Tailoring for small and simple projects

Small projects are where PRINCE2 has its worst reputation. People who have seen the full method applied to a three-month job, with a 40-page initiation document and weekly reports nobody reads, conclude that PRINCE2 is only for large projects. The manual says the opposite: the method is meant to scale down, and a small project run well with PRINCE2 can have its whole management framework on a few pages. This lesson shows how, using Harbor City Libraries' makerspace refit.

The makerspace refit

Harbor City's central branch has a room that used to hold microfilm readers. The library board has approved 85 thousand to turn it into a makerspace with a laser cutter, three 3D printers, two sewing machines, a vinyl cutter and workbenches, ready for a launch in ten weeks. The work involves electrical upgrades and ventilation by the city's facilities department, equipment purchase through the library's normal procurement, furniture, safety signage, staff training and booking rules. The head of branch services, Alicia Fernandes, sponsors the project. Dev Sharma, a branch manager who ran the kiosk pilot in the previous project, will manage it.

On the tailoring factors, the project is small in money and time, has one main supplier (the city's own facilities team, under an internal agreement), uses an experienced project manager and is mostly predictive: the room is designed, built, equipped and opened. Two factors stand out, though. The laser cutter and 3D printers bring real safety risks, from fire and fumes to burns, and the space is used by the public, including children, so accessibility and safeguarding matter. A small project can still have important risks, and good tailoring follows the risk, not just the budget.

Combining processes

On a small project, starting up a project and initiating a project are often combined into one short piece of work. Both purposes are still met: before committing to the work, the board confirms that the project is worthwhile and viable (the purpose of starting up), and the project manager establishes a solid foundation, with plans, approaches and controls agreed (the purpose of initiation). Combining them saves a board meeting and a document, not the thinking.

Dev runs a single week of startup and initiation. He interviews Alicia and the facilities manager, reviews the kiosk project's lessons, drafts a combined brief and initiation document and holds one board meeting at the end of the week, where Alicia authorizes both the project and the delivery stage. That one week counts as the initiation stage, so the project still has two management stages: initiation and a nine-week delivery stage. Directing a project continues throughout, but mostly by email and a short fortnightly call.

Combining roles

Role combination is the most powerful tailoring tool on a small project. The rules are few: there must be a single executive; the project manager should not also hold the executive or senior user role, because that removes the separation between directing and managing; and project assurance should not be held by the project manager, because assurance exists partly to check the project manager's work. Within those limits, almost any combination is possible.

The project assurance arrangement is a good example of tailoring that follows risk rather than size. A small project would normally rely on the executive for assurance, but Alicia has neither the safety expertise nor the time to check ventilation calculations. She asks the city safety officer to provide assurance on safety, a few hours of reviewing the design and inspecting before launch, and the finance analyst to check spending once a fortnight.

Combining and shortening management products

The makerspace has a short list of management products, most of them merged or shortened. What matters is that someone joining the project could find the business case, the plan, the risks, the issues and the controls within a few minutes.

budget, tolerance = 85.0, 0.05     # thousands; +5 % cost tolerance
plan = [4, 6, 10, 18, 30, 44, 58, 70, 80, 85]          # cumulative plan by week
actual = [4, 6, 11, 20, 33, 48]                         # cumulative actual to week 6
forecast_rest = [plan[i] - plan[5] for i in range(6, 10)]
eac = actual[-1] + forecast_rest[-1] * 1.08             # remaining work 8 % dearer
limit = budget * (1 + tolerance)
print(f"spent to week 6: {actual[-1]:.0f}k vs plan {plan[5]}k")
print(f"estimate at completion: {eac:.1f}k; limit {limit:.2f}k")
print("action:", "raise exception to executive" if eac > limit else "within tolerance; report in highlight")

spent to week 6: 48k vs plan 44k
estimate at completion: 92.3k; limit 89.25k
action: raise exception to executive

At week 6 the refit has spent 48 thousand against a plan of 44, mostly because the ventilation duct needed a longer route. The forecast at completion is about 87.8 thousand, inside the 89.25 thousand limit set by the 5 % cost tolerance. Dev reports the position in his next email highlight and takes corrective action within his authority: he moves the vinyl cutter purchase to the library's equipment budget, as Alicia had already allowed for small items. No exception is needed, but the tolerance did its job: it told everyone in advance when Alicia would need to be involved.

Tailoring the practices to the risks

Most practices on the makerspace are light. Plans are one chart. Issues go to Dev by email and into the workbook. Progress is controlled by the cost and time tolerances and the fortnightly email. Two practices are deliberately heavier than the project's size would suggest. The risk practice includes a proper safety risk assessment for each machine, done with the safety officer, because a fire or injury would close the space and harm patrons. The quality practice includes a formal acceptance inspection of the ventilation and electrical work before any equipment is switched on.

Because the makerspace is a public space, the quality criteria include accessibility: the ADA Standards for Accessible Design set expectations for routes, workbench heights, reach ranges and controls in public facilities. The library's own child safety policy adds rules for youth sessions.

Closing a small project

Closure on a small project is often one meeting and one short report, but the purposes of closing a project still have to be met: confirm acceptance of the products, hand them over to the people who will run them, review the project's performance, record follow-on actions and lessons, and plan any post-project benefits review. The makerspace's end project report notes the cost (87.1 thousand against 85), the launch in week 10, two follow-on actions (a second extraction fan if laser use grows, and weekend staffing) and a benefits review at six months to count bookings and youth sessions.

When a small project grows

Small projects do not always stay small. Halfway through the makerspace refit, the city offered Harbor City a grant of 60 thousand for a second makerspace at a suburban branch, provided both opened in the same quarter. Accepting the grant would have almost doubled the project's cost, added a second site and a second facilities crew, and brought a grant agreement with its own reporting. Dev did not simply bolt the second site onto the existing plan. He raised an issue, described the effect on cost, time, risk and benefits, and asked Alicia whether the change should be handled within this project or as a separate one.

Alicia took the question to the library's leadership team, acting as the business layer, because the grant changed the project's scope beyond the tolerances they had set. They decided to treat the second site as a separate small project, run by another branch manager, using Dev's PID as a starting point. The decision kept the makerspace refit's tailoring intact and avoided stretching a light framework over a larger, more complex job.

The lesson for small projects is that light tailoring is a decision for a known context. When the context changes, the tailoring must be reviewed, and sometimes the right answer is a new project rather than a heavier version of the old one.