Skip to content

Vendor Selection and RFP Project Management Explained

8 min read · The Eddie System

Almost every organization runs competitive sourcing continuously, and almost nobody is taught how. Here is how an RFP actually works, what the project manager owns, why these projects fail quietly rather than loudly, and how to practice one before you are responsible for the award.

What Vendor Selection Project Management Actually Is

Vendor selection is the project of choosing who a company will buy something from, and running it so the decision holds up afterwards. The formal instrument is usually an RFP (Request for Proposal) — a document setting out what the organization needs, issued to a field of suppliers, whose responses are scored against criteria agreed in advance.

It sounds administrative. It is not. A sourcing project decides where money goes for the next three to five years, which systems the business runs on, and how expensive it will be to change course later. The technology choices made in an RFP outlive most of the people who made them.

What makes it distinctly hard is that the project manager has almost no formal authority over the outcome. You do not pick the winner — a scoring model and an evaluation panel do. You cannot compel the internal stakeholders whose requirements you need, and you cannot compel the suppliers whose bids you are waiting on. Your job is to design a process fair enough to survive being challenged, and then hold it steady while people with more seniority than you push on it.

Every organization above a certain size runs these continuously — technology platforms, facilities, logistics, professional services — and most of them are run by project managers rather than career procurement specialists.

How an RFP Actually Runs, Stage by Stage

A competitive sourcing event is stage-gated by its own integrity rules, and the stages are less negotiable than they look:

  • Requirements gathering — Collect what the business actually needs from every group that touches the thing being bought. This is the messy, political part, and it is where the project is usually won or lost.
  • Requirements freeze — Close the requirements set. Nothing enters afterwards. The freeze exists so every supplier answers the same question.
  • Scoring model design — Agree the criteria, their weights, and the anchors that say what a given score means. This must be fixed before anyone sees a bid.
  • Issue and moderated Q&A — Release the RFP. Suppliers ask questions through one controlled channel, and every answer goes to all of them simultaneously.
  • Sealed bid open — Responses are opened together, usually with an independent witness, and no evaluator sees anything early.
  • Evaluation — Panel members score independently first, then moderate. The disagreements are recorded, not smoothed away.
  • Reference calls — Speak to the suppliers' existing customers. This is the only stage where you hear from someone with nothing to sell you.
  • Negotiation and award — Legal and commercial close the terms; the recommendation goes to whoever holds the decision.

The sequence matters more than any individual step. Requirements freeze before issue so the field is comparable. The model is fixed before bids are visible so nobody can shape the criteria around a preferred answer. These are not bureaucratic flourishes — they are what makes the eventual decision defensible.

Why Sourcing Projects Go Wrong

Sourcing projects rarely fail loudly. They fail quietly, months later, and the causes repeat:

  • Requirements were never really agreed. Signatures were collected from people who had not read the document. The disagreement surfaces after award, when it is expensive.
  • The scoring model rewards the wrong thing. Weighting price too heavily reliably selects a supplier that is cheap to enter and expensive to leave. The saving shows up this year; the lock-in shows up in three.
  • An accurate bid gets misread. A supplier states a real limitation plainly, an evaluation panel reading forty responses under time pressure registers the headline and misses the qualifier, and nobody notices until a reference call.
  • Exit terms go unpriced. Egress fees, minimum terms, and data-portability limits are rarely scored, and they determine what the next change costs.
  • The people who have to live with it were not consulted. The users of the chosen system have no formal veto and total practical power to make adoption miserable.
  • The process bends under seniority. A late requirement arrives from someone important, the freeze reopens, and the competition is no longer comparable.

Notice that most of these are not procurement failures. They are project management failures wearing procurement clothing: requirements discipline, stakeholder consent, and the nerve to hold a boundary when the person pushing on it outranks you.

What the Project Manager Owns in Vendor Selection

On a sourcing project the PM does not choose the supplier and does not sign the contract. What the PM owns is everything that determines whether the choice is any good:

  • The requirements set — driving four or five groups with different interests to one signed document, before the freeze, while disagreement is still cheap.
  • The evaluation design — criteria, weights, scoring anchors, and panel composition, all fixed before any bid is visible.
  • Process integrity — the supplier-contact log, the moderated Q&A, the sealed bid open, and the evaluation record that shows how a panel got from disagreement to a number.
  • Evidence over assertion — asking suppliers to demonstrate the capability that matters rather than describe it. A vendor's claim and a vendor's demonstration are different things.
  • Stakeholder consent — securing agreement from the people who will use the thing, before the decision rather than after.
  • The defensible record — a file that survives being read closely by a losing supplier who wants to know why.

That last point is the one newcomers underestimate. In delivery projects, documentation supports the work. In sourcing, the record *is* the work: compliance is demonstrated through what was written down, and a project manager who behaved impeccably and logged nothing is indistinguishable from one who did not.

It is also judgment work, and judgment is difficult to learn from a textbook. You learn it by making the calls — whether a late bid gets opened, whether a reference-call finding can touch a closed scorecard, what to concede in a negotiation and what to walk away over. For more on the underlying skill, see stakeholder management for project managers and how to manage project risk.

Practice a Vendor RFP Before You Run a Real One

The fastest way to build sourcing judgment is to run a competition end to end, without a real supplier relationship riding on your first attempt. That is what The Eddie System is built for: realistic 27-day simulations where you act as the project manager from initiation through award.

Each simulation is a fictionalized scenario inspired by a real company and project type. You make the decisions a sourcing PM actually faces — what enters the requirements freeze, how heavily to weight price, whether to open a strong bid that arrived late, what to do when a reference call contradicts your scorecard — and a live project-health dashboard tracks budget, schedule, scope, risk, and stakeholder sentiment as the consequences accumulate. Named stakeholders push competing agendas, and phase gates force the same discipline a real programme demands.

For sourcing specifically, start here:

Want the adjacent disciplines? Vendor management runs through much of the catalog: try a treasury management system implementation at BP or an Oracle Cloud ERP rollout at Starbucks, both realistic simulations where managing the supplier is as decisive as managing the build. You can browse the full set on the vendor selection and sourcing hub.

Every completed simulation leaves you with real PMO deliverables — a charter, plan, SteerCo deck, and closure documents — plus a verified completion record for your portfolio. No PMP or prior experience required.

Try the free first day with no account to see how it works, browse the full simulation catalog, or check pricing when you are ready to run a sourcing project end to end.

Frequently asked questions

What does a project manager do in vendor selection?

The PM owns the process rather than the choice. That means driving stakeholders to a signed requirements set, designing the scoring model and fixing it before bids are visible, running the moderated Q&A and sealed bid open, keeping an evaluation record that shows how the panel reached its scores, and securing consent from the people who will use whatever is bought. The supplier is chosen by the model and the panel; the PM makes sure that choice is defensible.

What are the stages of an RFP process?

Requirements gathering, requirements freeze, scoring model design, issue and moderated Q&A, sealed bid open, independent evaluation and moderation, reference calls, then negotiation and award. The order is the point: requirements freeze before issue so every supplier answers the same question, and the scoring model is fixed before anyone sees a bid so criteria cannot be shaped around a preferred answer.

Why do vendor selection projects fail?

Usually quietly and months later. The common causes are requirements that were signed but never genuinely agreed, a scoring model that over-weights price and selects a supplier who is cheap to enter and expensive to leave, exit terms nobody priced, an accurate bid whose limitation the panel misread, and users who were never consulted and have no reason to adopt the result. Most of these are project management failures rather than procurement ones.

Do I need a procurement background to run an RFP?

No. Most competitive sourcing in large organizations is run by project managers rather than career procurement specialists, with commercial and legal colleagues owning the contract itself. What you need is requirements discipline, evaluation design, and the judgment to hold a process boundary under pressure — all of which can be practised before you are responsible for a live award.

How can I practice vendor selection project management without a real job?

Run a sourcing simulation. On The Eddie System you act as the project manager across a 27-day fictionalized RFP, deciding what enters the requirements freeze, how the scoring model is weighted, and what to do when a reference call contradicts a locked scorecard. You finish with real PMO deliverables and a verified completion record for your portfolio. Try the free first day at /demo or browse sourcing simulations at /explore/type/vendor-selection-sourcing.

Start building real PM experience

Run a 27-day project management simulation at a real company — and walk away with proof.

Keep reading