AWS used re:Invent last week to reshape how products get found, priced and bought in AWS Marketplace. Five changes, covering discovery, pricing, bundling, services and configuration.
The one that changes day-to-day work for a seller is express private offers. Here is what it is and when it is the wrong tool.
Why manual offers stop scaling
Every private offer today is assembled by a person: pull the customer, pick the discount, set the term, attach the EULA, route for approval, send. Fine at ten offers a quarter. Not fine at two hundred.
Four costs compound:
- Time to offer is deal cycle time. An offer that takes two days to produce is two days of a deal not moving.
- Approval chains become the bottleneck, usually one person who is also doing three other jobs.
- Offers expire unnoticed, as covered in private offers explained. The common loss is not rejection, it is lapse.
- Renewals grow linearly with customers. Every customer you add is a recurring manual task forever.
That last one is the structural problem. Manual offer creation means your offer workload grows at the same rate as your customer base, which means it eventually consumes a team.
Express private offers
The new capability is opt-in: you define the terms, AWS qualifies the customer and delivers personalised pricing automatically, in minutes rather than days.
What you give up is case-by-case judgement. You are pre-committing to a pricing policy and letting it execute without you. That is a real trade, and it is the right trade for a specific slice of your pipeline:
Suits rules: mid-market deals inside a known discount band, standard terms, no unusual legal requirements, renewals at existing pricing, expansion on an existing agreement.
Still needs a human: anything with negotiated legal terms, multi-year commitments with custom payment schedules, deals where the discount depends on strategic value rather than size, anything involving a reseller, first deals with a new large account.
The useful mental model is that you are not replacing your offer process, you are carving out the predictable portion so your people only touch the deals that genuinely need judgement.
Bundles
The second change worth noting: selling multiple components together as one purchase. If your product has modules, or if you habitually sell software plus implementation, this collapses several offers into one transaction for the customer.
Check how each component meters inside a bundle before you commit to it. Bundling is a packaging change with metering consequences, and the metering is where revenue leaks.
Discovery is now part of the pricing conversation
AWS also introduced conversational, AI-driven discovery — customers describe what they need in natural language, including use case and constraints, rather than searching keywords and browsing a catalogue.
This is the quiet one with the longest tail. Listing content written to rank for search terms is now being read by something matching a described requirement. Specificity starts to beat keyword density: a listing that plainly states what the product does, for whom, and what it integrates with is easier to match than one optimised for search.
I flagged this direction when the agent categories opened in July. It has now arrived in the main buying flow, which makes reviewing your listing copy a concrete task rather than a future one.
Doing it yourself through APIs
Express offers are AWS’s rules engine. If your pricing logic does not fit it, the Catalog API remains the route to building your own, including automating channel resale authorization rather than issuing each one by hand.
One principle if you build this: automate assembly and delivery, keep approval human for anything outside a defined band. Fully automated discounting without a ceiling is how you discover a pricing policy error at scale.
What to do with this
- Segment your offer volume. What fraction is standard-band, standard-terms? That is your automation candidate.
- Write down the policy — discount bands, term options, approval thresholds. You cannot automate a policy that only exists in someone’s head.
- Start with renewals. They are the most repetitive, the most predictable, and the most costly to lose to timing.
- Review your listing copy against the discovery change.
CloudX’s marketplace platform covers the offer layer with approvals and renewals as tracked records, and managed operations covers running it across clouds if you would rather not staff it.
The summary
Express private offers trade judgement for speed, which is correct for the predictable majority of deals and wrong for the ones that need a negotiation. Write down your pricing policy first — automation turns a documented policy into throughput and an undocumented one into a liability.