Because your crm wasn't built for partners.

Co-Selling 101: What It Is, Why It Broke, and How to Fix It in 2026

Co-selling is the fastest-growing motion in B2B software and the worst-instrumented. What it is, why AWS ACE and Azure Partner Center are pushing it, why spreadsheet reconciliation is killing it, and what a fixed operating model looks like in 2026.

Co-selling is the fastest-growing motion in B2B software, and the worst-instrumented. Every ISV building on AWS, Azure, or Google Cloud is being pushed into it. Every services partner with a hyperscaler practice is being measured on it. And almost everyone doing it is running the motion out of a spreadsheet that gets rebuilt the Sunday night before every QBR.

This post is the anchor for our co-selling hub. If you’re new to the term, start here. If you’ve been doing this for three years and are tired of it, skip to “Why it broke.”

What co-selling actually is

Co-selling is when two companies work the same deal, at the same time, for the same customer, and split the credit.

That definition sounds obvious until you compare it to the motions it gets confused with:

  • Channel / reseller. The partner owns the transaction. They buy from you at a discount and sell to the customer. You may never talk to the end buyer. Co-selling is the opposite: both parties are in the room, and the customer usually pays each vendor directly (or pays one through a marketplace with a payout to the other).
  • Referral. A partner hands you a lead and walks away. Maybe they get a spiff. There’s no shared account plan, no joint forecast, no co-owned close plan. Co-selling is a two-vendor sales cycle with two account teams pursuing one outcome.
  • OEM / embed. Your product ships inside theirs. There’s no joint selling motion at all; there’s a contract and a royalty.
  • Alliances / marketing partnerships. Logos on a slide, a joint webinar, a press release. No deal-level collaboration.

Co-selling is deal-level. That’s the whole distinction. Two sellers, one opportunity, shared visibility, shared credit, shared close plan.

Why co-selling is spreading

Three forces, all pointing the same direction.

1. The hyperscaler programs matured. AWS ACE (APN Customer Engagements), Microsoft’s Partner Center co-sell workflow, and Google Cloud’s Partner Sales Console all now let partners register opportunities, request field help, and get co-sell credit that shows up in the account team’s compensation. That last part is the unlock. When an AWS AE gets quota relief for closing an ISV deal, they suddenly answer their phone.

2. Marketplaces became the transaction layer. Buyers are burning down committed cloud spend through AWS Marketplace, Azure Marketplace, and GCP Marketplace at record rates. A private offer through marketplace can pay the ISV, pay the SI who influenced the deal, and retire the customer’s cloud commit in the same transaction. That structure requires everyone to agree on who did what, before the paperwork gets cut.

3. Buyers stopped tolerating uncoordinated vendors. A modern enterprise IT deal has, on average, a cloud provider, an ISV, and one or two services partners. If those vendors show up with three different discovery calls and three different implementation timelines, procurement notices. Coordinated selling is now a competitive requirement, not a nice-to-have.

The result: partner-sourced and partner-influenced pipeline is the fastest-growing revenue category at most cloud-native ISVs. IDC pegs the ecosystem opportunity in the trillions. Every board is asking about it.

Why co-selling is broken today

The programs work. The motion doesn’t. Here’s what’s actually happening inside partner orgs in 2026.

The spreadsheet reconciliation ritual. Before every QBR with AWS, Azure, or a major SI, the partner manager exports their side’s pipeline from Salesforce, gets a CSV from the partner’s PDM, and spends a day (sometimes three) matching them by hand. Account names don’t match. Amounts are stale. Half the opps on the partner’s list were closed-lost two quarters ago. The other half are missing from the ISV’s CRM entirely. The “reconciled” sheet is out of date the moment the meeting ends.

Deals falling through the cracks. An ISV AE closes a $400K ARR deal. Nobody remembers to update ACE. The AWS PDM never sees it. It doesn’t count toward the partner’s Marketplace revenue targets, doesn’t retire the AWS AE’s quota, and doesn’t show up in the co-sell attach rate the ISV is being measured on. Multiply by dozens of deals per quarter. That’s not an edge case; that’s the average.

No shared pipeline visibility. The ISV’s RevOps team has a pipeline view. The partner’s PDM has a different pipeline view. Nobody has the merged view. When a customer calls the AWS account team asking about the ISV’s implementation, the AE has to Slack the ISV’s partner manager to find out where the deal actually stands. If she’s out that day, the customer waits.

Status drift. The ISV moves an opp from Stage 3 to Stage 5. The partner still thinks it’s in discovery. Two weeks later, the partner’s marketing team sends the customer a “considering us?” nurture email. The customer, mid-negotiation on the SOW, is confused. The AE is furious.

Credit disputes. Who sourced it? Who influenced it? Was the SI in the deal or just cc’d on a thread? Without a shared record with timestamps, this becomes a political argument every quarter. The bigger partners win those arguments. The smaller ones stop bothering.

Data hostage-taking. Neither side wants to expose their full pipeline to the other, so they share the minimum — usually via a static export. Which brings us back to the spreadsheet.

The common thread: there is no shared source of truth. Each side is managing its own CRM record for the same deal, and the two records drift the moment they’re created.

What “good” looks like

The fix isn’t a better spreadsheet template. It isn’t a bidirectional Salesforce-to-Salesforce sync either — those projects take 18 months, cost seven figures, and break the first time a picklist changes on either side.

The fix is a shared workspace that sits between the two CRMs and holds the co-sell record as a first-class object. Three properties matter:

1. One deal record, two owners. A single opportunity object that both the ISV AE and the partner PDM can see and update. Not a copy synced from Salesforce. The record itself. Amount, stage, close date, next step, decision maker — one field, two editors, full audit trail of who changed what. When the ISV moves the deal to Stage 5, the partner sees Stage 5 immediately. No export, no reconciliation.

2. Automatic status sync to each side’s system of record. The shared record pushes updates back into each party’s CRM (and into ACE / Partner Center / Google’s console) so the finance, RevOps, and comp systems on both sides stay accurate without human data entry. The shared workspace is the source of truth for the joint motion; each side’s CRM remains the source of truth for their own forecasting.

3. Scoped visibility by default. Neither side dumps their full pipeline. The shared workspace only contains the deals both parties have opted into. That’s what makes it politically viable. RevOps at the ISV isn’t nervous about a partner seeing net-new logos they weren’t invited to. The partner isn’t nervous about the ISV seeing their other ISV relationships.

Add the operational surface both teams need on top:

  • Shared close plans and next-step tracking, so a deal isn’t stuck because “I thought you were sending the SOW.”
  • Joint account planning at the whitespace level, not just the deal level.
  • QBR dashboards that generate themselves from the live record, so the Sunday-night spreadsheet dies.
  • Attribution rules encoded once, so credit isn’t relitigated every quarter.

The 2026 mandate

The hyperscaler programs are only going to get more demanding. Marketplace co-sell is becoming the default motion for how enterprise software gets bought. If your co-sell operating model is still “export, reconcile, email, repeat,” you’re not going to lose to a competitor with a better product. You’re going to lose to a competitor whose partner team can actually answer the question “where does this deal stand?” in real time.

A partner relationship management layer built for the co-sell era — shared workspace, one deal record, automatic sync — is the infrastructure that makes the motion work at scale.

That’s what Recordbook is built to do. If your Q1 kickoff includes the words “we need to fix the way we co-sell with AWS,” start with the shared record. Everything else follows from there.


Next in this series: The Co-Sell Operating Model — roles, cadences, and metrics that separate the top-decile partner orgs from everyone else.