A rejected ACE opportunity almost never comes back with a useful reason. The status changes, sometimes a short note appears, and the partner is left to reverse-engineer what went wrong from a field that was not designed to explain anything.
So partners guess. The common guesses are that the account was already covered, that a competitor got there first, or that the partner needs a competency they do not hold. Occasionally one of those is true. Far more often the submission failed for a reason that was visible before it was ever sent, and which has nothing to do with the account being contested.
Rejection is a resource decision, not a verdict on the account
It helps to stop thinking of a PDM as an approver and start thinking of her as someone allocating a scarce, non-renewable resource: her own hours, against a book of partners far larger than she can serve properly, measured on deals that actually close.
A submission is not a request for permission. It is a request for attention. When it is rejected, the answer is usually not this is not a real account — it is I cannot tell from this whether it is real, and I do not have time to find out.
That reframing matters because it changes what you fix. If rejection were a verdict on the account, the response would be to find better accounts. If it is a failure to transfer evidence, the response is to change what travels with the submission — which is a much more tractable problem.
The five reasons submissions actually fail
1. No named stakeholder who has responded.This is the single most common cause. The submission names a company and a title but not a human being who has said anything back. From the PDM’s side this is indistinguishable from a list entry. A partner who has had a real exchange with a real person has something no filter can produce, and it should be the first thing in the submission.
2. A why-now that is about the partner, not the account. “We are targeting financial services this quarter” is a fact about your plan. “They posted four roles against a workload they are visibly migrating, and the person who owns it replied” is a fact about the account. Only the second survives scrutiny, because only the second would still be true if you had never existed.
3. A scope the PDM cannot map to a motion. If the opportunity does not clearly correspond to a workload, a program, or a recognisable co-sell motion, she has to construct that mapping herself before she can route it. Most of the time she will not.
4. Timing that is a hope. A submission with no event anchoring it — a contract date, a migration milestone, a funded initiative, a hiring pattern — reads as an account that has looked like a fit for two years and will continue to. Perpetual fit is not timing.
5. Accumulated pattern. The most damaging one, and the one partners never see, because it does not attach to any single submission. If your previous ten were thin, the eleventh is read through them. Rejection at that point is not about the eleventh opportunity at all.
Why the rejection reason you are given is rarely the real one
Disposition fields are coarse by design, and PDMs are not incentivised to write coaching notes. “Not enough information” is accurate and useless. It does not distinguish between an account that was genuinely unknowable and a submission that simply did not carry what the partner already knew.
That second case is more common than partners expect. The account knowledge often exists — in a rep’s head, in an email thread, in a call recording — and never made it into the submission because nothing in the process required it to. The knowledge was real and the transfer failed.
What AWS now tells you that it did not before
The feedback loop has genuinely improved. Partner Central now returns an opportunity quality score and its trend, the co-sell motion AWS routed the deal to, and recommendations and next best actions on shared opportunities — and at the account level, signals like solution score, in-market workload, and which programs an account qualifies for.
This is a meaningful shift: it converts a private judgment into something a partner can see. But read it carefully — it tells you how an opportunity is being assessed after you have shared it. It does not tell you which opportunities to create, and it cannot fix a submission whose underlying evidence never existed. It is a scoreboard, not a coach.
Used well, it is a fast feedback loop on your own sourcing quality: if the quality scores on your shared opportunities are trending down, the problem is upstream of ACE, in what you are choosing to submit.
Rejections are not free
There is a persistent belief that a rejected opportunity costs nothing — you submitted, it did not work, you move on. That is true mechanically and false relationally.
Every submission is a withdrawal from or a deposit into a credibility account. A partner who submits ten thin opportunities has not run ten independent experiments; they have taught one person that their name in the queue predicts low-value work. That lesson persists, and it makes the next genuinely good opportunity harder to get championed than it should be.
Which is why “submit more and see what sticks” is not a neutral strategy with an unremarkable expected value. It has a negative one.
The pre-submission test
Before anything goes into ACE, it should survive four questions. They are the PDM’s questions, asked in advance.
Who at this account has actually responded to us, and what did they say? What specifically is happening at this account that makes this the moment? Which workload or program does this map to? And what would the PDM have to take on faith if she championed this — is that amount of faith reasonable to ask?
If three of the four have crisp answers and one does not, the submission is not ready. That is not a delay; it is the work. And it is far cheaper to find out at your desk than in a disposition field six weeks later.
Where the fix actually lives
Notice that every one of those four questions is answered — or not — long before the submission form is opened. They are determined by which account was chosen, how the conversation was opened, and whether anybody wrote down why this account and why now before outreach went out.
A qualification record produced before the first message, and carried forward with the opportunity, answers all four by construction. A qualification reconstructed afterwards from a call note answers none of them reliably, because it is memory rather than evidence.
Which means the rejection problem and the empty-pipeline problem are the same problem seen from two ends. Both are decided upstream, in the part of the motion that happens before any co-sell tool is involved.