to call the OO CASE instead of object modeling area. But one thing Select has promised the Forte user base that has really got them pricking up their ears is an OLE-based bi-directional link between Forte objects and Select models. Forte hasn’t missed the murmurs of excitement. It’s a pretty nice little product and they’ve priced it very aggressively. (Select Enterprise costs less than $3,000). Select has also made
investments – like the bi-directional link with our product – that stands out. We’re impressed, says Horst. The point about these partnerships is that Select has made a reasonable bid for – if not dominance, then certainly a strong role in – the object modeling end of the emerging n-tier client/server development market. This is pointed out in a statement from another kind of partner – specifically, a financial backer. Laurence Garrett, speaking from his Bristol, UK, office, works for one of Select’s most important backers, European investment company 3i (from Investors In Industry). It and Select have been involved since 1990. Why is 3i interested? Because Select is going to become the de facto standard for object oriented CASE tools – it will be the gorilla. The product offering is robust, its alliance with Microsoft and VB and Windows NT offerings are quite key, and objects are where the market is moving. Watch this space. Well, he would say that, right? We spoke with Dave Kelly, director of analysts Hurwitz Consulting Group’s Applications Development Group in Newton, Massachusetts. Select’s done some very nice integration with its work on Forte and Visual Basic; what it’s doing are certainly the directions these tools have to go. It may be a better fit with the kinds of skills associated with Forte developers than Visual Basic, who are more used to tools like LogicWorks’ ERwin, but it’s done well. It’s hard to create an overwhelming message for this market, but Select seems to be doing that. It hasn’t locked it up yet, but it has got the potential to become the gorilla. To sum up the whole Select approach to partners, we’ll quote Frost: There were other companies that did a superb job of building a large number of partnerships before we came in. But they were very thin. We picked off the ones we wanted and deepened them. Precisely!
Happy customers
But it can only become the gorilla if it’s got satisfied customers – it can make money for itself and allies, but it has to convince customers it’s got something special. Here again, Select wins. We spoke with a consultant from Cap Gemini working on-site at Microsoft’s HQ, who prefers not to be named. (But guess what – they like OMT in Microsoft!) She’s someone with extensive knowledge of CASE tools on platforms like the IBM mainframe. Boy, is she impressed with Select. I loved it – I tried it right out of the box without the manual and it felt just right. It’s not a perfect tool, but the development team over there are incredibly quick and are very responsive. In January we quoted Mark Joyce, supervisor, Fibre Systems Engineering, for the Telecommunications Products Division of Corning, based in Wilmington, North Carolina. Joyce particularly likes Select’s support for OMT and Jacobson and its close integration with Forte tool constructs such as events. He also favors the stated direction Select is taking to bi-directional OLE-based round trip integration with Forte, which he thinks Select might be first with. He would like greater support for certain other Forte constructs and closer integration with design, but overall: I’m very impressed with its customer support and speed of response. I’m also generally very pleased with its general direction. The design of the product is well thought through and laid out, as well. Gene Rooney is president of a 20-strong San Francisco- based consulting firm, Sage Solutions, and he adds: We’re focused on the enterprise and three- tier client/server, and we’re very pleased with Select. And another: Tom Reidy, senior consultant for CSC’s Consulting and Systems Integration unit, told Software Futures that he’s very encouraged and excited about Select’s aggressiveness in pushing forward the bi- directional link idea. He sees a good philosophical and methodological match between Select and CSC’s internal Lynx client/server development method, citing in addition the company’s excellent strategies and vision and responsiveness.
What next?
So what will happen with Select? Can it c
apitalize on these friendships and the good will it seems to have created? One thing that might impact Select is the market acceptance of objects, and it may have to be patient. A lot rides on the fact that folks are using Select to help fireproof future maintenance and design of bleeding edge applications. But in business, as in life, nothing is certain, and risk is what turns on entrepreneurs like Frost, who’s already turned down one offer to buy him out. Everything’s got a price but they’ve not come up with it yet, he jokes quietly. While there is talk of an IPO in 1997, he clearly wants to continue having fun and steering Select. And given his performance so far, that’s the best place for him. Indeed, a case of watch this space.
