CMS procurement: Time for open source?

I attended the Plone in Business seminar this week, organised by Netsight as part of the annual Plone conference, being held this year in sunny Bristol.

To be honest, I didn’t learn an awful lot about Plone but it was a very useful afternoon as there were some good speakers on the procurement process for CMS’s generally and where open source software (OSS) fits into the market.

First up was Mike Grafham from Deloitte. He said that the CMS market is massive these days and it’s actually very hard to pick a CMS. You need to be very clear about what you want to use the CMS for, which means spending time  thoroughly analysing your business and user needs. The technology is in fact not half as important as content governance models and he noted that projects usually crash because of requirement or process failure, not failure of the underlying tech.

Mike posited 4 rules for CMS selection:

1. Know what you need it for (and what you might want it for in the future). You might also want to consider that the locality of vendors and getting on with them are pretty key requirements
2. Run a structured selection process. You need to win hearts and minds and bring along the people who are ultimately going to be using the product. When looking at CMS’s, it is useful to know its provenance and its core market e.g. higher education
3. Consider a proof of concept – helps reduce risk. I’m increasingly of the opinion that this is essential. You need to see the CMS actually working with real content or you simply do not know if it is suitable. So you should try to get vendors to build you a prototype. However, it was noted that this might actually put open source vendors at a disadvantage as they do not have the sales budgets of proprietary software to take the strain. Although, the point was made that Plone, for example, is very quick to prototype in
4. ‘Open source doesn’t matter’. Controversial comment to make at an OSS conference! But it does drive home the point that it is the processes, people and requirements that are important, not the flavour of software. Predictably, this stirred up some audience comment, such as open source protects the customer in the case of  vendors going bust; you own the code and data. But I guess this comes down to exit strategies, which all CMS projects should have. What do you do if it all goes horribly wrong (as it will at some point).

I was interested in how you got from point 1 to 2. Last time I ran a CMS implementation project (here at JISC), we opted for an open tender process and got absolutely inundated. Mike suggested that once you have your requirements together, that is the point to call in the experts that know the market and can draw up a shortlist of suitable candidates based upon these requirements, the market you are in, budgets etc.

I suggested that the days of the monolithic, enterprise CMS might be coming to a close. Mike felt they would persist because it was in the interests of IT departments to maintain one system, both from an administrative and a financial point of view. A pragmatic assessment, but a little depressing nonetheless.

Next was Graham Oakes, a consultant who helps organisations select CMSs, talking about OSS in the public sector. He felt there were some really strong open source CMSs these days that were the equal of any proprietary system; these include Drupal, Plone, Joomla, Umbraco, Hippo, Typo3 and eZ. The benefits of OSS for him were:

1. Low up-front cost, which allowed you to do proof of concepts or pilots with no major outlay;
2. Easier to work with (but only if you have the skills);
3. The lack of licensing issues means apps built upon them scale up and down well;
4. They demonstrate openness, which is important for public sector bodies generally but particularly at the moment.

The risks are:

1. People misunderstand costs. OSS is not ‘free’. There’s still hardware, implementation, migration and content generation and management. These dwarf the cost of licensing, which is a small part of total cost of ownership.
2. You are still ‘locked in’. Moving from one system to another is difficult, time consuming and expensive. It is also not easy to reuse modules developed for Plone, say, in Drupal as the underlying architecture is so different.
3. Mismatched scale. OSS vendors are smaller and therefore higher risk for public sector managers. They also cannot absorb sales costs of £20K into future licensing revenue. Graham asserted that public sector procurement (e.g. OJEU) is ‘broken’ and is biased against open source vendors.
4. Unreasoned decisions i.e. you get Sharepoint for ‘free’ in a Microsoft package but the cost of implementing and servicing it outstrips any ‘savings’.

In conclusion:

1. Not all OSS is the same. Variable quality and capabilities;
2. Don’t underestimate the cost of OSS;
3. The team, not the tech, creates success;
4. No better or worse than proprietary software. Look for the best tool for the job.
5. OSS aligns well to evolutionary delivery e.g. ‘try before buy’, agile development. This is best practice for software development anyway
6. Public sector procurement is broken.
7. OSS aligns well with current public sector drivers to reduce costs (lower up-front costs), engage stakeholders (OSS underpins most social media) and to be innovative and agile (supports low cost experimentation)

Really useful session I felt. Still wouldn’t mind finding out more about Plone though!