François Nonnenmacher (EN) | FR

CNet has a rather long story on the challenges that the IT services firms are facing. Worth reading only if you care about that particular industry, which is indeed facing a profound overhaul. This article has a rather deep coverage of the issues at hand of both clients and suppliers.

Before I go on with a little comment, I have to introduce two French words:

• Informatique (n. f.), which covers both concepts of Information Science and Information Technology. It is rare to see a word or concept being much shorter in French than in English, and I see more and more appearances of the word Informatics (although I am not sure it has gained significant adoption yet). For a little history (en français ici et ici), this word was crafted by Philippe Dreyfus in 1962 (I had the chance to meet him at Cap Gemini, of which he is a retired veteran).
• Artisanat (n. m.), which translates into craft industry. The translation is actually unfaithful of the French concept, the art of the craftsman, craftswoman, l'artisan(e). In that sense, l'artisanat is often seen as the antithesis of l'industrie, and something artisanal (produced by an artisan) is opposed to something industriel (an industrial product).

After 23 years of heavy exposure to computers and all sorts of IT things, the apprentice has learned two fundamental things about informatics:

1. it is not an exact science
2. it is an artisanat before being an industry

The first one will strike people who think that computers are one of the most predictable things around. They may well be (although I still doubt that), but informatics is not just about computers. It is very much about humans, their needs, their requirements and expectations interpreted, written and rewritten into requirements, specifications, which get re-interpreted then transcripted into code which enjoys a cycle life full of debugging, patching, updating in minor and major versions before dying in obsolescence. Informatics is a world of uncertainty where the conservative metrics are: whatever the scope of a project, the timescale and budget will be twice as much as initially predicted.

The second one is what, in my eyes, makes the whole informatics world really interesting. A significant part of the IT services business is done by individuals doing haute couture work as free-lancers, or in small firms or in small expert teams in bigger companies. The innovation and top-notch expertise mainly comes from individuals (whatever their work status is), and the whole open-source movement is the quintessence of artisan IT work (one could argue that it is the only place where you see innovation en marche nowadays). The current IT services industry is basically split into four areas: consulting, technology, outsourcing and professional services. There are artisans working in all those areas and in all sorts of companies but -- because consulting is essentially an individual, client-facing job and professional services are very local -- the pressure to industrialize IT services is only sensible in technology and outsourcing. Besides, industrialization makes sense only where economies of scale are beneficial but may come at the price of lesser flexibility, adaptability, customization and -- mother of all business fears - competitive differentiation.

Informatics is somehow like cuisine. The art of the Chefs Information Officers consists of knowing all the grades in between the artisanal sublime and the industrial extreme, the ingredients and recipes, and how to pick venues and menus relevant for one situation.