Wednesday, April 15, 2009

How to procure a new CMS

In my previous post, I railed against the PQQ for upping the cost of software procurement and for favouring the larger suppliers/vendors. Well, how should a company or local govt buy their new CMS (content management system)? The following is based upon Kaushik's recommendations (p.104-7) in Web Analytics - an hour a day, & my own experience.
  1. Appoint a project manager (PM) and make them responsible for the purchasing decision. This PM should also be the person who is primarily responsible for managing the company/council website. They must have the support of the CIO, Council Leader, etc.
  2. The PM should appoint a small committee (2-3) of people who will use the CMS on a daily basis. One member should be technically competent if possible.
  3. Write an outline technical requirements document.
  4. Based on the expertise and experience of the committee members, select 3-4 systems that you believe will meet your outline requirements. At least one of these systems should be open source.
  5. Set up test environments and install each system - spend several months getting to know each system, testing them, extending them, implementing a test solution (a special form or content type, etc).
  6. When you have selected the tool that best fits the requirements, is easiest to use and has the potential to grow/keep pace with your organisation, then write a PQQ for a company to implement your chosen CMS. Don't write the PQQ in a way that favours the larger suppliers - a small/local company may provide a better, more tailored service.
  7. Inform the business that you have chosen a tool and supplier for the new CMS.
At each point in the above process you might say you don't have the expertise in-house - to setup test environments, for example. Well, appoint a company that can do that for you - setting up a server to host a CMS should only take a day. If this is the situation, you should probably go for a hosted solution. If you don't have in-house expertise, buy it in. Money spent up front now will save even more money down stream later.

The benefits of the above approach are:
  1. That you get to play with software before purchasing. Too often you only learn of the problems after you've spent loads of money & its then too late to change supplier or system.
  2. By making a named individual (the PM) responsible for the purchasing decision, you encourage a sense of ownership and responsibility for the outcome. There is less room for back covering and a greater focus on delivering benefit for the business.
  3. You limit the potential for scope creep. The inclusive approach only leads to occasional (or never) users asking for complex tweaks that dramatically increase costs for limited business benefit. You also prevent your project from being hijacked by another department working to another agenda - by keeping control your destiny remains in your own hands.
  4. Finally, the process is quicker and cheaper and the outcome should be more certain.
  5. You might also learn something about the qualities that your staff have or lack.

Do PQQs drive up the cost of software?

I recently had the frustrating task of completing a PQQ for a university looking for a replacement CMS (content management system). The PQQ was in two parts: The first was a mini technical specification, along the lines of 'can your system do x, y, z', etc. This in itself was quite a time consuming document to complete.

The second posed as a supplier competence test, and alongside the usual questions about financial probity and accounting history, were very detailed requests for costings and project timescales. For example, it asked for costings for staff training and associated budgets. Essentially, it was asking for what I considered a PID (project initiation document), which is something for a later stage, after a provider has been selected. This part of the PQQ was also very time consuming and the further I progressed, the more annoyed I became at the bureaucracy involved.

Why did this PQQ (& this is not an isolated example but a growing trend I fear) ask for so much information, information that arguably should not be requested at such an early stage? Well, the key reason I fear is fear itself - public sector workers are risk averse. The PQQ has become a back covering exercise conducted along the way to chosing the most expensive provider. After all, no one gets sacked for chosing the biggest, best and most expensive system. There is certainly no reward for saving money - budgets must be spent!

I have the benefit of having been on the other side of the PQQ process, whilst working as a consultant for a local government, so I know something about the mindset at work while drawing up these PQQs. They are written for the precise purpose to absolve the individual/department concerned from any blame for the outcome - afterall, the chosen product/supplier was the best on paper!

There is another, not unrelated dynamic at play too. After completing the PQQ I started to read Avinash Kaushik's, Web Analytics - an hour a day. Here he outlines (p.103) why PQQs (or requests for proposal) drive up the cost of (web analytics) software, largely because everybody has input into the requirements (scope creep), you then add requirements about vendor suitability, set an aggressive response deadline, and then have a committe of 'cross-functional representatives' to select a winner. As the long and expensive process ends, the
"guaranteed result is that you will almost always pick the most expansive, and usually one of the most expensive, (web analytics) solutions."
PQQs therefore are usually expensive and lead to expensive outcomes.

Thus, the PQQ becomes a vicious circle. The more extensive the PQQ becomes, the harder it is for SMEs to compete. So, these PQQs favour the larger companies, with sales teams dedicated to responding to PQQs, so SMEs are discouraged from participating because they don't have the expertise to respond to PQQs. So in turn, the procurement cost goes up - those sales teams need to be paid for.

While the PQQ may help guard against corruption and promote transparency, it also tends to engender complacency and acts to absolve those involved of having to make a real decision, one that is based upon their expertise and knowledge. Finally the PQQ also tends to lead to more expensive outcomes as it favours bigger players in the market place.