Quite often the IT department is overwhelmed with more projects than it can actually complete. What seems an intractable problem that tests the most resilient leader, there are in fact some guidelines that can deliver breakthrough results. One of them, based on Little's law, is so counter intuitive that only practising it can sway the doubters. The best metaphor is congestion on the smart motorway: More people get to work on time by driving slower.
Question: How many projects did your IT department actually COMPLETE last year? Place them into three buckets of capital expenditure (otherwise use the budget figure):
- Between £100k and £1m
- Greater than £1m.
You've just worked out the throughput of your IT department. Roughly. Per year. Warts and all, including temporary staff, screw-ups, holidays, sickness, flexing outsourcing arrangements, planning in advance, overtime, stress, tears, urgent customer requirements, the system that should have never been approved, the favour for the finance director, the executive decree and the one from she-who-shouts-loudest.
Now ask yourself a 2nd question: How many projects does the IT department currently have IN PROGRESS? Place them into three buckets of capital expenditure (CAPEX), as above.
Divide the second number by the first - either as an aggregate figure or for each category. What's your multiple? 2? 3? 4? 5? In the example below, it's 2.9
How do you feel about that? I bet there are more projects in progress now than completed last year. That's a bad thing. Appeasement, starting but not finishing doesn't make you credible.
Guideline 1. You can make a huge difference in the management of demand by setting realistic expectations on what the IT department can actually deliver.
For the next part, ensure that each project in progress has a capital expenditure / budget allocated to it and a business owner (decision maker). Write each project on a small index card together with the CAPEX / Budget figure and owner.
Now, create 3 criteria similar to below.
- High priority: If we don't do this project, someone dies, the company ceases to exist, someone goes to jail, or we lose a major (strategic) client
- Medium Priority: Projects that add revenue, cut costs and / or improve efficiency
- Low priority: Everything else.
Get a representative sample of business decision makers together in the same room. Call them the Strategic IT Governance Board. Suggest to them that their attendance is required to support the prioritisation of their project. If they don't come there is a chance other projects may take it's place. At this meeting:
- Present your throughput analysis above.
- Ask them to prioritise the projects in progress firstly into the three criteria above and then, most importantly, ask them to force rank them from highest to lowest in numerical order. Highest = 1, etc. for the project size
- Write that number on the index card of the project.
Guideline 2. You can make a huge difference to the management of demand by creating or managing one organisation-wide board that represents the main decision makers in the company.
Your Strategic IT Governance Board will meet every quarter. So, divide your total COMPLETED projects last year by 4 for each category of project size. Round that number to the nearest whole number
Guideline 3. You can make a huge difference to the management of demand by assigning these decision makers, as a group, with the responsibility of making the final decision on which projects are started, stopped, postponed or advanced up the queue.
Next, on a very large piece of paper, stuck to the wall, create a 3x3 matrix grid. Let's call it a Kanban. The rows denote the budget size of the project. Columns are split according to :
- Work In Progress
Write the nearest whole number into the respective row on your "Work In Progress" Kanban:
Separate your list of projects into two piles - those with a priority number lower than last year's number of completed projects and those above. From the first pile, take the maximum number of highest priority projects for each project size category and place in the respective Work In Progress column. For the remainder of the first pile, put them in the respective "Pending Column". For 2nd pile, keep them and write the next year on each of them.
Congratulations! You have now determined what projects the IT department will be working on for the next quarter. Keep this Kanban in a public place where everybody can see it. Every time a project is completed, move it to the completed column. Once you have created a space in the Work in Progress, move the next highest priority project into that newly created space.
Guideline 4. You can make a huge difference to the management of demand by throttling it to the maximum throughput capacity of the IT Department. i.e. maximum input rate must be less than or equal to maximum output rate.
Guideline 5. You can make a huge difference to the management of demand by making the IT team accountable for realistic delivery of projects - You've removed one excuse by throttling demand. Another way is to have a visual display showing all stakeholders the current status.
Reconvene your Strategic IT Governance Board at the start of the next quarter. Review progress made, challenges and successes. Ask if the current priority of projects has changed. If it has, don't touch those projects in progress. Only change the priority for those in the Pending column.
Review progress at the end of the year. If there has been consistent application of the above guidelines then you will have been able to pull forward demand from next year! You've just made a huge difference to the management of demand!
- IT need to deliver more projects quickly, not less! I agree. In fact, by actually reducing work in progress and creating a smooth flow of projects, more projects will be completed. Consider the variable speed limits on congested motorways. It's exactly the same principle. More people get to work on time by driving slower. Here is a good article that relates the theory to the IT management
- What about dependencies on other projects or projects that start later? Lower the priority.
- What about lack of sponsorship from the business decision maker/ non-attendance? Q.E.D.
- What about new urgent requirements that falls between the quarterly meetings? As soon as a project is complete, speak to the business owner of the next project to start and check whether they are willing to give up their place in the queue for the new urgent requirement. At least get the two decision makers to discuss the impact without you becoming the middleman. Otherwise call an exceptional Strategic IT Governance board meeting
- Small projects get completed more quickly? Have a "small projects" board with delegated authority that meets more regularly.
- This is too simplistic. You haven't considered the leadtimes of projects, the multiple dependencies and complex scoping and planning required to accurately estimate the delivery of a project, the business case, and indeed whether they are ready to start! Correct. So why at the very beginning of this scenario did we start with more projects in progress than we could feasibly deliver? Can you reliably predict the delivery of your project otherwise? Reference guideline 4.
- This is too slow for my project. I'm going to get it done by myself. IF you can persuade the Strategic IT Board to fund / resource / agree to it separately AND the IT department see no data security issues / no dependencies on other elements of IT Infrastructure, why not? But please make it a joint decision and keep it on the matrix board so everybody can see where it is.
- What about IT projects for IT, or enabling infrastructure projects? Present them at the Strategic IT Projects Board too. The owner will be the IT Director.
- The business owner wants to know when the project will be delivered. IF the project is ready to start (i.e. your due diligence has been completed and the business case is approved), provide a forecasted date, together with a probability of success, based on your risk assessment. If they are looking for a guarantee, that's different - there's insurance. If they are looking for commitment, then there is your personal commitment, that of the IT department and that of the business team who need to be equally committed. The IT department's commitment will be a lot higher if it can focus on less work in progress. Reference guideline 4.
- What about the spare resource that hangs around waiting for the next priority project to start because of their skillset? Cross training to reduce the bottleneck. Continuous improvement to eliminate root causes of typical holdups and increase flow through the IT department.