It’s been said that the IT profession has the worse collective memory of all. Perhaps this is down to its relative immaturity or perhaps this is down to the sheer pace of innovation and adoption of IT in business. Many folks believe it is a combination of course. But whatever it is I do not believe it is an unfair criticism of the industry to say it appears to be pre-occupied with the future while left in its wake are a series of innovations (‘legacy’) which are causing significant management issues today. As Les Hatton says in the excellent 2004 BCS publication on ‘The Challenges of Complex IT Projects’ – “…a significant percentage of IT project failures, perhaps most, could have been avoided using techniques we already know how to apply. For shame, we can do better than this.” And we’re getting excited again, aren’t we? The concepts and emerging technologies of ‘Web 2.0’ are promising much. Even today’s Web is delivering new business models nobody took seriously 10 years ago and most derided when the dot.com bubble burst. And this time it isn’t an innovation that helps a department or several, or supports specific marketplaces. It’s an innovation that’s connecting social and business interactions on a global scale, challenging organisation boundaries and connecting individuals outside of traditional government and corporate controls en masse. This in the context of the average age of a business application today being in the range of 10-20 years. There is an argument that says despite all the risks and issues, we as an IT industry get there in the end – so does it really matter that we seem to have a collective ‘memory reset’ feature when it comes to the past? I think it does. Because for each high-profile failure we further dent the reputation of the IT function in the eyes of the business and we increase the opportunity cost to bring in real IT innovation that can really transform the business. Many CIOs I work with really do want to bring innovation in but are dealing with 50 legacy issues that are more pressing. It’s no surprise that many of the high-profile businesses leveraging Web models today to connect horizontal processes with their markets have started from scratch based on the model. I wonder if their own IT legacy will kick-in in 5-10 years time and instead of being an asset become an anchor? Is there a way to break the cycle? I think we have to find one otherwise this time things might get a lot worse. Business packages today are more like a lorry load of highly configurable services which can be orchestrated into agile processes rather then a pre-determined, tightly contained functional and data model. The Web model starts with the premise that everything can communicate with everything else. Departmental and enterprise systems – the ones we’ve been used to procuring and implementing – start with the opposite premise. Interactions across boundaries – and into a globalised economy - are the norm not the exception. I actually think, if we look back, that this has been less of a memory problem and more of a clash of management experience with specialist knowledge of the IT wave of the time. Each time there’s been a new fundamental commercial IT innovation (mainframes, departmental computing, PCs, Web evolving to “Web 2.0/Semantics”) there have been folks who specialise in the new (tended to have ‘grown up’ with the new) and those who understand how to bring in IT enabled business change based on the previous wave. I believe there’s a pattern that shows the IT management of the day being a cycle behind the IT specialists of the day. For sure there have been common aspects but the key factor in each cycle is that the differences outweigh the common. And I think it is this, rather than an inherent memory problem, that’s a root cause of the issue. Actually there are many views on these cycles – e.g. client/server, ERP, Web, SOA – the point is there is a tipping point in the cycle where new IT approaches clash with known IT management approaches. What can be done? We often focus much management attention on the ‘business / IT interface’. Perhaps a pressing focus we need is on the ‘IT management / IT specialist’ interface to create a language both can understand to apply the lessons of history in the context of the current. And the good news is that Enterprise Architecture as a developing field has spanned the last cycle – and there’s good evidence that it helps spans not just the business / IT interface but the IT management / IT specialist interface too. The opportunities are great – but I think with this wave the risks for some may be even greater.