What should a category strategy actually contain?

Templates vary enormously in length and almost not at all in what they need to cover.

1. 01Spend and demandWhat is bought, from whom, by which parts of the organisation, at what volume and on what terms, including the spend that is not on contract. Demand matters as much as spend: a category where volume is falling is a completely different negotiation from one where it is growing, and the supplier will know which it is.
2. 02Supply marketWho the credible suppliers are, how concentrated the market is, how suppliers in it make money, what drives their costs, and where capacity sits. The test of whether this section was worth writing is whether it contains something the organisation did not already know.
3. 03Requirement, and where it can flexWhat is actually needed, separated from what is currently specified. State which parts of the specification are fixed, which are preferences, and which exist because a supplier wrote them some years ago. Get that agreed by the budget holder in writing before going to market, because it is the section they will revisit when the price comes back.
4. 04Route to marketCompetition, negotiation with the incumbent, a framework call-off, aggregation with another organisation, or leaving a good contract alone. Say why, and say what the alternatives were. Where the procurement is regulated, the route and the reasoning behind it form part of a record that has to stand up later.
5. 05Commercial model and target positionThe section below. This is the one that decides what the contract is worth after it is signed.
6. 06RiskSupply continuity, financial standing, dependency, data and security exposure, and what happens if the supplier fails. For each, who carries it and what the mitigation costs, because a mitigation nobody has funded is not a mitigation.
7. 07ImplementationThe sequence, the dates, the internal decisions required and a named owner for each, at least one of whom is outside procurement. A strategy that ends at award has stopped before the part where the value is realised.
8. 08ReviewWhen the position will be revisited, and what would trigger an earlier look: a market movement, a supplier event, a change in demand, a regulatory change.

This is the section most often written as a list of nouns, and it is the one that determines whether a good price is still a good price in year three.

Pricing mechanism

Fixed price, rate card, cost plus, capped time and materials, or an output or outcome basis. Each moves risk somewhere. A fixed price where the requirement is genuinely uncertain buys you a contingency you cannot see and a change control conversation every quarter. A rate card where the volume is unpredictable protects the supplier and nobody else.

Indexation

Which index, and applied to what. A published general index, a sector or materials index, or a labour index all behave differently, and applying one to the whole price when only part of the cost base is exposed to it hands the supplier the difference. Say when it is reviewed and how much notice is required.

Cap, collar and floor

Whether indexation is capped, whether it is symmetric so that the price falls when the index does, and whether there is a floor at zero. A larger reduction with an uncapped index is frequently worth less over a term than a smaller one with a cap, and the negotiation should be run on that arithmetic rather than on the headline percentage.

Volume commitment against volume estimate

An estimate buys you nothing. A committed minimum buys price, and it costs you flexibility and creates an exposure if demand falls. Decide which you are offering before the supplier asks, and price the difference.

Cost transparency

Where the category justifies it, what you actually ask for is the build-up: base cost, on-costs, overhead recovery and margin. The margin line is the least useful one to attack because it is usually the smallest. The useful lines are the ones the supplier has room to be wrong about.

Term, break and exit

Length, break points and what happens on exit, including transition assistance and what data and assets come back. A long term is a concession and should be traded for something rather than granted because it makes the business case work.

The target position is the second half of this section: what you intend to achieve and what you are prepared to concede, written down before the negotiation. Concessions decided in advance are decisions. Concessions decided in the room are reactions.

The headline percentage is negotiated once. The indexation clause is negotiated every year, by nobody.

Where the category is going to market, the strategy fixes how the winner will be chosen, and that has to be decided before the documents are drafted rather than during them.

• The price and quality split. Moving from seventy thirty to sixty forty changes who wins, and the number should follow from what the category actually needs rather than from what was used last time.
• How price is scored. Scoring against the lowest bid, against the mean, or on a fixed scale produces different winners from identical submissions, and the method should be chosen with a view to how bunched the prices are likely to be.
• What the quality criteria are actually testing, and whether the answers can be evidenced or only asserted.
• How an abnormally low bid will be handled, decided before one arrives rather than after.
• Whether the market is likely to produce a real field. In specialist categories a single compliant bid is a common outcome, and a strategy that assumes three bidders and gets one has no fallback unless it was written with that possibility in it.

A category strategy earns its time only if it changes a decision. Name the decision and the person who is going to make it. If the answer is that it informs future thinking, it is analysis with a cover page.

Useful documents in this format change one of five things: what is bought, because the requirement was carrying cost nobody had questioned; how it is bought, because the current route produces a worse answer than an available alternative; how many suppliers it is bought from, in either direction; what the contract says, because the commercial model does not match how the category behaves; or when it is bought, because the timing of a renewal is worth more than the negotiation.

Length is a symptom rather than a target. A short main document that states the position and the plan, with the analysis held separately, is read at the point of award. A long one is not, with one legitimate exception: where the strategy doubles as part of a regulated procurement's decision record or an investment case, it has a second job and a different reader, and that justifies the length.

1. 01The specification is taken as givenIn many categories the requirement carries cost nobody has questioned since it was first written, and in labour-based categories the equivalent is the rate structure and the volume of demand rather than the specification itself. Either way, a strategy that does not touch what is being bought is working on the smaller half of the problem.
2. 02Written without the budget holderA strategy produced inside procurement and presented at the end will be negotiated at the end, when there is no time left to negotiate anything.
3. 03A portfolio matrix used instead of a decisionPlacing a category in a quadrant describes it and does not tell you what to do about it. The value of the exercise is that it forces supply risk into a conversation that was only about price. Take that and leave the quadrant.
4. 04Aggregation assumed to be the answerConsolidating volume helps where the market is fragmented and the requirement is common. It fails where suppliers price on ability to serve rather than volume, and where a single large lot exceeds what most of the market can deliver, which shrinks the field rather than sharpening it.
5. 05No position on renewalMost categories are re-bought rather than bought. A strategy with nothing to say about the next renewal point has to be rewritten before it is used.