Few organisations have the spare capacity to run a change like this entirely alone, so external help is often sensible. But the series’ founding question applies to the delivery model too: does the help you bring in build a capability that stays — or quietly become the next thing you can’t live without? Scaffolding should come down. The building should remain.
The roadmap flagged “delivery dependence” as a trap. This is where we design it out — by getting honest about consultants.
It can be. The default consulting model — they design it, they build it, they run it — produces a polished result and a quiet problem: when they leave, the capability leaves too. You’re left with a deck, a system you don’t fully understand, and a standing invitation to call them back. That is delivery dependence, and it’s the rescue cycle wearing a different badge.
External help isn’t the enemy — renting capability instead of building it is. The same money, spent differently, can leave your people more capable than it found them. The distinction is everything.
External support runs from genuinely capability-building to quietly captivity-creating. The art is matching the model to the moment — and steering away from the kinds that never let you go.
Experts who upskill your people, set up the approach, then deliberately work themselves out of a job. You do the work; they make you better at it.
Builds what stays ✓
Short, sharp expertise for a specific gap — a tricky governance question, a security review — scoped to a deliverable and a clear end.
Bounded, leaves cleanly ✓
Extra hands working alongside yours when capacity is short. Useful — but only if your people lead and learn, not just supervise from a distance.
Fine, if you stay in the lead ⚠
Hand the whole thing over and receive a finished result. Fastest to a deliverable, surest route to dependence. The capability never becomes yours.
Rents, never builds ✕
There’s a simple way to tell a capability-builder from a dependency-creator: ask, at the very start, how the engagement ends. The good ones have a clear answer. The trap ones change the subject.
Even with external help engaged the right way, ownership must sit firmly inside. Here’s the division that keeps capability in-house and help in its proper place.
The temptation is always to let the experts handle it — it’s faster, easier, and lower-effort today. That temptation is exactly the one the whole series exists to resist. Two disciplines keep external help honest.
Capability transfer is the deliverable — name it as one. Write it into the engagement: alongside the system or the design, the explicit deliverable is your people able to run it without them. Measure the engagement’s success by what your team can do once it ends, not just by what was shipped. If skill transfer isn’t a named outcome, it won’t happen by accident.
The same test as everything else in the series. Does this build a capability that stays, or a dependence that lingers? Applied to consultants, the answer decides whether external help advances the mission or quietly betrays it. Help that makes you more dependent has failed, however polished its output. The goal was never a finished change delivered to you — it was an organisation able to change itself.
That is the difference between scaffolding and a crutch. One comes down when the building stands; the other you lean on forever. Choose the kind of help designed to make itself unnecessary — exactly as the change itself must.
A strategic discussion series making the psychological and strategic case for AI augmentation in HR. If it landed for you, the next conversation is about your function.
Start the conversation →