Israeli company Magic Software Enterprises Ltd plans to be a $100m company within two years, and wants its rapid development tool to be as popular in New York as it is in Tel Aviv. Susan Amos checks its progress
A siren wails. You see blue and red flashing lights. A team of nurses and doctors flies past with a body on a trolley. It’s touch and go to revive the patient, but, yes, the team pulls it off. Is this the latest episode of hit hospital soap ER, starring gorgeous George Clooney and cute Noah Wyle? No, it’s an advert for Israeli company Magic Software Enterprises Ltd’s eponymously named rapid application development tool. If we are to believe the hospital analogy, the application development emergency is upon us. Around 80% of information technology projects miss deadline or are over budget, or both. This figure is often quoted by consultants trying to sell you services, testing tool companies trying to flog you debuggers, and now Magic is using it in its slideshow. If the apps backlog is threatening your career, buy our tool and save your skin, goes the advertising copy. The reason Magic’s so quick, explains the vendor, is that a lot of work is already done for you. Most functions you would want to carry out in a database-heavy application are already written into the Magic engine. All you have to do is decide what functions you want to use and what order you want to use them in. You do this by typing business rules into a table. You throw the table at the engine, or Magic Virtual Machine, as the company has started to call it, and it gets interpreted. So, speed of development is Magic’s trump card. The only trouble is, it plays it over and over again. Each year for the last four years the tool has taken first place in a US rapid application development – or RAD – tools competition, the Droege Developers Competition, run by a small consultancy operating out of Durham, North Carolina. Oracle Corp’s Developer/2000 could get no higher than 17th place, Centura Software Corp’s, formerly Gupta, SQL Windows was at number 20 and Powersoft Corp’s PowerBuilder at number 21. So, aside from rapid application developement races, what else has Magic Software been up to? Expanding operations, growing revenues and dipping in and out of the red. The company has just bought 51% of its Hungarian subsidiary, and also acquired the whole of its Brazilian distributor. There are now 310 people in the company, and in two years, president and founder David Assia aims to crank the total up to 600 staff in a $100m company. To ramp up its US sales, new offices are in the pipeline for New York City and Washington DC. If you lump in Latin America and Canada with the US, the bundle brings in 20% of revenue – the same amount of sales as Japan. Tot up the remaining pie chart figures and they give Europe 41% and Israel 10%. By the end of next year, the company rather ambitiously plans to go for 40% of sales in North America. Magic has predicted its total revenue for the year at around $36m, and claims to have returned to growth levels of 50% to 60% since the release of its Windows product, Version 6.0, in September 1995. Last year the company did $25m of business. While revenue is growing, income is not. The company has just posted a third quarter loss of $238,000, but this is down to the company’s spending on sales and marketing. And, according to a source close to the company, the loss is due to a couple of big fish deals that slipped out of the net, but have since come good. Assia, though, assures us that his company will bounce back into the black for the financial year as a whole. If it does, it will make a welcome break from the losses of last year, at $438,000 and the year before, at $1.4m. He believes a bright spot on the revenue horizon for next year will be a version for IBM Corp’s mid-range AS/400 box. Once we’ve got a server version of our product [due out in the first quarter], it could be 5% to 7% of our revenue. Indeed at the user conference, Software Futures noted a definite interest from small applications vendors in
converting to the large and solid user base that the AS/400 has behind it. Where is Magic in the $2bn tools market, we quizzed Assia? It’s not the low end, he says, describing that end of the market as supporting up to 50 users and having less logical complexity and performing two transactions per second. We’re running away from the low end, that is Visual Basic and Delphi. It’s not sensible to compete with Mr Gates. Nor is Magic at the high-end with thousands of users and massive volumes of transactions, say an airline reservation system, where you will find tools from Forte Software Inc, Dynasty Technologies Inc and Seer Technologies Inc. It’s camped out in the middle ground, with hundreds of users, and between two to 10 transactions per second. Magic goes head to head with tools from companies like Progress Software Corp, Uniface BV, Unify Corp and Cognos Inc. Magic version 7.1 was announced at the company’s Dana Point, California user conference last month. It promises a memory gateway so you can store the results of a SQL query on a personal computer, to boost performance, and an application programming interface that will make it easier to integrate third party tools. With Magic 8.0 users will get better support for Web applications, including automatic HTML generation. And beyond that is support for Java, ActiveX, and an object- oriented version, Leapfrog, slated for release in the second half of 1998. The research and development laboratory at Magic is laboring to make a conversion tool for smooth migration, but it can only take an application 60% of the way, 40% of an application will have to be re-developed. Support for green screens wins Magic contracts. Indeed it was one of the reasons the company won UK newsagent W H Smith & Co. But we hear that version 7.0 does not support green screens. In fact you have to stick with version 6.0 if you have dumb terminals. Magic’s a quick tool, it’s good if you’ve an applications backlog that goes out the door and round the corner. The tool has a following of small vertical applications vendors, and if it wants to push onwards and upwards into corporate America there are still a few rough edges to smooth over. It needs to standardize its support across all territories, communicate better with customers, and shove its tool in analyst’s faces, to get more visibility. It needs more visibility. Software developers find it hard to convince customers to buy systems written in Magic, a tool they’ve never heard of. Oh, and Magic needs to watch its back. It’s not enough to cross your fingers, throw salt over your left shoulder and chant: We don’t compete with Gates Inc. Before you know it, Bill will have you for breakfast. That’s not magic, that’s business!
