Closing the Seller Activation Gap in Co-Sell Partnerships

Ask a partner leader how their program is doing and you’ll hear about activity: alliances signed, pipeline influenced, certifications issued, QBRs held on schedule. All real, all worth doing. But none of it answers the main question a CRO cares about: “Is the partnership driving revenue that would not have closed without it?”

Between all that activity and that revenue sits a single moment: a seller, mid-deal, deciding whether to reach for the partnership or go it alone. Repeated across thousands of live deals, that decision is where a signed alliance either becomes revenue the business counts or quietly evaporates.

Call it seller activation: the one conversion point that turns everything a partner team builds into the only number a CRO rewards, and why it belongs at the center of the partner leader’s job.

Three Ways Co-Sell Revenue Leaks Before It’s Counted

When sellers engage partnerships, the upside compounds. When they do not, revenue leaks in three places:

• Co-sell that never happens. The referral never submitted, the partner never pulled in, the stronger position in the deal never claimed.
• Marketplace spend left unburned. Deals closed direct while the customer had committed cloud spend sitting there to draw down.
• Incentives earned but never claimed. Money the partner already earned, left uncollected because no one tracked the deadline.

All three of these ways to lose revenue are rooted in the gap between the partnership and the moment a seller decides whether to use it. None of this is a failure of effort or goodwill; it is an operating problem that can be solved.

Operating Discipline Beats Relationships, Tenure, and Scale

Not through relationships, tenure, nor scale. The programs that pull ahead do it on operating discipline: a few focused plays, run where the deal already lives, measured at the level of the individual seller, and reinforced each time a commercial milestone lands.

This co-sell field guide examines the quantitative shape of what sellers actually did (six years of real co-sell behavior, not surveys or self-reported health) and the qualitative how of the operators who have already activated sellers by hand. The numbers show the pattern; the practitioners show the play.

The Activation Gap

The activation gap is the distance between a partnership and the moment a seller decides, mid-deal, whether to use it. When the gap is wide, the seller goes it alone and the upside goes unrealized. When it is closed, the partnership multiplies. Six years of seller-level co-sell data show this distance tracks outcomes more reliably than relationship strength or team size ever has.

Salesforce research reveals 94% of sellers work with partners on their deals, yet 89% doubt they will hit quota. Partners are everywhere in the modern deal, and sellers are still drowning.

It is not that sellers do not care. To the average seller, bringing the partnership into a live deal looks like more trouble than it is worth. Reaching for the partner quietly asks three things of a seller with no time to spare: know which partner fits this deal, have it in front of them where they already work, and act on it before the steps pile up. Miss any one of these and going it alone wins.

6.3 Partners, One Seller, No Time to Choose

No seller can hold every partner and every use case in their head. The decision to go it alone is fast, private, and unsentimental.

“I run a go-to-market team. I myself have been a seller. Sellers are coin operated, and so they will look for the easiest path to close a deal.”

And the modern deal is crowded. Jay McBain, Chief Analyst for Channels, Partnerships and Ecosystems at Omdia, identifies an average of 6.3 partners surrounding an enterprise deal, and north of ten in large enterprises. A seller staring at that thicket, mid-deal, will not run a research project.

“You don’t have time to search through the enablement material and come up with the answer among the dozens of partnerships available to you. So you go with the one or two partners you already know and are comfortable with, regardless of if they are good or not good for that opportunity. To avoid all of this complexity, it is understandable to go it alone, because the friction isn’t worth the potential upside.”

The impossible numbers game is why more enablement does not fix it, and why the program cannot staff its way out either. Greg Sarafin, former global managing chair at EY, laid out the math: take your number of partners, add your number of sales motions, multiply by every seller who might touch a deal, yours and your partners’. At a firm his old size, those combinations run into the tens of millions.

“You cannot enable every sales motion, across every partner, across every one of your sellers, plus every one of your partner sellers, at the moment of opportunity.”

The trap tightens the better you do: more successful partnering means more partners and a more challenging job. Enablement scales linearly; complexity scales like a matrix. And the seller usually knew the partnership existed anyway. Knowledge was never the issue. The bottleneck is matching the right partner to this deal, fast.

Known Partner, Still Skipped: the Friction Is Mechanical

The second problem is mechanical. There are many small steps between deciding to use a partnership and actually using it: find the right contact on the partner side, create the referral, map the deal into a portal shaped nothing like your CRM, schedule the introduction, chase a status update later. It is not hard on its own, but together they are real delays, and they all land on the seller at the exact moment they are least willing to spend time on anything that is not closing.

That accumulated friction is the activation gap in practice: the distance between the partnership and the seller’s live decision, measured in steps nobody wants to take. But friction, unlike indifference, can be removed.

Recap & FAQ

Why don't sellers bring partnerships into their deals?

Bringing the partnership in is usually more work than closing without it, so it loses a fast, private cost-benefit calculation — no matter how strong the alliance looks from outside. It is not indifference and not a knowledge gap.

What is the activation gap?

The distance between a partnership and the moment a seller decides, mid-deal, whether to bring it in. It is made of two things: the impossible numbers game of remembering the right partner, and the manual steps of actually pulling them in.

Doesn't better enablement close it?

Rarely. Knowledge was seldom the bottleneck, and the partner-motion-by-seller combinations are too large to enable by hand. The bottleneck is mechanical — the steps between deciding to co-sell and actually co-selling. Removing steps moves activation; more decks do not.

Project Ground Truth: Six Years of Co-Sell Data, 92 Ecosystems

We stopped collecting opinions about what makes partnerships work and looked at what sellers actually did. Project Ground Truth, co-sponsored by AWS, covers six years of co-sell across 92 partner ecosystems and 5,709 named sellers.

Sort a partner’s sellers by how many deals each referred in a year, and depth is the share who kept coming back, three referrals or more. Breadth is how many show up at all. Depth is what separates a habit from a list.

Note: the study measures co-sell behavior, not revenue. It has no win-rate or deal-value field, so where this guide cites deal economics the numbers come from outside benchmarks, labeled as such.

Depth Is the Difference; Scale Is Not

Referral volume and seller count rise and fall together, and both follow from how a program is run; neither drives anything. Depth behaves differently. It barely tracks size, so large programs do not get it for free. A tightly run 24-seller partner routinely out-engages one with 400. Among the partners we can read cleanly, depth runs around 70 percent; across the fuller, noisier cohort it sits in the high 50s.

Depth is the closest read we get on operating discipline, though it is a proxy and not the thing itself. The data sees the behavior, not the choice behind it, which is why one of the findings below is about who can even see their own numbers. Read cleanly, depth is the one figure on this map that reflects a decision, and the one worth managing to.

Partners also fall into a few recognizable patterns, and each is best read against its own kind rather than one leaderboard.

Boomi: From a Quarter to Nearly Three-Quarters Depth in Four Years

Boomi is the cleanest picture of building it, the one partner in the study that grew breadth and depth at the same time. Over four years Boomi grew from about 130 active sellers to nearly 300, while its depth — the share of sellers reaching three-plus referrals — climbed from roughly a quarter to nearly three-quarters. Most programs watch depth thin as they add sellers. Boomi’s thickened.

“I have learned painfully through the years that it is difficult to scale an organization through sales and marketing alone. The real advantage, the leverage that you build, is through a massive partner ecosystem.”

Operating rigor drove the curve, but the inflections lined up with commercial events that let sellers close differently: a renewed Strategic Collaboration Agreement, a run of new competencies, a Marketplace listing. Boomi did not ride any single one. It stacked them, and a lean team stayed disciplined enough to convert each before the next arrived. Certified partners now drive about 60 percent of the company’s total ACV. Joe Estes, who runs the global AWS alliance, named the constraint when he accepted WorkSpan’s 2026 co-sell recognition:

“We have a lean team that operates highly efficiently and is trying to achieve scale well above our weight class.”

Recap & FAQ

What actually predicts whether a partnership activates?

Operating discipline. Its best visible proxy is depth — how many individual sellers come back, not breadth, tenure, team size, or inbound share — though depth only means what it says once the account-owner data is clean. Depth moves nearly independently of scale, which is why small, focused partners routinely out-engage far larger ones.

Does the study say depth drives revenue?

No — and the honesty is deliberate. The dataset has no revenue variable; it measures co-sell behavior. Depth is the signal of the discipline that independent benchmarks tie to revenue. We do not join the two datasets, and we do not pretend to.

Does pushing more outbound earn more inbound over time?

Not meaningfully. Among 45 tenured partners, 58% saw inbound fall, ~7% saw it rise, the rest flat. Inbound is set by structural partner type, not effort.

Why can't most programs see their own depth?

Their dashboards count deals, not sellers, and the account-owner field is usually too sparse to read — sometimes because a few operations people file for a larger engaged team. Cleaning up that one field is what turns the map into a live feed.

The Easy Path

Closing the activation gap means making the partnership the easy path to the close: already in the deal, already knowing the account, ready the moment the seller decides, costing them nothing to use. Project Ground Truth revealed the programs that pull ahead operate close to the live decision, with the discipline to be there every time. The easy path is that same idea pointed at your own operation.

From Relationship Manager to Revenue Operator

Someone builds that proximity. Twenty years ago the partner hire was chosen for relationships. Dan McAllister, who leads alliances and channels at Boomi, remembers what the job used to select for.

“They were looking for people who played golf. I don’t play golf. Can I outperform somebody who does from a business revenue and business metric standpoint?”

In recent years, the partnering role has converted from a relationship discipline into an operating one. The relationships still matter, but the job is now run with the rigor a sales team brings to a pipeline.

Rob Moyer, who ran partner strategy at Gong and co-wrote The Partnership Operator’s Manual for the AI Era, watched this happen once before, when CRM became sales’ system of record. Partnerships are living through it now. The relationship manager is becoming an operator who runs co-sell as a revenue motion, not a book of goodwill. His core move is to put that motion where the deal lives, and his warning is that you have to earn the right to automate it first.

“Parallel universes do not get budget. Systems of record do. And you cannot automate chaos.”

A partner program that runs in its own portal, its own tools, its own quarterly deck is invisible at the moment a seller decides, because the seller is living in the CRM. But point technology at a messy program and it just gets faster at being messy. The human work has to come first, and it is the operating discipline the data already rewarded.

The strongest operators, unprompted, keep describing the same habits: they start from the outcome and run fewer plays, not more; they own the hardest seam in the business, the part Munish Khetrapal of Palo Alto Networks describes as where “you neither own a product, nor a customer, nor a technology”; and they measure the motion in numbers a CRO recognizes.

Measuring it is where the discipline turns concrete. Rather than argue over who influenced a deal, Khetrapal’s team built a scoring method:

“We identified 12 fields. If you have three to four of those fields, which means you have jointly built a solution, jointly architected it, you start immediately creating irrefutable evidence. And that’s really the big holy grail for us.”

Fill in enough of those fields and a QBR stops being a fight over credit and becomes a look at a shared number. That is a discipline.

Qualtrics: Over 18x More Marketplace Transactions in a Year

Programs describe two problems: activating your own sellers, and activating your partner’s sellers. Those are the same problem wearing two faces. The partnership has to be the easy path at the moment of decision, or it gets skipped; only the identity of the decider changes. Brian Bohan, who built AWS’s Accenture partnership from nothing to a billion dollars, describes it done right inside the largest co-sell operation in the industry:

“When our sellers are sitting down to do their account planning, we are feeding them highly contextualized partner and solution recommendations.”

Qualtrics engineered that by design. Its buyer usually sits with the CMO or CHRO, not IT, while its cloud partner’s account team already has the technical-buyer relationship. That gap, two champions with no bridge between them, is where most co-sell motions quietly die. A two-person team built the bridge they call a landing zone.

“We create what we call landing zones… ‘What we really need you to do is go work with your champion and explain to them why it’s super valuable that this transacts through marketplace.’ There’s been multiple examples where the procurement team proactively reaches out to us. They’ve just got rid of maybe the biggest hurdle, procurement.”

A landing zone puts the partnership in position before the deciding moment arrives, instead of scrambling a response after. In a year it moved Qualtrics’s marketplace transactions up 1,775% and shared referrals up 700%, enough to win AWS’s Healthcare Partner of the Year.

One Operator, Thousands of Moments

Every discipline above has a ceiling made of human bandwidth. Greg Petrossian runs DataStax’s Microsoft partnership.

“We have roughly 100 sellers. Each seller has around 20 accounts. Each account could have one to three opportunities. And so that is a lot of activity happening for a single person like myself to support.”

That is a couple thousand individual moments where a seller might or might not think to bring in the partnership, and it is one person’s job to be present for all of them. A great operator removes friction deal by deal and still cannot stand next to every seller at the moment each one decides.

“Making that dependency of ‘Greg must be in every sales call’ is not scalable. AI should be able to very quickly sniff it out and tell the seller: ‘Here’s an opportunity where you can get some lift with the partnership.’”

His standard for whatever replaces that dependency is the cleanest statement of the easy path in the guide: it has to be “quick, easy, and nearly invisible.”

Look closely at what these operators actually do by hand. It is three jobs, not one. They know which partner fits this deal. They carry the better-together story into it. And they know the next move that advances it. Eventually, the partnership gets big enough where a person doing all three by hand, in every deal, stops being possible. Most programs hit this constraint sooner than they think. This is what technology has to solve.

Recap & FAQ

What does "the easy path" mean?

The partnership shows up in the deal already knowing the deal, at the moment the seller decides, without asking them to do anything first. It is the single standard this guide holds every program and tool against.

Is activating your own sellers a different problem than your partner's?

No. One problem, two faces. The partnership has to clear the same bar at the moment of decision; only the chair changes.

Why can't a great partner manager just be the easy path themselves?

The math does not allow it. Petrossian's own motion is roughly 100 sellers, 20 accounts each, one to three opportunities per account: thousands of decision moments no single person can attend.

Your Best Operator, in Every Deal

A human operator can be present at a few dozen decisions a week. The easy path at the scale of thousands of live deals needs something that can be in all of them. WorkSpan AI is built to do what your best partner leader does, in every deal, without the human limit. It is the operating discipline the data rewarded, run at machine scale, and built for the specific partnership rather than as the general-purpose AI on every desktop.

Three Moments: Think, Present, Act

Seller activation runs on three capabilities that work as one loop. They are the three jobs your best operators do by hand, in that same order.

WorkSpan AI agents think ahead. One agent per partnership, trained on that relationship rather than the internet: its playbooks, its rules of engagement, and the live transactional history of every deal the two companies have run together. That is what lets it do the work a seller has no time for: map a rep’s territory against where the partner is strong, and surface the lookalike accounts where the same play has already closed. One agent per partnership means coverage has no headcount ceiling, and this layer is coming online now, partnership by partnership. It answers the impossible numbers game directly, and it is what Sarafin called years ago the “cockpit around the seller.”

The Partner Advantage Card presents. It lives in the right rail of the opportunity in Salesforce or Dynamics, where the seller already works: which partners are relevant, how they map, propensity to buy, which incentives to request, and a better-together story on demand. The partner shows up in the deal the seller is working, not in a portal they have to remember to open. It passes Petrossian’s test, “quick, easy, and nearly invisible,” and is live today, already carrying the partner into live deals at customers like Boomi.

Next-best-action turns knowing into the move. Here is where the gap actually closes: a seller mid-deal, deciding whether to reach for the partner or go it alone. The agent has already read the account across both companies’ systems of record. It does not just show the seller that a partner is relevant; it surfaces the single next move most likely to advance this specific deal, and makes that move executable in a click, mapped natively into the partner’s system so there is no re-keying and no second window. That one move takes three forms, one for each of the revenue leaks the guide opened with — co-sell, marketplace, and incentives:

• Co-sell. Submit the referral from inside the deal. AWS co-sell benchmarks put the win rate on deals with a submitted referral at roughly 47%, versus about 22% without one, yet the manual version classically takes about eight steps across two systems and three days, which is why so many referrals never get submitted. Status flows back to the Card in near real time, so “referral sent, pending acceptance” is something the seller sees rather than something the alliance team chases.
• Marketplace. Committed cloud spend now sits in enterprise budgets in the hundreds of billions, and every deal closed direct that could have transacted through marketplace is upside surrendered, usually because the seller does not know, at the moment they are structuring the deal, that this account has unspent commit. The Partner Advantage Card now surfaces that signal inline, and the move carries the offer. Boomi grew its AWS Marketplace revenue by more than 3,000% year over year with a small team, and the study saw marketplace listings act as depth inflection points.
• Incentives. The quietest of the three, and often the most valuable to recover: money the partner already earned, expiring unclaimed across programs that each have their own portal and deadline. The capability tracks eligibility and deadlines and reconciles what was earned against what was paid, keeping a human at the submission step by design. That infrastructure is battle-tested and live in top-tier customers today.

WorkSpan built the first version of next-best-action with a global systems integrator and a hyperscaler, and it reaches the market this fall. It is the difference between a card that informs and an operator that acts.

Underneath these three moments is a fourth feature that makes the loop compound: Engagement & Attribution, which ties partner-engaged actions to outcomes with auditable lineage across both companies’ systems of record, then feeds what it learns back to the Agents.

Attribution is what turns a QBR from a credit negotiation into a conversation about a shared number, and it finally gives a program the clean seller-level data the study found most partners lack. Add the recovered upside across all three leaks and the case for closing the gap stops being a co-sell argument and becomes a revenue argument, now one you can actually measure, deal by deal.

The Three Bars General AI Can’t Clear

If the payoff is that clear, why hasn’t a general-use AI assistant already delivered it? Because the work is not what those tools are built for. They are single-company productivity tools that draft, summarize, and answer inside your own walls. A co-sell motion asks for what a single-company tool structurally cannot provide. It has to execute, not just generate: complete the referral, not draft an email about it. It has to see the live deal across a CRM record, a cloud portal, and the partner’s system at once, not a static document. And it has to operate across the trust boundary between two companies, which is the deepest bar of all.

Picture the version everyone recognizes. A global systems integrator and a large software vendor want to plan joint accounts. Both have rich CRM data, both have invested heavily in AI, and the quarterly planning still happens on a PowerPoint built from two exported spreadsheets, because neither company can let the other’s AI touch its data. That is the trust boundary asserting itself, exactly as it should. It is not a tooling gap, and it does not close as the models improve.

The resolution is not to ask one side to trust the other more, but to put a trusted party in the middle: the other half of Sarafin’s argument.

“There has to be a trusted service manager in the middle between any two partners. Somebody that can protect you from each other, can protect your data from within the organization, can combine your data safely with the counterparty’s data and a corpus of data that neither party A nor party B have.”

That is a description of an architecture: a cross-company platform, not a PRM, not a shared Salesforce instance, not an AI tool that reads your documents. It is a trusted space between two companies, with permissions enforced from both sides at the individual deal level. Account mapping happens inside that space. The overlap, the stage, and the next action get shared; each company’s confidential fields stay governed by that company’s rules and never leave its control. The agent is no longer reaching for someone else’s data. It is operating inside a space both companies authorized.

The scale evidence that this is real and not a whiteboard concept: WorkSpan has processed $513 billion in shared pipeline across its cross-company environment, with $196 billion already closed, none of it sitting inside any single organization’s AI stack. WorkSpan was also the first vendor to build against Google Cloud’s co-sell API and today runs co-sell for many of its largest partners. A single-company tool cannot follow into a space defined by belonging to both companies at once. That is an architecture limit, not a feature gap.

WorkSpan Report

Running AI-Native Partnerships

The full three-bars framework

A 2026 report on what AI is doing to the partnership function — beyond time savings, into category transformation. Featuring ten partner-revenue leaders across cloud, ISV, and SI ecosystems.

Read the report ›

Recap & FAQ

How do I do account mapping with a partner without sharing confidential data?

In a sanctioned space that sits between the two companies rather than inside either one, with field-level entitlements enforced from both sides at the deal level. Overlap, stage, and next action get shared; each company's sensitive fields stay governed by its own rules. A single-company AI tool cannot do this; by definition it only sees one side.

Why can't we just use general-use AI?

Those are single-company productivity tools; co-sell is a cross-company execution problem. They clear none of the three bars: they generate rather than execute, see a document rather than the live deal across three systems, and cannot operate across the trust boundary. That is a limit of how the tools are built, not a feature they are missing.

Does this replace the partner manager?

No. It removes the constraint that a partner manager can only be in one place at a time. The deal understanding gets encoded into a per-partnership Agent and carried into every opportunity, freeing the human for judgment, exceptions, and strategy.

Leadership Before Technology

The mechanism is real, and the parts already in the market work. But no mechanism installs itself, and the leaders who have lived through this kind of change are candid that the hardest part is not the software.

Understand Your Atlas in Four Weeks

The first step is not a rollout; it is a diagnosis. WorkSpan builds the same activation breakdown Project Ground Truth ran across 92 partnerships, this time for your own AWS, Microsoft, Google, and Salesforce partnerships, anonymized to your team and mapped to outcomes: about two weeks to pull the data, two more for a delivered Atlas that sorts your partnerships by archetype and maps them one by one. You find out where each partnership actually sits, where the dormant pool is, and where depth is real versus where the dashboard is only counting deals. Everything after that is a decision made against your own data instead of a deck.

Book a demo →

Adoption Comes Before Technology

Even with the map in hand, the next constraint is human.

“So many of us in go-to-market are being told to reinvent our jobs while we’re doing our jobs. The quarterly target doesn’t go away, and now we have to reinvent our jobs on top of executing.”

This, then, is a leadership challenge first and a technical one second.

Superhuman, Not Replaced

The reframe worth ending on changes what the AI is even for.

“We’re getting it wrong when it comes to talking about AI. We talk about AI as replacing humans, versus, how do we use AI to turn us into superhumans?”

The point is people doing what one person alone never could. The mechanism does not remove the operator. It removes the ceiling on how many decisions the operator can be present for.

The Ceiling Just Moved

The rare and valuable skill in co-sell was always three things held together: knowing which partner fits, carrying the better-together story into the deal, and knowing the next move that advances it. All three lived in one person who could be in one room at a time, the ceiling every plateaued program eventually hit.

Six years of behavior, across 92 partnerships, point at the same undramatic thing: the programs that pull ahead build the discipline to be in the room at the moment it counts. They did not ask their sellers to care more; they made the partnership the easiest thing in the deal. What just changed is the ceiling on how many rooms that discipline can be in at once: the thinking encoded into a per-partnership agent, the presence carried by the Partner Advantage Card, the move executed by next-best-action across both companies’ systems, and the proof measured and fed back so the next one is sharper. The question left is whether your programs will be in the room when it counts, and now, for the first time, that is a decision you can make.

Recap & FAQ

Is closing the activation gap a technology problem or a leadership one?

Both, and leadership comes first. The mechanism works, but adoption is hard because teams are asked to reinvent their jobs while still hitting the same targets. Naming that cost, and framing AI as something that makes people superhuman rather than replaceable, is what gets a program through.

What is the single takeaway?

The programs that activate their sellers do it on operating discipline, not relationships or scale — the discipline to put the partnership in front of the seller at the moment they decide. That discipline used to live in one person who could be in one room at a time. The mechanism ends that singularity, carrying it into every deal at once.

Where do we start?

With your own map, not a purchase. The activation breakdown shows which of the five shapes each partnership is and where the depth is real — so the first decision is made against your data, not a deck.

What is a WorkSpan AI Agent?

A WorkSpan AI Agent is a per-partnership AI trained on one specific partner relationship, not the open internet: its playbooks, its rules of engagement, and the live transactional history of every deal the two companies have run together. That is what lets it do the work a seller has no time for, mapping a rep's territory against where the partner is strong and surfacing the lookalike accounts where the same play has already closed. One agent per partnership means coverage has no headcount ceiling, and the capability is coming online now, partnership by partnership.

What is the Partner Advantage Card?

The Partner Advantage Card is the in-CRM surface where WorkSpan's per-partnership AI becomes something a seller can act on directly: which partners are relevant to this deal, how they map to it, propensity to buy, which incentives to request, and a better-together story on demand. It lives in the right rail of the opportunity in Salesforce or Dynamics, where the seller already works, so the partner shows up inside the deal instead of in a portal no one remembers to open. It is live today, already carrying the partner into live deals at customers like Boomi.

What is Next-Best-Action in co-sell?

Next-Best-Action is the capability that turns a relevant partner into a specific move. It reads the account across both companies' systems of record and surfaces the single action most likely to advance this deal, a referral, a private offer, a joint meeting, then makes that action executable in a click, mapped natively into the partner's system so there is no re-keying. WorkSpan built the first version with a global systems integrator and a hyperscaler, and it reaches the market this fall.

WorkSpan’s mission is to help businesses achieve more together.

Cloud built the infrastructure of trust between businesses. AI is how that trust scales.

Sources: WorkSpan primary interviews with partner-revenue leaders across cloud, ISV, and SI ecosystems — including Mala Tejwani (DataStax), Jay McBain (Omdia), Mayank Bawa (WorkSpan), Greg Sarafin (Sarafin Advisory), Steve Lucas & Joe Estes & Dan McAllister (Boomi), Rob Moyer (BlueThread), Munish Khetrapal (Palo Alto Networks), Brian Bohan & Steven Partie (AWS / Qualtrics), Greg Petrossian (DataStax), Sam Gong (WorkSpan). Quantitative findings: Project Ground Truth, co-sponsored by AWS — six years of co-sell across 92 partner ecosystems and 5,709 named sellers. Deal-economics figures drawn from AWS co-sell benchmarks where noted.