The lock-in decision.

The lock-in issue is always a serious, and often unrecognised and unconsidered risk. In a long life, I have seen many wonderfully productive platforms – from IBM, Microsoft, Oracle and so on – that locked their adopters in to the platform. And I have sometimes seen these, initially enthusiastic, adopters wondering how they can escape (usually, expensively and disruptively) when their business strategy diverges from that of their platform owner. Lock-in worries me – which is why, despite my enthusiasm for the wonderful Adobe Lightroom product, all my photos are in my own file structure and I keep high quality, lossless, TIF images of the final edited versions.

Nevertheless, when I have long conversations with developers that really hate lock-in, I have to point out that if lock-in to a particular platform generates sufficient business benefit – money – to cover the cost and disruption when the time comes to move away (and that time always comes, in my experience), and to still leave an adequate profit, then the lock-in doesn’t matter. Of course, this argument implies that the “victim” of lock-in is mature enough to do the risk analysis and has the metrics to support its decision, but that’s really a governance issue.  

These thoughts were prompted by continuing discussions with high-productivity development platform vendors around the advantages of “no lock-in” and their true value. Then, I had the chance to ask Antoine de Kerviler, CIO of Eurostar, about just this issue.

Eurostar is adopting the Salesforce platform, and developing applications with Force.com, to transform the company. It has a very strong story around rationalisation of its automation environment (with serious reductions in complexity, siloisation – and simply the number of different application solutions it has to maintain). More importantly, perhaps, it is focusing on transforming the Eurostar user experience, for both customers and employees, with collaboration enabled by the excellent Chatter product. I was really impressed by the vision, which even includes employees using their own devices or, rather, owning their own interfaces to Eurostar systems – in line with the emerging freedom people expect these days, to work the way they think best. And the Eurostar vision is very much enabled by the Salesforce platform – it is quite hard to think of anything else with exactly matching capabilities.

Nevertheless, as I asked de Kerviler, doesn’t this also deliver the mother of all lock-ins? Not just technology lock-in, which is no worse than the IBM, Microsoft, Oracle, etc lock-in we’ve all learned to manage (more or less) in the past. Eurostar faces a cultural lock-in, because Chatter and the Salesforce capabilities will become intimately combined with the Eurostar business culture.

His answer was, I think, the right one. Eurostar has had considerable internal debate over the lock-in issues, he says, so it is going into this with its eyes wide open. Moreover, de Kerviler has reviewed Salesforce’s roadmaps in detail and Salesforce has had to convince him that its strategies will continue to support Eurostar for the foreseeable future. Yes, walking away from Salesforce would be difficult but it would have to be managed, if it became necessary. It seems to me that Eurostar’s transformation story is impressive enough to justify a lock-in risk – as long as Eurostar continues to manage the relationship with Salesforce, as I’m sure it will, so it has the time it might need to manage walking away, if it ever needs to. The immediate risks of not transforming the company probably do outweigh the potential risks of lock-in.

Could Eurostar have adopted alternative platforms, possibly with fewer lock-in issues? Doubtless it could have, but I’m not arrogant enough to criticise its choice on the basis of an hour on the phone. I did ask about whether Open Source Software (OSS) solutions were considered, and de Kerviler said that he thought that they had issues of their own: such as a possible focus on low-level code rather than the business; and their own platform lock-in issues, because of the customisation usually associated with OSS.

Whatever the technical issues, Eurostar’s new strategy looks to be well-thought-through and firmly focused on removing silos, increasing collaboration, and delivering better user experiences. If its choice of Salesforce helps to deliver all that effectively, it can probably live with any potential lock-in issues.

Nevertheless, note that my generally positive feeling is based on my belief that the risks were assessed early in the platform adoption process and that Eurostar has a strong, two-way, relationship with Salesforce. Neither of these might be true for some potential Salesforce customers; but such customers would probably take a short-term view of any vendor relationships they got involved in, anyway. I remember working for a bank, many years ago, that adopted an Open Systems strategy, in part, to escape from its lock-in to IBM mainframe solutions. It managed to move itself to a Unix platform with expensive hardware and limited connectivity options, which soon disappeared; and a lot of expensive, proprietary, end-user oriented spreadsheets that became a governance nightmare. With hindsight, the mainframe might have been a better strategic bet! The point is, that avoiding lock-in to proprietary technology is a very good thing, but achieving it isn’t always trivial and a holistic business benefit analysis should always trump achieving even a worthy technical objective.