8 questions from the RFI stage, free
These come from the module as sold. The workbook adds follow-ups, a response format, a weight and a score column to each.
1. Provide a URL to your public pricing page. If specific enterprise pricing applies, or if pricing is not public, attach your standard rate card.
Why it matters. Published pricing shortens procurement cycle time. For fair shortlisting and TCO modeling, buyers need a written rate reference and cannot rely solely on sales-gated quotes, which create information asymmetry and downstream renewal leverage problems.
- Provides a direct link to a public pricing page with current rates.
- Attaches a standard enterprise rate card.
- No NDA is required for baseline rate disclosure.
- No published pricing at any tier; requires sales call for all pricing.
- Refuses to provide a standard rate card attachment.
- Public pricing exists but is materially different from what is quoted.
2. State your standard initial contract term length and the minimum term length you are willing to offer.
Why it matters. Term length is a primary lever for buyer flexibility. Vendors that require multi-year commitments to unlock baseline pricing constrain procurement options and increase exit risk.
- Offers a 12-month term as a standard option.
- Discloses whether shorter pilots (30/60/90 day) are available.
- Multi-year terms are optional, not required for standard pricing.
- Three-year minimum with no shorter option at any price.
- Pilot terms are blocked behind a multi-year commitment.
- Refuses to quote month-to-month at any rate.
3. Describe your volume-discount tiers, including the consumption thresholds at which discounts apply and the discount percentage at each tier.
Why it matters. Discount structure transparency lets buyers model breakpoints and avoid stranded commitments. Opaque or sales-only discount discussion creates information asymmetry favoring the vendor.
- A published or shareable tier schedule is provided.
- Thresholds are defined in concrete units (tokens, seats, $).
- Tier method is stated: all units repriced at the new tier, or only units above the threshold.
- Discounts are only available via 'contact sales' with no schedule.
- Tier thresholds change between customers without rationale.
- Tier treatment (all units repriced at the new tier, or only units above the threshold) is not stated.
4. Provide a comprehensive list of all billable consumption units and all separately priced offerings. For each item, specify the unit of measure (e.g., per token, per GB, per seat) and state whether it is included in a base subscription or priced as an add-on. This list must include, but is not limited to: all types of token consumption (input, output, cached), embeddings, fine-tuning, storage, data egress, premium support tiers, professional services, dedicated capacity, and onboarding fees.
Why it matters. Headline pricing for AI tools may not show the full cost. Enumerating every billable unit and add-on cost category up front exposes hidden cost surfaces, prevents post-signature surprises, and supports accurate Total Cost of Ownership (TCO) modeling.
- Provides a comprehensive table enumerating all billable axes.
- Explicitly identifies items priced as add-ons (e.g., egress, support, pro-serv).
- Unit of measure and per-unit price are clear for each item.
- Asserts 'all-inclusive' pricing when the product has clear add-on features.
- Omits common hidden costs like data egress, storage, or premium support.
- Uses vague phrases like 'usage-based' without enumeration.
5. Confirm whether pricing differs by deployment region or data residency zone, and if so, attach a region-by-region price differential summary.
Why it matters. Some AI vendors charge more for specific regions, data-residency options, or sovereign and air-gapped deployments. Buyers with mandated residency need this differential disclosed before contract signature.
- Clear yes/no on regional differentials.
- Lists each region with applicable surcharge or notes parity.
- Identifies any region-exclusive SKUs.
- Claims parity but cannot produce a regional schedule.
- Surfaces regional uplifts only late in negotiation.
- Sovereign deployment pricing is 'on request' with no indicative range.
6. Describe your renewal mechanics, including whether renewal is auto-renewing by default, the notice period required to opt out, and any price-uplift cap at renewal.
Why it matters. Auto-renewal with an early opt-out deadline and no uplift cap can renew the contract at a higher price before the buyer acts. Buyers need explicit renewal terms before signature, not at year-end surprise.
- Auto-renewal is opt-in or easily disabled.
- Opt-out notice period is stated in the contract and fits the buyer's renewal review cycle.
- Renewal price uplift is capped (e.g. CPI or a stated percentage).
- Auto-renewal is the default, with an opt-out deadline set long before term end.
- No contractual cap on renewal pricing.
- Renewal opt-out requires certified mail or other burdensome process.
7. Describe any committed-use discount program, including minimum commitment levels, term length options, and the treatment of unused commitment at term end.
Why it matters. Committed-use economics can offer meaningful savings but introduce stranded-commit risk if usage is over-forecast. Buyers must understand forfeiture rules before sizing the commit.
- Multiple commit tiers are offered with stated discount percentages.
- Annual or monthly rollover of unused commit is available.
- Burndown reporting is available in-product.
- Unused commitment is 'use-it-or-lose-it' with no rollover.
- True-down is not permitted.
- Commit can only be increased, never decreased.
8. Describe how overages above committed or subscribed capacity are handled, including the overage rate, billing cadence, and whether soft limits or hard caps can be configured.
Why it matters. Uncapped overage exposure is a commercial risk in usage-priced AI tooling. Buyers need to know whether they can configure spend ceilings before a runaway agent or integration creates a budget event.
- Overage rates are explicitly published.
- Hard spend caps are configurable by the buyer.
- Real-time alerts at configurable thresholds are available.
- Overage rates are higher than list price (penalty pricing).
- No hard-cap option exists; only soft alerts.
- Alerts are delayed by hours or a day.
What the audit changed
A language model drafted these questions and a second model critiqued them. Three audit passes followed and made 198 changes. Three examples:
Wrong or outdated citation
Draft: No termination-for-convenience remedy on material change.
Now: No termination right on material change.
Termination for convenience means termination without cause. A right to terminate because of a price increase or a material change is an event-triggered termination right, not termination for convenience.
Wrong or outdated citation
Draft: No retroactive application of discounts (cliff structure).
Now: Tier treatment (all units repriced at the new tier, or only units above the threshold) is not stated.
Mislabeled: a 'cliff' is what retroactive all-units pricing creates (total cost can drop when a threshold is crossed). Incremental tiers are not a cliff. The real gap is an unstated tier method.
Wrong or outdated citation
Draft: Discount applies retroactively to consumption above a threshold.
Now: Tier method is stated: all units repriced at the new tier, or only units above the threshold.
Self-contradictory: a retroactive discount reprices all units, not only consumption above the threshold. The signal now matches the red flag pass 1 wrote for this question.
