The Life of a Partner Ecosystem: Why Consultancies Get Bought When They Do

We have now mapped eighteen software partner ecosystems: 1,364 verified partner-firm deals between 2015 and 2025 across the eleven we count deal by deal, from SAP’s 369 to NetSuite’s 29, plus the IBM channel’s 46 traced transactions, the Snowflake, Databricks and Atlassian Marketplace ecosystems and the Shopify agency market covered in dedicated reports, and the youngest of all, the AI platforms, OpenAI’s deployment layer and Anthropic’s Claude Partner Network, mapped before their partner M&A begins. Put side by side, the eighteen markets stop looking like eighteen stories. Every ecosystem moves through the same four phases, and the phase it is in determines who buys its partners, what they pay, and when an owner should sell. For an owner, reading that phase correctly is worth real money: it decides which buyers are at the table and which are already gone, and the difference between selling into scarcity and selling after the consolidators have filled their gaps is the difference between a competitive process and a single offer.

The Four Phases

Phase One: The Open Door

A new platform needs implementation capacity faster than it can build a services arm, so the vendor recruits anyone. Joining the partner program is a form and a fee, the early partners are generalists adding a logo to a wall of logos, and every partner’s real business is still somewhere else. A specialist appears only once the platform can feed one, and that point is still years away. The AI platforms sit in this phase today: Anthropic tiers its Claude Partner Network from the start, and OpenAI is building its deployment layer by acquisition.

Phase Two: The Crowd

The platform starts winning accounts, and the partner count explodes. Entry is still easy, and the work is still done by generalists learning the platform on the client’s clock. Certifications exist but discriminate little; the directory grows faster than the delivery quality behind it.

Phase Three: The Gate

The vendor tightens the program: certified-headcount thresholds, specializations, tiers that take years to climb, controlled entry. Getting in becomes hard, and this is the phase with the sharpest imbalance, because genuine specialists stay scarce while demand for them compounds. The firms that committed early hold positions that would take a newcomer years to replicate at any price.

Phase Four: Maturity

Specialists finally exist in numbers, and the partner count starts falling rather than rising: consolidation removes firms faster than the gated program admits them, and new entrants lose interest because the land grab is over. The ecosystem stops being a frontier and becomes a market share table, and the independent specialist, the firm every buyer wants, becomes the scarcest thing in it.

Why the Lifecycle Sets the M&A Cycle

The deals wait for scarcity. Through phases one and two, a buyer who wants the capability can simply join the program and hire, so acquisitions stay rare. The buying begins at the gate, and it begins with the incumbents.

Incumbents Buy Because Building Fails Them

The pattern repeats across the whole series: a global integrator or IT major that could hire thousands decides instead to pay for a few hundred certified consultants. Building the competency in-house fails for structural reasons, and every buyer knows them. Certification tiers take years to climb and the vendor controls the clock. The scarce people prefer specialist firms, where the platform is the business rather than a line item. And a large firm’s economics work against incubation: a twenty-person team inside a hundred-thousand-person integrator is invisible to the vendor, to reference clients, and to the partner-of-the-year shortlist. So the majors wait until phase three and buy their way in: Infosys took GuideVision to enter ServiceNow, Wipro took Appirio, Cognizant took Collaborative Solutions to enter Workday at scale, and every hyperscaler page we publish shows the same entry purchases. Buying replaces joining once joining has become hard, and the seller’s price reflects exactly that.

Roll-Ups Need a Confluence

Somewhere in phase three a different buyer appears: the sponsor assembling certified firms into a platform, the Thirderas and NewRockets of ServiceNow, TopBloc’s backers in Workday, Qodea in Google Cloud, the country consolidators of the Dynamics market. It is tempting to read roll-ups as a phase every ecosystem reaches, and the record says otherwise. A roll-up needs a confluence: enough fragmented specialists to assemble, scarce against demand yet numerous enough to buy in series, a certification moat that makes the assembled capacity defensible, a recurring revenue share that underwrites the leverage, a vendor that tolerates or even backs consolidation, and a visible exit, a strategic buyer who will pay for the finished platform. ServiceNow had all five, and its vendor co-invests in the consolidators through ServiceNow Ecosystem Ventures. Adobe’s ecosystem, mapped on its own page, skipped the ecosystem roll-up entirely: the specialists there are agencies whose capability spans platforms, and the buyers assembling them are agency holding companies, so the moat never formed around the platform itself. Oracle’s market shows the third shape: sponsors bought services groups and let the product label follow the customer, a services roll-up rather than an ecosystem roll-up.

Maturity Is the Endgame in Two Acts

First the consolidation: serial buyers work through the specialist supply, the application layer consolidates separately from the consultancies, and the accountants, resellers and regional groups mop up their home markets. Then the scarcity: the pure specialists run out. Workday is the clearest case we cover, and its page records the numbers: pure firms at real scale in single digits, four of them already sold, a market that went nearly silent through 2023 and 2024, and the buyers of 2025 and 2026 competing for what remains. For a seller, the lesson of the whole cycle compresses into one line: the best time to sell is when your kind of firm is what the current phase makes scarce.

The Ecosystems We Cover

Ecosystem	Verified deals 2015-2025	Partners in program	Typical buyers	Lifecycle position (our judgment)
SAP	369	~24,000	global integrators led by NTT DATA, PE roll-ups, listed SAP specialists	maturity, deep consolidation
Microsoft Dynamics	210	~1,400 Dynamics-focused	country consolidators like 9altitudes and Columbus, PE platforms, regional IT groups	maturity
Salesforce	197	~3,400 consulting partners	global integrators, now TCS, Inetum and sponsor-backed mid-market platforms	maturity
Oracle	115	20,000-30,000	PE-backed services platforms, IBM and Accenture	maturity
Microsoft Azure	106	inside Microsoft’s 500,000+	PE-built Microsoft platforms like 3Cloud, global integrators	late gate
ServiceNow	97	~2,200, six Global Elite	global integrators, sponsor pure-play platforms, resellers, European IT groups	gate
AWS	96	~140,000	resellers and managed cloud groups, sponsors, integrators	gate
Google Cloud	55	~120,000; ~2,900 services partners	sponsor platforms like Qodea, integrators making a first purchase	crowd to gate
Adobe	52	tiered program, count not published	agency networks led by dentsu and Omnicom, IT majors, PE platforms	maturity
IBM	46 traced	four lanes, one vendor	CDW and H.I.G.’s Pellera among resellers, two Maximo consolidators, software buyers	maturity, endgame run
Workday	38	~150 services partners	HR, payroll and accounting firms, staffing groups, PE	maturity, scarcity endgame
NetSuite	29	~800	accounting and CFO-advisory firms, Zone & Co on the application layer	gate to maturity
Snowflake	mapped in the report	1,000+ services partners, 93 Elite	PE building data platforms, integrators; the vendor invests in its own channel	crowd to gate
Databricks	mapped in the report	~4,000 directory firms, pure-plays under 1%	PE, global integrators, AI labs; first pure-play roll-up under way	crowd to gate
Atlassian Marketplace	mapped in the report	1,800+ partners, 8,000+ apps	PE-backed app consolidators, Appfire near 200 apps, Tempo behind it	gate to maturity, on the app layer
Shopify agencies	mapped on the page	open program	global consultancies and PE groups, buying since 2020	crowd to gate
OpenAI	vendor-side buying has begun	deployment layer forming	OpenAI itself bought Tomoro to anchor a $14B deployment company	open door
Claude (Anthropic)	the market is forming	tiered from the start: Select, Preferred, Global Premier	early entrants building positions; partner M&A has yet to begin	open door

Program sizes count different things, and the differences are the point. The hyperscaler figures count every partner type, resellers and ISVs included; the Salesforce, Dynamics, Google Cloud services, Snowflake and Workday figures count services firms only; SAP’s includes free registrants. Figures are 2024-2026 vintages from vendor statements and partner-directory tallies, sourced in our research notes. IBM’s count is traced web-first from acquirer announcements rather than database-counted, and its page says so.

What Partners Sell For, by Ecosystem

Ecosystem	Median deal size (disclosed)	Revenue multiple	EBITDA multiple	Nearest benchmark	Reading
SAP	$12M (n=31)	1.2x (n=31)	7.1x (n=15)	systems integration	benchmark-grade samples
Microsoft Dynamics	$14.4M EV (n=33)	1.06x (n=12)	8.63x (n=8)	systems integration	published with samples
Salesforce	$35.3M EV (n=26)	1.97x (n=6)	–	IT services	multiple is a direction, above other services lines
Oracle	$22M (n=18)	1.3x (n=4)	–	IT services	thin multiple sample
Microsoft Azure	$15M (n=17)	–	–	cloud services	too few multiples to publish
ServiceNow	$85M (n=7)	–	–	IT services	skews to the platform trades
AWS	$50M (n=12)	–	–	cloud services	skews large; no reliable AWS multiple exists
Google Cloud	$248M (n=7)	–	–	data & analytics and cloud services	only the giants disclose
Workday	platform trades $110M-$500M	–	–	IT services	five prices, all at the ceiling
NetSuite	not published (n=3)	–	–	IT services; apps on software	prices stay private
Adobe	$8.1M (n=11)	–	–	IT services	skews small, listed buyers disclose
IBM	no median; three prices $1B-$2.5B	–	–	IT services or software, by lane	four lanes, incomparable populations
Snowflake / Databricks	$50M (data & analytics comps)	–	16.3x (data & analytics comps)	data & analytics	the highest-valued subsector in IT services
Atlassian Marketplace	mapped in the report	–	–	software and SaaS	app businesses price on product revenue
Shopify agencies	mapped on the page	–	–	IT services	people businesses, priced as services
IT services benchmark	$21M (n=1,051)	1.1x	10.4x	the reference for everything above

Each row shows only what its page publishes, with the sample stated. Two ecosystems carry enough disclosures to read as benchmarks, SAP and Dynamics. The rest are directions at best, and where disclosure skews to the largest deals we say so rather than print a misleading number. One pattern in the column is worth naming: the deep, mature markets, SAP and Dynamics, price at the IT services baseline on revenue, 1.2x and 1.06x against 1.1x, while the biggest disclosed medians, Google Cloud’s $248M, Workday’s platform range, ServiceNow’s $85M, come from markets where only the platform trades reach the public record.

Set the two number columns against each other and the lifecycle shows up as a ratio. Workday has seen 38 partner acquisitions against a program of roughly 150 services firms, so buying on a scale of a quarter of today’s program has already happened, and that is what the endgame of the tightest program looks like. ServiceNow runs at 97 deals against 2,200 partners, SAP at 369 against 24,000, and AWS at 96 against 140,000. The tighter the gate, the larger the share of the ecosystem that has already been bought. The comparison is rough, the denominators count different things, and a hundredfold gap in deal density survives any counting choice.

Where This Leaves an Owner

Read your own ecosystem off the chart before you read any offer. The phase names your buyers: in a gated market the integrators pay for entry and the sponsors pay for platform potential, and in a mature one the consolidators pay for fit, a client book, a country, a capability gap. The phase also sets your leverage. A specialist in phase three sells what a buyer can obtain in one other way only, years of certification work, and a firm in phase four sells into a market that knows exactly what it is worth. The worst position is the one the lifecycle punishes: holding a scarce firm through the window where scarcity peaks, and selling after the consolidators have filled their gaps. Eleven ecosystems counted deal by deal, 1,364 verified transactions, the IBM channel traced beside them, four mapped reports, the AI networks mapped at the door, one cycle: the platform-by-platform detail, who the buyers are, what they pay, and who buys next, sits in each linked page.