Where Growth Operations came from
Growth Ops did not emerge from a McKinsey deck. It emerged from a very specific gap that opened between 2023 and 2025 as revenue teams tried to absorb generative AI without blowing up their existing operations.
Look at what the adjacent functions were doing:
- RevOps got stuck in reporting. Forecast accuracy, funnel dashboards, quota planning, board decks. Important work. Not the work that ships new AI-driven motions.
- Marketing Ops got stuck in campaign attribution and MAP hygiene. HubSpot, Marketo, six-touch attribution models, UTM taxonomies.
- Sales Ops got stuck in CRM hygiene, territory design, and comp plan administration. Essential plumbing that has nothing to do with signal orchestration.
Meanwhile the interesting work — the work that actually moved pipeline in 2024 — sat in the seams. Wiring Clay to Apollo to Outreach with a custom scoring model. Standing up an internal AI SDR that drafts personalized sequences from a buying signal. Turning a job-change alert into a warm intro request routed to the right rep in Slack. Building relationship-graph enrichment that finds the 60-80% of warm paths your CRM never captured.
Nobody in the existing org chart owned that layer. The result was a chaotic scramble where SDR managers built n8n workflows on the side, senior AEs bought Clay seats out of their own commission, and the CRO's dashboard still showed the same three funnel stages while the underlying motion was quietly being rebuilt by whoever had time.
Gartner made the shift official. In 2024 they formalized Revenue Action Orchestration (RAO) as a distinct category and published a Market Guide for GTM Data Applications. The functional owner they described looks nothing like a classic RevOps analyst. It looks like a systems engineer with a go-to-market brain. That is Growth Ops.
The forcing function was performance. Gartner's own numbers tell the story: 45% of CSOs missed their 2024 goals. 60% of B2B buyers report deal regret. AI tools promise to save reps five hours a week, but 72% of that time is wasted because the workflow around the tool is broken. The gap between "we bought AI" and "AI moved the number" is exactly the gap Growth Ops fills.
The Growth Ops charter
The function has a specific scope. When companies get this wrong they either bloat Growth Ops into a shadow product org or shrink it into a glorified Zapier team. The charter has five areas.
GTM system architecture
Growth Ops owns the map of how data flows across the stack. Data source to enrichment to signal to routing to sequence to CRM back to reporting. When a new tool gets added, Growth Ops decides where it plugs in and what it replaces. When a workflow breaks, Growth Ops is the on-call.
AI workflow deployment
This is the load-bearing new responsibility. AI SDRs, meeting summarizers, deal-risk scoring, next-best-action prompts, custom GPTs trained on your ICP. Growth Ops ships these workflows, measures them, and kills the ones that do not lift a metric. Marketing does not own them because they are not campaigns. Sales does not own them because they are infrastructure. RevOps does not own them because they are prospective, not retrospective.
Signal-to-play mapping
A signal is only worth what the play behind it produces. Growth Ops owns the translation layer: when a champion changes jobs, what does the rep do in the next 24 hours? When intent data spikes on a target account, who gets pinged and with what context? See the warm-intro signal library for the shape of this work. The rule is simple: no signal ships without a play attached.
Relationship data hygiene
This is where Boomerang plugs directly into the Growth Ops charter, and it is where most stacks quietly bleed pipeline. Your CRM captures the relationships your reps remembered to log. It does not capture the relationships your alumni network, your board, your investors, your former customers, or your extended team actually hold. Our data shows CRM undercounts warm paths by 60-80%. Growth Ops owns closing that gap — because if your relationship graph is broken, every downstream play is guessing.
Cross-team metric alignment
Marketing counts MQLs. Sales counts pipeline. CS counts NRR. Growth Ops owns the metric layer that stitches these together into a single view of revenue efficiency. If marketing's MQL definition and sales' SQL definition do not reconcile, that is a Growth Ops problem, not a "let's have a meeting" problem.
The 5 skills a Growth Ops leader needs
I have hired for this role and interviewed for it. The five skills below are the ones that separate a Growth Ops leader from a very senior operator who cannot ship.
- Systems thinking. Can you draw the full data flow from lead source to closed-won on a whiteboard from memory? Not the marketing funnel — the actual system. If you cannot, you cannot lead this function.
- SQL and Python. Not to become an engineer, but to build the workflow layer yourself when needed. A Growth Ops leader who cannot write a SQL query or a Python script waits on engineering for six weeks and misses the quarter.
- Workflow orchestration. Fluency in n8n, Workato, Zapier, and native APIs. The specific tool matters less than the ability to model a workflow, ship it, and iterate.
- AI prompting and fine-tuning. In 2026 this is table stakes. You should be able to build a custom GPT, evaluate it against a rubric, and know when to switch to a fine-tuned model or an agent framework.
- Change management. The technical rollout is 30% of the job. Getting sellers to actually use the new workflow is 70%. If your Growth Ops leader cannot sit in a QBR and coach an AE through the new motion, the workflow ships to nobody.
Growth Ops vs RevOps vs Sales Ops vs Marketing Ops
The single most common question I get is "how is Growth Ops different from RevOps?" The table below is the answer I give.
| Function | Primary output | Time horizon | Key tool ownership | Reports to |
|---|---|---|---|---|
| Growth Ops | Shipped workflows, AI plays, signal-to-play maps | Next quarter and beyond | Clay, n8n, Boomerang, custom GPTs, agent frameworks | CRO or CTO |
| RevOps | Forecast, funnel reporting, planning, quota | Last quarter, current quarter | Salesforce, BI stack, forecast tooling | CFO or CRO |
| Sales Ops | Territory, comp, CRM hygiene, quota admin | Current quarter | Salesforce, CPQ, comp tooling | CRO |
| Marketing Ops | Attribution, MAP hygiene, campaign ops | Current quarter | HubSpot, Marketo, attribution stack | CMO |
The load-bearing distinctions: Growth Ops is prospective and workflow-first. RevOps is retrospective and reporting-first. If you do not have a clear owner for the prospective, workflow-first work, one of two things is true. Either you are small enough that it does not matter yet, or your CRO is quietly doing this work at 11pm on Sundays. Neither is a permanent answer. For a deeper comparison see Sales Ops vs RevOps in 2026.
The 2026 Growth Ops tech stack
Growth Ops does not have "one stack" any more than engineering has "one framework." But the layering is consistent across the teams doing this well. Below is the reference architecture I share with founders and heads of revenue.
| Layer | What it does | Representative tools |
|---|---|---|
| Data layer | Firmographic, technographic, contact data | Clay, Apollo, ZoomInfo, UserGems |
| Signal layer | Intent, job change, hiring, funding, warm-path signals | Bombora, G2, Boomerang, LinkedIn Sales Navigator |
| Orchestration layer | Route, score, enrich, decide next action | n8n, Workato, Boomerang, native APIs |
| AI layer | Draft, summarize, score, prioritize, decide | Custom GPTs, internal AI SDRs, agent frameworks |
| Execution layer | Where the play actually lands | Outreach, Salesloft, Gong, Salesforce |
Two things to notice. First, Boomerang shows up in two layers — the signal layer (warm-path signals your CRM does not surface) and the orchestration layer (routing an intro request from signal to booked meeting). That is not a coincidence; the whole platform was designed to sit inside a Growth Ops stack rather than replace one. Second, the AI layer is where the moat compounds. Anyone can buy Clay. The team that trained a custom scoring model on their own closed-won cohort has an edge nobody else can copy. See AI-driven pipeline acceleration across deal stages for the stage-by-stage picture.
The Growth Ops moat
Here is the contrarian point I want to leave you with. The moat is not the tools. Everyone can buy Clay. Everyone can buy Apollo. Everyone can subscribe to ChatGPT Enterprise. The moat is the specific combination of workflow orchestration, custom AI, and proprietary relationship data that your Growth Ops team assembles into a motion your competitors cannot replicate.
Two customer stories make this concrete.
Armis. The Armis Growth Ops team built a warm-path activation motion on top of Boomerang that surfaced 26,000+ warm paths their CRM had never captured. They compressed sales cycles by activating those paths systematically — not one rep at a time, but as an org-wide play routed through their orchestration layer. The result: 10× ROI, 1,400+ hours of rep time saved, and a measurable lift in enterprise deal velocity. The tool did not do that. The Growth Ops motion around the tool did. This is the pattern for relationship intelligence at enterprise scale.
Narvar. Narvar's team stood up a warm-intro motion in the first quarter after deploying Boomerang and generated $800K in pipeline within three months. What made that possible was not "the software worked" — it was that Narvar had a Growth Ops muscle capable of wiring signal to play to rep, of getting sellers to actually run the motion, and of measuring the output honestly. Same tool, same market, without that muscle: nothing happens.
The pattern in both cases is the same. Buying signal + relationship graph + orchestration + rep execution = pipeline. Break any link in that chain and the motion collapses. Growth Ops is the function that owns the whole chain. This is also why "will AI replace sales" is the wrong question in 2026. The right question is which teams have the Growth Ops muscle to make AI actually pay off, which we break down in AI in B2B sales — what works vs hype.
When to hire your first Growth Ops leader
The single most common hiring mistake is bringing in a Growth Ops leader too early. Two examples of when to hold and when to hire.
Do not hire at Series A. At Series A you have a founder-led sales motion, a couple of AEs, maybe an SDR, and a CRM that a well-organized ops generalist can run. You do not have enough workflow surface area to justify a specialist. Hiring a Growth Ops leader at Series A means either they get bored, they build for a scale you do not have, or they become a very expensive Salesforce admin.
Hire at Series B, around $10M ARR. This is the inflection point. You have three or more segmented motions (SMB, mid-market, enterprise, or product-led plus sales-led). You have five or more tools that need to talk to each other. Your CRO is spending Sunday nights fixing routing rules in Salesforce. Your best SDRs are secretly running Clay tables their manager does not know about. That is when Growth Ops earns its keep.
Reporting line: CRO or CTO. The answer depends on where AI orchestration ownership lives in your company. If AI is a CRO-owned initiative, Growth Ops reports to the CRO. If AI is engineering-led (increasingly common at technical companies), Growth Ops reports to the CTO with a dotted line to the CRO. Either works. What does not work is reporting to a VP of Sales — the mandate is too broad for that reporting line.
.png)
