Sequence transformation programmes; do not run them in parallel
Mid-tier firms running multiple concurrent transformation programmes hit a coordination ceiling that makes any single programme stall; complete critical migrations before adding AI complexity.
A common pattern at mid-tier firms is to run multiple transformation programmes concurrently — practice management migration, document management upgrade, governance refresh, AI adoption, IT infrastructure renewal — on the reasoning that each is independently necessary and waiting risks falling further behind. The reasoning is correct on each programme considered alone. It misreads the constraint that determines whether any of them succeed.
The constraint is shared. Coordination capacity, change-fatigue tolerance and senior attention are all drawn from the same fixed pool. A firm running three transformations in parallel is not running three programmes at one-third pace; it is running three programmes that compete for the resources each requires to clear its critical-path bottlenecks. The result is that all three slip, and at least one of them stalls long enough that earlier momentum is lost.
The heuristic is to sequence rather than parallelise. Identify the critical-path programme — usually the one with the most binding deadline, or the one whose completion enables the others — and run it to completion before adding the next. The cost is that some programmes wait. The benefit is that the ones that proceed actually finish.
Two cautions apply.
First, sequencing is not single-threading. Some programmes can run as background workstreams that do not require the same intensity of attention as the critical-path one. Knowledge management investigation, vendor research and governance drafting can sit in parallel with a critical-path implementation; the rule applies to programmes that require active change management, not to all activity.
Second, the discipline is hardest at executive level. Pressure to “show momentum” usually arrives as a request to start everything, and the executive sponsor who insists on parallelisation against the working team’s advice is typically right that nothing is moving — and wrong that starting more things will fix it. The honest framing is that the firm has chosen which programme to slip; sequencing makes that choice explicit rather than letting it happen by accident.
Sequencing aligns with the broader The mid-tier AI adoption threshold dynamic at the programme-portfolio level: payoffs only clear the BAU-pressure threshold if attention is concentrated. Spreading attention across programmes guarantees that none of them clears it.