Cloud Marketplaces

What do private offers actually do?#

A private offer is a negotiated, customer-specific version of a listing: custom price, custom term, custom EULA, custom payment schedule. It is the mechanism that makes a marketplace usable for enterprise deals rather than credit-card self-service.

Two mechanics catch founders out. First, entitlement and metering are your engineering problem, not the provider's — the marketplace bills what you report, and reconciliation errors become invoice disputes. Second, contract-duration limits are structural: on Microsoft, a private plan caps at three years while a private offer built on absolute pricing can run far longer. If your deal desk assumes a five-year term is always available, it will not be.

How do you decide? Route-to-transaction economics#

Compare the marketplace route with the best realistic direct route for the same account.

channel-adjusted contribution = contract value × gross margin − listing fee − reseller or referral share − incremental sales and operations cost − expected channel-conflict cost

Then add probability and time:

expected deal value = probability of close × channel-adjusted contribution ÷ (1 + r)^(days to cash ÷ 365)

The discount factor is negligible over a single quarter at ordinary rates. It is not negligible when you are managing runway and a 60-day acceleration changes what you can hire. Use it as a reminder that cash timing has value, not as a precision instrument.

Measure incremental lift, not gross lift#

Never credit the marketplace with a deal that would have closed directly on the same terms and timeline.

incremental marketplace value = expected value through marketplace − expected value through direct route − fixed marketplace investment

Fixed investment includes listing review, metering and entitlement engineering, testing, legal templates, partner management, finance reconciliation, and ongoing release maintenance. It is mostly a one-time cost with a recurring maintenance tail — which is exactly why the payback question is "how many incremental deals," not "what is the fee percentage."

The marketplace readiness test#

Worked example: is a marketplace listing worth it?#

A fictional infrastructure startup, Meterline, is evaluating an AWS listing. A representative deal is a $120,000 annual SaaS contract. Product gross margin before channel costs is 82%.

gross contribution before selling cost = $120,000 × 82% = $98,400

Route 1 — direct#

Expected cycle: 150 days. Account-specific commission, legal, billing setup, and sales engineering cost $24,000.

direct contribution = $98,400 − $24,000 = $74,400

Route 2 — marketplace via a channel partner#

The customer will close through a channel partner private offer in 90 days. Using the current AWS schedule for a sub-$1M SaaS private offer sold through a partner:

marketplace contribution = $98,400 − $4,200 − $12,000 − $4,000 = $78,200

Contribution is $3,800 higher than the direct route, and cash arrives roughly 60 days earlier — but only because the assumed direct route carries $24,000 of account-specific selling cost that the partner-led route avoids. Change that assumption and the ranking flips.

Sensitivity, stated plainly. If Meterline listed without a channel partner, the fee would be 3% ($3,600) and there would be no $12,000 partner share — but also no partner-sourced deal and no reduction in direct selling cost. The partner share is only worth paying if the partner actually sources or accelerates.

Now test the real question: incremental lift#

Suppose marketplace procurement raises close probability from 55% to 75%:

direct expected contribution = 55% × $74,400 = $40,920

marketplace expected contribution = 75% × $78,200 = $58,650

expected-value gain per deal = $58,650 − $40,920 = $17,730

If listing, metering, entitlement, and legal work cost $90,000:

$90,000 ÷ $17,730 = 5.08 deals

Meterline needs roughly five similar incremental deals to recover the fixed investment — and "incremental" is doing all the work. If three of those five would have closed directly anyway, the listing has not paid for itself.

Three caveats that keep this honest. The 55% → 75% close-rate lift is an assumption, not a measurement; it should come from a matched comparison of accounts, not from the two deals that went well. The 82% gross margin quietly assumes support and metering labor stay inside it, which is the first thing that breaks at scale. And the model prices a transaction, not an outcome — see the caution below on shelfware.

What are the common mistakes?#

• Treating a listing as demand generation. Products sit undiscovered without a sales or partner motion behind them. A listing is a checkout counter, not a storefront window.
• Using one fee percentage for every deal. Fees vary by provider, offer type, deployment method, contract value, renewal status, channel involvement, and region. The AWS gap between a 3% SaaS public offer and a 20% server public offer is not a rounding error.
• Quoting a committed-spend cap you cannot source. Drawdown rules differ by provider and are partly contractual. Google publishes a 25% cap for channel private offers; AWS publishes nothing. Verify per account, in writing.
• Adding the channel after pricing is final. Listing fee plus partner share can turn an already-approved discount into a loss-making deal. Model the channel before the discount desk approves anything.
• Assuming procurement review disappears. Marketplace contracting helps with vendor onboarding and payment. Security, privacy, data-residency, and architecture review remain.

When does the marketplace strategy break?#

• Self-serve, card-paid motions. If customers buy with a credit card in minutes, marketplace overhead adds cost and friction with nothing to compress. See Product-Led Growth.
• Deals too small to absorb channel overhead. Fixed listing investment plus per-deal fees needs contract values that can carry them.
• Pricing too bespoke to represent. If every deal has a custom meter or a hand-built commercial structure, the platform will misrepresent it and reconciliation will fail.
• Immature security and support posture. Enterprise buyers arriving through a marketplace still expect enterprise controls. Arriving faster at a review you cannot pass is not progress.
• The provider is both channel and competitor. Reach comes with platform dependency, data exposure, fee risk, and bargaining asymmetry. Preserve direct customer insight, portable entitlements, and a credible alternative route. See Switching Costs and platform strategy.
• Procurement speed is not adoption. A purchase funded by committed budget the customer had to spend anyway can produce shelfware. Measure deployment, usage, renewal, and expansion after the transaction — not gross transaction value. See Customer Success and Land-and-Expand.