The margin is already in the contracts you have signed
Spend is where margin and resilience are actually decided, and in most businesses nobody owns it above the level of a purchase order. This practice takes it on at board level: the categories, the contracts, the supplier you could not replace, and the AI vendors nobody has commercially reviewed.
A margin function, filed under admin
In most businesses between scale-up and mid-cap, spend is managed transactionally. Purchase orders get raised, invoices get matched, and once a year someone is asked to find a percentage. That is an administrative function doing an administrative job well.
What it cannot do is answer the questions that actually sit at board level. Which categories are worth a strategy and which are not. What happens if the supplier you cannot replace fails. Which contracts renew on their own next quarter because renegotiating them takes longer than anyone has. What was committed to in the deal you closed last year, and how hard it is to get out of.
Nobody owns those questions, so nobody answers them, until one of them answers itself.
Four kinds of work, all of them commercial
Every one of these is a commercial judgement before it is a process, which is why they sit at board level rather than inside a category team.
Category strategy and sourcing
Where the money actually goes, which categories justify a strategy, and which sourcing approach fits each one. Consolidation and supplier-rationalisation calls follow from that analysis rather than from a savings target set before it.
Contract, supplier and AI-vendor risk
Renewal timing, exit terms and concentration, including the question most businesses cannot answer about their largest supplier. AI and software vendors are the category procurement functions have caught up to least, and the one signing the least-reviewed contracts.
Tail spend, and the data underneath it
The spend nobody has sized because the data has never been cleaned. Tail and maverick purchasing are usually the quickest thing to fix and the last thing anyone examines. But any analysis is worthless until the classification is trustworthy, so that comes first.
M&A, carve-out and integration
Contract forensics across both sides of a transaction: what has been committed to, what can be exited, where the recoverable value sits and what integration costs commercially. Diligence prices the balance sheet thoroughly and skips the supplier book.
We do not sell what we select
A vendor selection is independent only if the people running it have nothing to gain from the answer. Where we run a selection, we exclude ourselves and our delivery partners from the work it awards, not as a courtesy but because a process whose organiser can win it is not a selection.
That constraint costs us the engagement we would otherwise be best placed to win, which is precisely what makes the recommendation worth having. It is also why leadership and delivery are quoted separately here rather than bundled: a blended rate is a quiet incentive to recommend more delivery.
A baseline before a strategy
Most engagements here open with a maturity review rather than a category plan, because a category plan written against unreliable spend data is a guess with a spreadsheet attached.
The review looks at four things: how spend is captured and classified, how suppliers are managed and how concentrated they have become, how contracts are governed across their lifecycle, and how commercial decisions get made and by whom. It produces a scored baseline and a ranked view of where the exposure and the recoverable margin actually sit.
That is what a board needs before it approves a programme, and often enough to act on without approving one.
Seven areas, scored out of five. Assess only the areas that apply to you, see your rating on screen, and take the detailed PDF if you want it.
Agile sourcing works, and the boundary is the more useful half
Sprint-based sourcing has a published, attributable case record: a signed ERP contract in four weeks, a competitively bid public-sector award in forty-five days, and one documented case where a 9% savings target was delivered at 15%. Compressed cycles are assumed to cost you commercially. That last case shows the opposite, and the mechanism is not mysterious: suppliers solving the real problem in the room propose things a specification written six months earlier had already ruled out.
It does not fit everywhere. Where the process is itself the control, as in some public procurement regimes, regulated financial services and defence, the sequence and its documentation are the point, not overhead to be compressed. Where a requirement genuinely cannot be understood inside a sprint, compressing the cycle compresses the wrong thing. Choosing where the method applies is as much a part of the work as running it, and we will say so when it does not.
Agile Sourcing: the real numbers
Six documented cases, the sources behind every figure, and the regulated categories where the method should not be used at all.
Read the researchWho leads this work
The remit, the four areas it is accountable for, and why we do not badge it "Fractional CPO".
See the leadershipWhere it meets the operating model
Supplier decisions are operating-model decisions. Reconciling the two is what the disciplines are for.
How we reconcile themStart with the spend you cannot see
Supplier concentration, renewal timing, tail spend, or a transaction about to close. A direct conversation about where the exposure is.