The software RFP, rebuilt for real price discovery.

The pattern vendors expect.

A typical enterprise software RFP runs as a 12-week process — discovery, requirements, vendor outreach, response evaluation, demos, scoring, shortlist, BAFO, award. The shape of it is so familiar that vendor capture-team leaders model the customer's response curve before bid submission and time their concessions to the milestone calendar. The result is a process that looks competitive but is engineered to produce a known outcome: the incumbent wins on a small discount, the runner-up provides covering bid data, and the customer reports a 7–12% saving against list that the vendor already had in its bid model.

In our experience across 340+ engagements, the RFPs that actually shift price share four characteristics. They include a credible alternative vendor with an operational fit, not just a paper presence. They name the unbundling levers in the requirements rather than letting the vendor bundle around them. They lock contractual terms before commercial terms — not after. And they treat the BAFO as the second-to-last round, not the last.

The competitive alternative test

A vendor reads an RFP's alternative list before reading the requirements. If the alternative is a credible production replacement — same workload class, same regulatory acceptance, named reference customers in the buyer's industry — the vendor responds with its A-team pricing. If the alternative is a "best-of-breed" boutique with no operational footprint at the buyer's scale, the response goes through the standard channel. The decision a procurement lead makes when selecting the alternative — six weeks before pricing arrives — sets the ceiling of what the RFP can achieve.

What the document actually has to do.

The RFP document is the single largest source of customer leverage in the entire procurement cycle, and most teams underuse it. The vendor response is partly a sales pitch and partly a legal commitment — the requirements language that survives into the master agreement is the language the customer wrote first in the RFP. Three structural choices compound:

First, separate functional requirements from commercial requirements from contractual requirements. Vendors will give ground on commercial terms (discount, ramp, payment) and on functional requirements (feature roadmap commitments) much faster than on contractual terms (audit clauses, exit terms, price protection). Mixing them invites trading.

Second, include a deal-shape question. Most RFPs ask for a single price; this hides the fact that the same software comes in five different commercial constructions — three-year enterprise agreement, consumption-based, named-user perpetual, hybrid, marketplace. Asking each vendor to price all five reveals which construction the vendor wants to sell, which often differs from what the customer should buy.

Third, embed the contract draft. Attaching a customer-paper MSA to the RFP and requiring vendor markup as part of the response shifts the negotiation calendar by 60–90 days and prevents the late-stage "legal review" stalling tactic. The customers who run this play save not only on price but on cycle time — carrying that vendor markup straight into buyer-side contract negotiation is what converts RFP leverage into signed terms.

Requirements written for leverage

Specific requirements outperform general ones because they prevent the bundling that bloats the price. A requirement that says "the system shall support 12,000 named users" is bundled. A requirement that says "the system shall support 12,000 named users priced at named-user volume with separate consumption pricing for a list of 14 named integration accounts" is unbundled. The second version produces a 10–18% lower price in most enterprise SaaS RFPs we have benchmarked.

What happens inside the vendor.

Understanding how vendors process an RFP is half the battle. At Oracle, Microsoft, SAP and the major SaaS vendors, an inbound RFP creates a deal record that runs through a deal-desk workflow with specific approval gates. The customer's leverage is determined by which gate the deal reaches. A standard discount the rep can sign off needs no approval. A non-standard discount needs deal-desk approval. A non-standard commercial term — credits, ramp, future-state pricing protection — needs finance approval. A non-standard contractual term needs legal approval. Each gate adds weeks to the cycle and increases the cost of "no" inside the vendor.

RFPs that reach only the rep gate produce single-digit-percentage discounts. RFPs that force deal-desk involvement produce mid-teens. RFPs that force legal and finance involvement — usually through contractual terms the rep cannot sign — produce 25–40% discounts off list. The lever is the language in the RFP, not the volume in the deal.

Reading the response document

Vendor RFP responses follow a tactical pattern. The executive summary is a sales document. The functional response is largely accurate but optimistically scoped. The commercial response is anchored to a price the vendor wants the customer to accept as the reference point. The contractual response is where the real signal sits — what the vendor agreed to, what it pushed back on with explanation, what it pushed back on with stock language. The stock pushbacks are the negotiation surface; the explained pushbacks are real constraints.

BAFO is round two, not the final round

"Best and Final Offer" is the line in the procurement process where buyers most often hand back the savings they have already earned. A well-structured BAFO is followed by a "deal-shape exception" round, in which the customer asks specific structural questions (a ramp curve change, a price-protection clause, an exit window) that the vendor still has authority to grant after the headline price is anchored. This round typically produces another 4–8% of value on top of the BAFO.

Where good RFPs go wrong.

Three failure modes recur. The first is the late-stage scope creep, where the vendor adds in-scope items after BAFO that change the comparison. The defence is a frozen scope baseline and a strict change-control protocol that runs through procurement, not through the relationship owner.

The second is the executive-relationship intervention, where the vendor escalates to the customer's CIO or CFO directly, bypassing procurement. This is engineered, not coincidental. The defence is a pre-agreed escalation discipline: the executive sponsor receives no commercial-content meetings during the active procurement window, only progress updates.

The third is the contractual swap-back, where the vendor agrees to customer-paper in the RFP response and then re-papers to vendor-paper after award. The defence is to sign the negotiated MSA at award, not before contract preparation begins.

When to use independent advisory support