Zero to co-selling with AWS, in the right order

This post starts a third track on this blog. The build track covers how we ship, the product track covers what we ship, and this one, the partner track, covers the ecosystem both of them live in: what a working AWS partnership looks like, and where our product sits inside one. Same rules as the other two tracks. One system per post, real numbers, and the exact place a human belongs.

Twice this spring I sat on a first call with a newly hired alliance lead, and both calls opened with the same question: "How fast can we get on Marketplace?" It's the natural first question. The listing is visible, the badge goes on the website, and both feel like progress you can show a board. But in every partnership we've watched succeed, from the inside of our customers' alliance teams and from running our own, Marketplace came last. The partnerships that stall are the ones that did the visible things first and the boring things never.

The order that works

A note on the clock before the order. None of these stages is a fixed number of weeks; each one is gated by what the stage before it produced. The only one with real duration in it is the middle stage, the deal data. Done by hand, that work is what makes people budget a quarter for all of this; handled by tooling, it takes days. The order is the part you can't compress.

The order itself fits in four letters: PPDM. People and paper, then data, then money. If that looks one letter away from PDM, it should. Run the sequence in that order and the Partner Development Manager is what falls out of it.

The first stage is people and paper, and mostly people. Register in Partner Central and understand which tier you're actually in (not the one the sales deck assumes). Here's the part the sales deck also skips: at the entry tiers you don't have a named Partner Development Manager yet. AWS keeps the early stages deliberately self-serve, and a PDM gets assigned when your co-sell activity gives AWS a reason to put a person on you. So this stage's quiet goal is earning that assignment, and the deal data work below is exactly what earns it. When a PDM does pick you up, treat that person like what they are: a colleague with a quota that your deals can retire. A PDM who can describe your product accurately to an AWS account team is worth more than any badge, and they get that fluency from talking to you, not from your solution brief.

The second stage is data, and this is where the boring work lives, because co-sell runs on your CRM and your CRM is probably not ready. Before an opportunity can go to AWS through ACE, someone has to know which deals have AWS underneath them, who the customer's cloud stakeholders are, and what the workload actually is. We wrote about what deal data usually looks like in the 16x problem: at one customer, 0.74% of emails were attached to the opportunity they were about. An alliance lead inherits that same emptiness. The instinct is to ask reps to fill in AWS fields by hand; that initiative dies the same death every CRM initiative dies. This is the first problem we built Aithon to answer: resolving the emails, calls, and transcripts into deals and stakeholders so the AWS-shaped facts exist without anyone typing them. Tooling or intern, get the data handled, then submit your first two or three ACE opportunities by hand to learn the machine. Ours went in 47 days after registration, and the manual walk-through taught us most of what we later automated.

The third stage is money. AWS runs more than eighty partner funding programs, and roughly four of them will matter to you at the start. Pick the one or two your deal profile hits most often, run them on real deals from the pipeline you just cleaned up, and let the first funded deal teach the sales team why the CRM hygiene was worth it. When Aithon ran this check across one customer's existing pipeline, it surfaced 39 co-sell-eligible opportunities worth roughly $3.2M in potential funding. Nothing in that pipeline was new. The checking had just never been anyone's job.

And Marketplace? It enters when a buyer asks for it, which is usually the moment their procurement team realizes the deal can draw down committed AWS spend. Listing before you have a deal that wants to transact is decorating a store on a street nobody walks down yet. When a buyer does pull it, the listing and the private offers that flow through it become a recurring workflow, one Aithon runs from inside the deal; that machinery gets its own post later in this track.

Real numbers

1. 47 days from Partner Central registration to our own first ACE opportunity submission, done by hand on purpose.
2. 0.74% email-to-opportunity association at one customer before cleanup - the state of the CRM every alliance lead inherits.
3. 39 opportunities ≈ $3.2M in potential AWS funding found in one customer's existing pipeline once checking became automatic.
4. 80+ funding programs in the wild; 1-2 are the right number to run in your first quarter.

Where the humans sit

Everything durable in an AWS partnership is a human relationship: your PDM, the account teams your deals introduce you to, the AWS seller who decides your co-sell opportunity is worth their week. Portals don't return calls. The automation in the later posts of this track exists to feed those relationships, never to replace them, and the irreversible steps (submitting to AWS, signing an offer) stay behind named humans throughout.

Steal this

PPDM, as a checklist you can run from a standing start:

Stage 1: people and paper

1. Register in Partner Central and write down your actual tier, plus exactly what the next tier requires of you.
2. Write the two-sentence description of your product an AWS seller could repeat without you in the room. You'll use it on every AWS human you meet, starting with the PDM you're working to earn.
3. Name the one AWS service your product genuinely deepens (for us: Bedrock) and collect the evidence.
4. List your five biggest open deals and, for each, whether anyone knows if the customer runs on AWS.

Stage 2: data

5. For every open enterprise deal, answer the four co-sell facts: platform, workload, who owns cloud decisions, real stage. This is the part Aithon automates; at small scale, an afternoon per account and a spreadsheet gets you started.
6. Submit two or three ACE opportunities by hand, and keep notes on every field that bounces.
7. Turn the bounce notes into CRM fixes (required fields, a validation pass) so the next submissions don't repeat them.

Stage 3: money

8. Pick the one or two funding programs your deal profile hits most often; ignore the rest for now.
9. Run eligibility across the whole pipeline, not just the big deals, and apply on one real deal.
10. When the first funded deal lands, make it loud internally; it's the argument for all the data work above.
11. Marketplace: only if a buyer has pulled it. Otherwise leave it alone until one does.

If you only have budget for one hire's worth of effort, spend it on items 5-7, the deal data. Every later motion, co-sell, funding, and Marketplace alike, reads from that same table, and it's the only investment on this list that compounds.

Next on this track: the step where most co-sell actually dies, the swivel-chair between your CRM and Partner Central, and the one-click submission we built to remove it.

---

This post is part of the partner track of How a Tiny Team Ships Like a Big One, a series on how six builders run a production AI company. Building at Aithon - if this is how you want to work, talk to us.