An engineering lead buys a tool, and legal finds out three weeks later that the vendor owns the telemetry data. Here is the sequence that prevents it.
An engineering lead needs a monitoring tool. She finds one, likes it, and asks finance to pay the invoice. Finance asks for a purchase order. There isn't one. Legal finds out three weeks later, when the vendor's terms turn out to include a perpetual licence to the telemetry data flowing through the product. The tool is already in production. Removing it means a sprint nobody budgeted for.
Nothing here involved bad faith. Every person acted reasonably given what they knew. The failure was structural: there was no sequence, so the controls fired in the wrong order — after commitment rather than before it.
The procurement process exists to prevent exactly this. Its value is not bureaucratic thoroughness; it is that each step surfaces a specific category of problem at the point where fixing it is still cheap. If you want the broader picture of what the function covers, start with what procurement is. What follows is the operational sequence.
Someone in the business needs something. The first job is not to find a supplier — it is to confirm the need is real, properly scoped, and not already met.
This step gets skipped almost universally, and skipping it is the origin of a large share of wasted spend. The questions worth asking take minutes:
A validated requisition with a defined need, a budget owner, an approximate value, and a timeline. The value estimate matters more than it appears — it determines which path the request takes through everything that follows.
Turn the need into requirements precise enough that different suppliers can be compared meaningfully, and loose enough that you are not accidentally writing a specification only one supplier can meet.
That tension is the whole craft of this step. Over-specify and you eliminate competition and pay for it. Under-specify and you receive incomparable proposals and end up choosing on presentation quality.
For services, this is where a statement of work takes shape, and scope ambiguity here becomes scope disputes later. Acceptance criteria deserve particular attention: how will you know the work is done?
Build a shortlist. Sources include existing approved suppliers, market research, industry analysts, peer referrals, and incumbent alternatives.
Qualification runs in parallel and asks whether a supplier is viable before you invest in evaluating their proposal:
Three to five serious candidates is usually the right number. Fewer limits leverage; more imposes cost on your team and on suppliers, who eventually stop bidding for processes they perceive as theatre.
Formally solicit proposals. The instrument depends on what you are buying and how well you understand it:
Give suppliers the same information at the same time. Set a clarification window and circulate all questions and answers to every participant. Allow enough time — compressed deadlines produce padded pricing, because suppliers price uncertainty. Score against the criteria you defined in step two, using more than one evaluator, and document the reasoning. In regulated and public sector contexts this documentation is not optional; award decisions can be challenged.
The step with the widest gap between effort invested and value captured. Most negotiation attention goes to price, which is often the least valuable thing on the table.
Then execute properly: correct legal entities, authorised signatories, all schedules and exhibits attached, blanks filled. A signature from someone without authority produces a document whose enforceability is uncertain — one of the essential elements of a valid contract that commercial teams most often overlook.
The transactional core, and the part most amenable to automation.
A purchase order is raised against the agreement, referencing it explicitly. The supplier delivers. Goods or services are receipted against the order. The invoice arrives and is matched — conventionally a three-way match between purchase order, receipt, and invoice — and paid on the agreed terms.
The step that gets least attention and generates the most avoidable loss. Everything to this point produced an agreement. This step is where the agreement either delivers what it promised or quietly fails to.
The structural reason this fails is that the contract stops being usable the moment it is signed. It becomes a flat PDF, its obligations buried in prose, its amendments filed separately, its renewal date known only to whoever negotiated it — who may have left. Disciplined contract obligation tracking is what closes the loop, and it is far cheaper to start at signature than to reconstruct two years later.
Running all seven steps at full rigour on every purchase is a good way to make procurement universally resented and universally bypassed. Tier it.
Set the thresholds explicitly and publish them. An unpublished threshold is a threshold people learn by violating.
The common sequence is: identify and validate the need, specify requirements, identify and qualify suppliers, run a sourcing event, negotiate and contract, order and receive and pay, then manage the supplier and contract ongoing. Some models split or combine steps — you will see five-step and nine-step versions — but the substance is consistent. The final step is the one most often omitted from both diagrams and practice, and it is where the value negotiated in earlier steps is either realised or lost.
An RFI gathers market information when you do not yet know what solutions exist or what they cost; it is exploratory and does not lead directly to award. An RFP is used when the outcome is defined but the approach is open, inviting suppliers to propose how they would deliver it — typical for services and complex solutions. An RFQ is used when the specification is precise and you are comparing price and commercial terms on a known item. Using an RFP where an RFQ would do wastes everyone's time; the reverse produces incomparable responses.
It varies by an order of magnitude with value and complexity. A catalogue purchase against an existing framework can complete same-day. A competitive tender for a significant services contract commonly runs three to six months from need identification to signature, with the sourcing event and negotiation consuming most of it. Public sector procurement is typically longer due to mandated notice periods and standstill requirements. The most common cause of delay is not the process itself but waiting — documents sitting with approvers who do not know it is their turn.
A control that compares three documents before an invoice is paid: the purchase order (what was ordered), the goods receipt (what arrived), and the invoice (what is being charged). All three must agree within tolerance. It prevents payment for goods never received, quantities never ordered, and prices never agreed. It works poorly for services, which have no physical receipt event, so services procurement needs an explicit acceptance step to serve the same function.
Because notice windows are calculated from the effective date rather than the signature date, are often short, and open well before the renewal itself — and because the obligation to act lives in a clause inside a PDF rather than in any system that generates a reminder. The person who negotiated the agreement frequently knows the date and equally frequently has moved on. The fix is structural: extract renewal and notice dates at execution and calendar them, rather than relying on memory or on rereading the contract.
Maverick spend is usually a symptom rather than a discipline problem. People bypass processes that are slower than their need. The durable fixes are making the compliant path genuinely fast for low-value purchases, setting and publishing clear thresholds so people know which path applies, providing catalogues and pre-approved suppliers so the easy option is also the compliant one, and making spend visible so the scale of the problem is known. Enforcement alone, without addressing cycle time, tends to move the behaviour rather than stop it.
HERO keeps supplier agreements structured after signature, so the terms procurement negotiated stay queryable rather than dissolving into a PDF. Renewal windows, price escalators, service levels, and reporting obligations remain connected to the clauses that created them, through every amendment — which is what makes step seven possible at all. Book a demo.