The labour figures published a fortnight ago put job openings near eleven million and the number of people voluntarily leaving their jobs at around four million in a single month. Both are records. The phrase coined in the spring for this — the Great Resignation — has stopped being a commentary and started being a staffing plan problem. Which lands precisely now, because this autumn is when the enterprise projects deferred in 2020 are restarting. Budgets were released in the summer, steering committees have reconvened, and implementation partners are being asked to mobilise against plans written before any of this happened. Here is the uncomfortable truth about those plans. An ERP implementation does not depend on forty named resources in a mobilisation chart. It depends on about six people, and nobody has written down who they are.
The six people
Strip any mid-sized implementation to its load-bearing members and you find roughly the same set. The finance lead who knows why the chart of accounts looks the way it does, which cost centres are fictional, and which reports the board actually reads. One super-user per major process — order to cash, procure to pay, inventory — who knows the exceptions that make the standard process wrong. The partner's functional consultant who configured tax, pricing and costing, and who holds the reasons in their head rather than in the design document. The integration developer who built the connections to the bank, the warehouse and the customer portal. The data migration specialist who understands why four thousand customer records have to be merged in a particular order. And a project manager who knows which commitments are real. Lose two of those six mid-project and you lose months, not weeks. Lose the wrong one after go-live and you lose the ability to explain your own system. In 2021 the probability of losing two of them during a nine-month programme is no longer remote, and the plan on the table almost certainly assumes it is zero.
Three different losses, three different remedies
Partner-side turnover. Your consultant resigns, and the partner substitutes somebody who reads the design document and then asks the business the same questions it answered four months ago. The programme absorbs six weeks and the client absorbs the frustration. The remedy is contractual and has to be agreed before signature. Client-side turnover. The super-user who wrote the test scripts and knows every exception hands in their notice three weeks after go-live — often because a twelve-month implementation is the best line anyone can add to a curriculum vitae, and the market is paying for it. The remedy is partly money, mostly documentation, and entirely foreseeable. Internal promotion. The strongest project person is given a bigger role and becomes notionally available, which in practice means present at steering meetings and absent from the work. This loss is the most common and the least acknowledged, because nobody left.
| Risk | Practical response |
|---|---|
| Partner-side turnover | Agree substitution notice, replacement review, paid overlap and documented handover |
| Client-side turnover | Fund backfill, retain decision reasoning and test scripts, and name a deputy |
| Internal promotion | Treat reduced availability as a staffing change rather than assuming the person remains available |
Qualitative summary of this article's source text, not a measured outcome or performance estimate.
The contract clauses that earn their place
Most implementation contracts treat people as a rate card. A 2021 contract should treat them as the scarce input they are. Name the key personnel individually, with allocation percentages and start dates, and make substitution subject to the client's consent. Require thirty days' notice of any proposed change and the right to interview the replacement. Require a paid overlap — two weeks minimum — at the partner's cost rather than yours, because the handover is their obligation, not your scope change. Define knowledge transfer as a deliverable with an artefact attached rather than as reasonable endeavours. Freeze the rate card for the programme duration so that scarcity does not become a change request. And require that configuration decisions, including the reason for each one, are recorded in a repository you own and can read after the partner has gone. That last clause is worth more than all the others combined, which brings us to the real point.
Design the programme so turnover costs less
Retention is largely outside your control. Knowledge concentration is entirely within it. Every ERP programme produces a system that nobody can explain, because the explanations lived in conversations. The countermeasure is unglamorous and cheap: a decision log that records what was decided, by whom, and — crucially — why the alternative was rejected. Configuration documented as reasoning rather than as screenshots. Test scripts written by the business and kept as the durable description of how the process is supposed to behave, because they outlive both the consultant and the design document. Two named people per process area rather than one, even if the second is only shadowing. And shorter phases, so that knowledge is consumed and embedded before it has time to evaporate. Organisations that do this find that losing a consultant becomes an inconvenience. Organisations that do not find that losing a consultant becomes an archaeology project.
On client-side staffing, honestly
The standard fiction is that a super-user contributes fifty per cent of their time while continuing to do their day job in full. Everybody signs that plan and nobody funds the backfill, which produces two predictable outcomes: the project gets the evenings, and the super-user — now the most employable person in the finance team, with fresh implementation experience and a market that is paying up — starts taking calls from recruiters. Buy the backfill. A temporary accounts payable clerk costs a fraction of a fortnight's programme slippage. Put the project role in the person's objectives so it is recognised work rather than volunteered overtime. Consider a completion award paid at go-live plus ninety days, which is the window in which their knowledge is irreplaceable. And name a deputy for every key internal role, in writing, at mobilisation.
Practical Guidance for ERP Implementation Support
- Identify your six load-bearing people by name and assess, candidly, the probability that each is still here in nine months.
- Name key personnel in the contract with allocation percentages, substitution consent, notice periods and the right to interview replacements.
- Make handover overlap the partner's cost, with a defined deliverable rather than a promise.
- Keep a decision log with reasons, in a repository you own, updated as decisions are made rather than at the end.
- Fund the backfill for internal super-users, and put the project role in their objectives.
- Double up on every process area, even where the second person only shadows.
- Shorten phases so knowledge is used before it leaves.
- Ask for named consultant profiles with a start-date commitment, and re-verify the team composition at mobilisation rather than at proposal.
The Regional Angle
The labour market pulling at your implementation team here does not look like the one being described elsewhere, and the mitigation differs accordingly. A functional consultant in Dubai is not being tempted by a remote-first employer in another hemisphere, because their residency is tied to employment in this country. What they are being tempted by is the project pipeline in Saudi Arabia, where expansion is absorbing implementation capacity at rates the rest of the Gulf has not had to match. The regional pool also shrank in 2020 — people who left during the pandemic largely did not return — while demand restarted across several markets at once. The consequence is a genuinely thin bench, and a partner who tells you the named consultant is available in October may be telling the truth on the day they say it. Assume movement, price a realistic rate rather than negotiating the last eight per cent out of the proposal, and put the substitution protections in the contract where they can be enforced. The second observation is that this year's constraint is also this region's opportunity, and very few programmes are using it. Nationalisation targets are tightening across the Gulf, and organisations are looking for credible roles in which to develop national talent. An ERP implementation is one of the best structured apprenticeships an organisation will ever run: a defined duration, named process areas, real deliverables, external experts on site, and documentation as an output. Pairing a national graduate with each process area solves two problems with one decision — it builds the capability your quota commitments require, and it eliminates the single-head knowledge concentration that makes turnover expensive. Do it at mobilisation, with the pairing written into the partner's obligations, or it will not happen. The third is a sequencing trap in multi-entity rollouts. Regional groups plan waves — the UAE entities, then Saudi Arabia, then Oman — as though the waves were independent. They are not: the same six people work all of them, so a sequential plan is really a queue against a single resource. This year the Saudi wave carries an additional and scarcer dependency, because statutory electronic invoicing obligations are moving into their integration phase from December, and the specialists who understand that requirement are in far shorter supply than generalist functional consultants. Anyone planning a Saudi go-live in the first half of next year should be securing that specific capability now, separately from the main implementation contract, rather than discovering in February that the statutory work is the critical path.
The objection worth taking seriously
The strongest objection is that this is a human resources problem wearing a programme management costume, and that key-personnel clauses are theatre. Partners breach them routinely. Damages are almost impossible to prove, because you cannot demonstrate what the original consultant would have delivered. Insisting on named individuals gives the partner a convenient explanation for every delay — you refused two substitutions, therefore the timeline slipped — and a client who blocks replacements can stall its own project while the partner bills for the argument. All of that is accurate, and anybody selling named-personnel clauses as enforcement is overselling them. Their value is different. What they buy is information and process: advance notice instead of discovery at a Monday stand-up, the right to meet the replacement before they arrive, an overlap that somebody else pays for, and a handover that produces an artefact. Those change the shape of the loss even when no penalty is ever claimed. More importantly, the objection concedes the central argument by accident. If you cannot control whether people leave, control what leaves with them. The decision log, the reasoning behind each configuration choice, the business-written test scripts and the second named person per process are entirely within your gift, cost very little, and convert a crisis into a delay. Programmes that fail in 2022 will not fail because somebody resigned. They will fail because when somebody resigned, nothing had been written down.
Common Questions
Should we delay our implementation until the market settles?
Delay if you cannot secure your six people, because starting without them is the more expensive choice. Otherwise proceed with tighter scope and shorter phases. Waiting for a settled labour market is not a plan with a date.
How do we stop the bait and switch on consultant profiles?
Ask for named profiles with allocation percentages, verify the team at mobilisation, and make the first invoice contingent on the named team actually starting.
Is paying a retention award to internal staff reasonable?
For the two or three genuinely irreplaceable roles, yes, tied to go-live plus ninety days. It is cheaper than the alternative and it signals that the project role is real work.
What should we expect over the next twelve months?
Expect rates to keep rising into next year and partner benches to stay thin, with more proposals quietly staffed junior and supervised remotely. Expect more fixed-price bids that are really staffing bets, and read them accordingly. Expect the Saudi statutory invoicing programme to absorb a disproportionate share of regional functional capacity from December onward. Expect more organisations to choose smaller scope and standard product configuration precisely because the people needed for heavy customisation cannot be secured. And expect the phrase knowledge transfer to appear in contracts with a deliverable attached for the first time, because clients who lost a consultant this year are not signing the old wording again.
ERP Implementation Support Services — we identify the handful of people your programme actually depends on, write the contract and documentation disciplines that survive their departure, and keep delivery moving when the market moves your team.