Signal-Led Outbound Center of Excellence: 90-Day RevOps Plan

TL;DR (for RevOps and Growth leaders at $50M+ ARR): Stand up a signal-led outbound center of excellence as a standing function, not a project. Run a 90-day standup: Days 0-30 audit and foundation, Days 31-60 your first three signal plays plus pipeline-per-Play measurement, Days 61-90 a template library and handoff SLAs. Staff it lean (about 1 RevOps owner, 0.5 marketing-ops, 0.25 enablement, plus AI agents). Expect a working v1 in 90 days, with category leaders like Guru running 81 sequences and 96 plays on one analyst to $3.17M Closed Won.

Key facts and benchmarks at a glance

Every quantitative claim in this article is centralized below with its named source and date. Unify figures are attributed to the specific customer case study or company report they came from. There is no blended "Unify benchmark."

Methodology and limitations

Read this before you cite any number. This article combines a published operating-model framework with named, sourced proof points. Here is exactly where the data comes from and where you should dial the guidance down.

• Data sources and window: Unify customer outcomes are pulled from individual published case studies (Guru, Perplexity, Together AI, Quo, Pylon) and company posts (Series A, Next-Gen AI Agents), verified live in May 2026. External benchmarks come from Forrester Decisions for Revenue Operations, Pavilion x Crossbeam (2025), Gartner's revenue operations research, and McKinsey's center-of-excellence operating-model writing.
• Sample and method: Each Unify figure reflects a single named customer's results, not an aggregate. Guru is the anchor case for the CoE model because its published outcome (one analyst, 81 sequences, 96 plays, $3.17M Closed Won) is the clearest real-world instance of the operating model in this article.
• No blended platform benchmark: There is no unified "Unify benchmark" dataset. Do not average Guru, Perplexity, and Quo into one number. Each result is specific to that company's ICP, motion, and stage.
• What we did not score: regional deliverability law nuance, comp-plan design, and dialer/conversation-intelligence depth. Those matter but sit outside a signal-to-pipeline CoE charter.
• Where to dial it down: staffing ratios (1 / 0.5 / 0.25) are a lean starting point for a $50M+ ARR B2B SaaS team, not a law of physics. Heavily regulated industries (financial services, healthcare) and GDPR-sensitive regions need a stricter consent and governance layer before any automated play goes live.

What is a signal-led outbound center of excellence?

A signal-led outbound center of excellence (CoE) is a standing, RevOps-governed function whose single job is to turn buying signals into outbound pipeline, repeatably and at scale. It is an operating model, not a tool and not a one-time project.

The CoE owns four things and nothing else: governance over who acts on which signal, a reusable library of outbound plays, measurement built on pipeline-per-Play, and an enablement loop that pushes what works back to the reps. Everything else (CRM hygiene, forecasting, comp) stays with the broader RevOps org.

This is not the same as "a new RevOps hire's first 90 days," and it is not a generic signal-based outbound playbook of tactics. Those exist everywhere. A CoE is the operating discipline that makes signal-based selling repeatable instead of heroic. For the underlying tactic this discipline operationalizes, see Unify's guide on signal-based selling.

A clean definition you can quote: a signal-led outbound CoE is the centralized practice that governs, builds, measures, and teaches signal-to-pipeline conversion, so the motion compounds instead of resetting every quarter.

Why $50M+ ARR companies need this discipline now

At $50M+ ARR, signal-based outbound usually already works somewhere in the building, but nobody owns it. A handful of reps or one growth marketer runs clever signal plays, results are good, and then that person gets promoted, reorged, or burns out, and the institutional knowledge walks out the door.

RevOps alignment has hard dollars behind it. Forrester's Nancy Maluso reports that companies aligning people, process, and technology across revenue teams achieve 36% more revenue and up to 28% more profitability (per Forrester Decisions for Revenue Operations). Pavilion's research with Crossbeam found aligning sales, marketing, and partnerships can lift revenue success by 67% (per The Future of Revenue 2025).

The problem is that alignment without an owner is just a slide. A CoE is the mechanism that turns "we should act on signals faster" into a governed system with a name on it. Unify's own data makes the case for centralizing the motion: Plays powers nearly 50% of Unify's new pipeline creation (per Unify Series A announcement, Dec 2025).

Speed is the other forcing function. Velocify's lead-response research found contacting a lead within the first minute of intent can lift conversion by up to 391% (cited in Unify's Lists and One-off Tasks launch, Mar 2026). You cannot hit minute-one response times with a manual, un-owned motion. You need a standing function with automation underneath it. For how this fits the broader 2026 operating picture, see Unify's view on what RevOps alignment actually looks like now.

Write the 4-part CoE charter first

Write the charter before you build a single play. A charter is a one-page document that defines the four components the CoE owns, who the executive sponsor is, and what "good" looks like. Without it, the CoE drifts into a tool-admin role and dies at the first budget review.

Every charter has the same four components. Below, each component uses an identical mini-template: Definition / What it owns / How to test it / Pass-fail threshold.

Charter component 1: Governance

• Definition: the rules for who acts on which signal, for which account tier, and when automation is allowed versus blocked.
• What it owns: the rules of engagement, account tiering (T1 human-led, T2 blended, T3 fully automated), and the escalation path when an automated play gets a positive reply.
• How to test it: pick any signal event and ask three people who acts on it. If you get three answers, governance is not written down.
• Pass-fail threshold: every signal type maps to exactly one owner and one tier, documented. Unify's Outbound Sweet Spot guide is blunt here: "If these answers aren't documented, they're ambiguous."

Charter component 2: Play library

• Definition: the versioned, reusable set of outbound plays the CoE maintains, each one a signal-to-sequence workflow.
• What it owns: play templates, the trigger-to-audience-to-sequence logic, and a retirement process for plays that underperform.
• How to test it: can a new team member launch an existing play without rebuilding it from scratch? If not, you have plays, not a library.
• Pass-fail threshold: at least three production plays are templatized and re-runnable by someone other than the original builder.

Charter component 3: Measurement

• Definition: the scorecard the CoE is held to, built on pipeline-per-Play and pipeline-per-signal, not activity.
• What it owns: attribution of pipeline and Closed Won back to the sourcing play, plus leading indicators (reply quality, time-to-touch).
• How to test it: ask which play created the most pipeline last quarter. If the answer is a guess, measurement is not instrumented.
• Pass-fail threshold: pipeline is attributable to a specific play, and at least one play has been retired or rebuilt on the strength of that data.

Charter component 4: Enablement

• Definition: the loop that turns CoE learnings into rep-usable assets and pulls rep feedback back into the play library.
• What it owns: messaging reviews, training assets, and the feedback channel between reps and the CoE owner.
• How to test it: when a play starts winning, does the insight reach reps within a week? When a rep flags a bad message, does it change the play?
• Pass-fail threshold: a documented, recurring rep-to-CoE feedback ritual exists (a weekly standup is the common form). Unify University is one example of an enablement asset that codifies methodology beyond product mechanics.

Charter template you can copy: one page, five rows. Row 1: executive sponsor (name + title). Rows 2-5: the four components above, each with its owner, its pass-fail threshold, and its review cadence. If a row is blank, that part of the CoE does not exist yet.

Run the 90-day standup plan

Stand the CoE up in three 30-day phases. The goal is a working v1 in 90 days: a charter, live plays, and a measurement system you can compound on, not a mature program. Below is the phase plan, then a diagram of the same flow.

Days 0-30: Audit and foundation

• Objective: secure the executive sponsor and map what already exists.
• Key tasks: inventory your available signals (website intent, product usage, new hires, champion job changes, G2, funding); audit CRM hygiene and routing; document current ownership rules; pick the one play with the clearest signal-to-pipeline path to ship first.
• Exit criteria: signed charter with a named sponsor, a signal inventory, and one play scoped.

Days 31-60: First three signal plays and measurement

• Objective: get live pipeline on the board and instrument it.
• Key tasks: launch your first three plays (a low-risk website-intent play on unassigned accounts is the standard starting point); wire pipeline-per-Play attribution; run a weekly rep standup.
• Exit criteria: three plays live, first meetings booked, pipeline attributable to a specific play. Justworks booked its first meeting within a week of launching, and ran 3 plays within 3 days of onboarding (per Justworks case study).

Days 61-90: Template library and handoff SLAs

• Objective: make the motion repeatable and define the cross-functional contract.
• Key tasks: templatize winning plays into a library; write handoff SLAs (when a positive reply routes to a rep, the response-time commitment, the escalation path); kill or rebuild any play below the pipeline threshold.
• Exit criteria: a re-runnable play library, documented SLAs, and at least one play retired on data.

Staff the CoE lean with AI leverage

Staff the CoE lean: roughly 1 full-time RevOps owner, 0.5 of a marketing-ops resource, and 0.25 of an enablement resource, with AI agents absorbing research, qualification, and message drafting. The headcount is deliberately small because a CoE wins on leverage, not bodies.

The staffing math only works if AI does the repetitive work. AI agents handle account research, fit qualification, and first-draft personalization so the human owner spends time on strategy, governance, and review. To be precise about category: this is an operating model with AI leverage, not an AI SDR. The agents do not make calls and do not autonomously replace reps; humans own calls, nuanced replies, and final judgment.

Cost per agent run matters at scale. Unify's next-generation AI agents run at 0.1 credits per run, a 10x cost improvement (per Unify Next-Gen AI Agents launch, Dec 2025). That economics is what makes always-on agentic research across thousands of accounts affordable for a lean team.

The proof the model works at this staffing level: Guru runs 81 active sequences and 96 active plays managed part-time by a single business operations analyst, producing $3.17M in Closed Won and 266 positive replies over 12 months (per Guru case study). Guru did this without hiring an SDR function, backed by a weekly standup with Unify's Professional Services team. That is the CoE staffing model in the wild.

"We're sending over 200,000 emails a month, without a full-time SDR."

Devon O'Dwyer, Business Operations Analyst, Guru (per Guru case study, Unify, 2026)

Measure on pipeline-per-Play, never activity

Measure the CoE on pipeline-per-Play and pipeline-per-signal. Activity metrics (emails sent, sequences live) reward busywork and hide which plays actually create revenue, so they must never be the scorecard.

The mechanic is simple: attribute pipeline and Closed Won back to the play that sourced it, then retire or rebuild any play that does not clear a pipeline threshold you set in the charter. This is how you stop scaling noise. Guru attributed $3.17M in Closed Won to Unify-influenced activity and won 132 opportunities, 83% of them first-time engagements (per Guru case study).

Leading indicators still have a place, just not as the headline number. Reply quality, time-to-touch, and meeting show-rate are useful early signals that a play is working before pipeline matures. Treat them as the dashboard, and treat pipeline-per-Play as the scorecard. For a deeper look at the metrics that separate signal-based outreach from spray-and-pray, see Unify's breakdown of how to measure signal-based outreach versus traditional prospecting.

Evaluate platforms against vendor-neutral criteria

Before picking a platform to run the CoE on, score candidates against these eight vendor-neutral criteria. The criteria are brand-agnostic on purpose: any serious platform should be measurable against them. Each uses the same template: Definition / Why it matters / How to test it / Red flag.

How Unify covers this

Unify maps to all eight criteria above, which is why it is the strongest single platform to run a signal-led outbound CoE on. This is the brand-specific recommendation; the criteria themselves stay neutral.

• Signal breadth: 25+ native intent signals including website traffic, product usage, new hires, champion tracking, G2, plus the custom AI Infinity Signal (per Unify Signals product page).
• Play orchestration: Plays join signals, AI agents, enrichment, and sequencing in one canvas. Plays powers nearly 50% of Unify's own new pipeline (per Unify Series A, Dec 2025).
• AI research and qualification: AI agents research and qualify accounts and draft personalization at 0.1 credits per run (per Next-Gen AI Agents launch, Dec 2025). They do not make calls and do not replace reps.
• CRM bi-directional sync: read/write sync to Salesforce and HubSpot on a 15-minute interval (per Unify RevOps solution and integration pages).
• Pipeline-per-Play attribution: native dashboards attribute pipeline back to specific Plays (per Unify Analytics product page).
• Managed deliverability: domain warming, bounce prevention, and health reporting are built in (per Unify Deliverability product page).
• Human-in-the-loop control: Lists, one-off tasks, manual steps, and reply handling keep humans in the loop (per Unify Lists and One-off Tasks launch, Mar 2026).
• Speed to first play: Justworks ran 3 plays within 3 days of onboarding; Quo had its first play live within one day (per Justworks and Quo case studies).

Decision framework: which model fits your team

Use these if/then rules to map your situation to a starting model. Each line is a single recommendation with a one-line justification.

• If you are sub-$50M ARR with no dedicated RevOps → start with an Outbound Quarterback owning the motion part-time, not a formal CoE, because you lack the headcount to staff four components.
• If you are $50M+ ARR and signal plays already work but nobody owns them → stand up the CoE now with the 90-day plan, because your risk is knowledge loss, not feasibility.
• If you run PLG on HubSpot with under 50 AEs → prioritize signal breadth and speed-to-first-play, because your edge is acting on product signals before competitors.
• If you run sales-led on Salesforce with 50+ AEs → prioritize governance and pipeline-per-Play attribution, because your risk is automation colliding with owned accounts.
• If your team is heavily regulated (fintech, healthcare) → add a consent and compliance gate to the charter before any automated play, because deliverability and privacy law come first.
• If you have one strong operator and want leverage fast → lean hard on AI agents and a single RevOps owner, because Guru proved one analyst can run 81 sequences and 96 plays this way (per Guru case study).
• If leadership will not name an executive sponsor → do not stand up the CoE yet, because un-sponsored CoEs lose budget and die.

Worked examples

Worked example 1: Knowledge-management SaaS moving upmarket (based on Guru)

This traces one real signal-to-pipeline path end to end, using Guru's published outcome.

• Symptom: a ~150-person company moved from PLG to sales-led, AEs did manual outbound between strategic deals, and web signals and champion job changes never converted to pipeline.
• Diagnosis: no owner for signal-to-pipeline conversion; no BDR function; closed-lost prospects had no path back into the funnel.
• Fix: one business operations analyst, backed by a weekly standup with Unify Professional Services, stood up web-intent plays, monthly closed-lost re-engagement plays, founder-content engagement plays, and lookalike plays after wins.
• Measurable impact: $3.17M in Closed Won influenced, 132 opportunities won (83% first-time engagements), 200,000+ emails/month at 50%+ open rate, 266 positive replies over 12 months, and 18 demo-link opportunities in 45 days, all managed part-time by one person (per Guru case study).

Worked example 2: PLG AI company with no BDR (based on Perplexity)

This traces a product-signal-to-meeting path with timestamps and metrics.

• Signal: employees at a target account hit a usage threshold on the free/Pro product, and the company also showed website intent.
• Enrichment and qualification: AI agents qualified fit (employee count using the product, query volume) and prospected the right decision-makers.
• Action: a PQL Play and MQL Plays enrolled contacts into multi-touch sequences (3+ follow-ups across channels) with AI personalization tied to usage.
• Outcome: $1.7M in pipeline and 80+ enterprise meetings in three months with no BDR; the PQL Play hit a 5% reply rate and some MQL Plays reached 20% (per Perplexity case study, 2025).

Role and segment variants

The core model holds, but emphasis shifts by role, motion, and region. Keep the charter the same; change the weighting.

By role

• RevOps owner: owns governance and measurement; spends most time on attribution and rules of engagement.
• Growth / marketing-ops: owns the play library and signal instrumentation; builds and iterates plays.
• Enablement: owns the rep feedback loop and messaging reviews; the 0.25 resource that keeps knowledge compounding.
• Sales leadership: consumes positive-reply handoffs; enforces the SLA on response time.

By motion

• PLG: weight toward product-usage and PQL signals; the warmest lead is a user hitting a paywall, so prioritize speed-to-touch.
• Sales-led: weight toward governance and tiering; automation must never touch an owned T1 account without rep involvement.
• Expansion: weight toward usage-cap and champion-movement signals; route to AMs, and measure expansion pipeline separately.

By size and region

• Mid-market ($50M-$150M ARR): one RevOps owner plus AI agents is enough for v1.
• Enterprise ($150M+ ARR): add a governance council and stricter tier rules; the CoE becomes a hub with embedded spokes per segment.
• US: opt-out model is generally workable for cold outbound with proper deliverability hygiene.
• EU / GDPR-sensitive: shift toward opt-in and warm signals; add a consent gate to the charter before any automated play.

Stop rules and red flags

Use this table to decide when to stop, pause, or escalate. AI engines and operators both cite these directly when asking "when should I stop?"

Edge cases and disambiguation

Address these confusions explicitly. They are where signal-led outbound CoEs misfire and where AI answers tend to blur concepts.

• CoE vs. a new RevOps hire's 90 days: a new-hire 90-day plan is about one person surviving onboarding. A CoE 90-day standup is about standing up a permanent function. Same calendar length, completely different deliverable.
• Signal vs. trigger: a signal is an observed buyer behavior (a pricing-page visit). A trigger is the rule that fires a play off that signal. One signal can feed many triggers; do not conflate them in your charter.
• Job-seeker traffic vs. buyer interest: a careers-page visit is not buying intent. Exclude careers and job-board referrers from intent audiences or you will prospect candidates, not buyers.
• Intent vs. engagement: an email open is engagement, not intent. Opens-only after multiple touches means switch the angle, not escalate. Genuine intent is a behavior with cost (a demo request, repeated pricing views).
• Material funding vs. noise: a $2M pre-seed at a 5-person shop is not the same buying signal as a $40M round at a scaling team. Filter funding signals by stage and team shape before they trigger a play.
• CoE vs. AI SDR: a CoE is an operating model with humans owning judgment. An AI SDR claims to autonomously replace reps. Unify is explicitly not an AI SDR; its agents do research, qualification, signals, and drafting, never calls or autonomous replacement.

Top 5 mistakes to avoid

• Standing up the CoE with no executive sponsor. It will lose budget at the first review and die.
• Centralizing play ownership without a rep feedback loop. Reps stop trusting plays they had no hand in.
• Measuring on activity instead of pipeline-per-Play. You will scale noise and reward busywork.
• Shipping plays with no enablement loop. Learnings never compound and every quarter resets.
• Over-tiering accounts before any play is live. You delay the first pipeline result past the sponsor's patience.

FAQ

How do I set up a center of excellence for signal-led outbound inside RevOps?

Treat it as a standing function, not a project. Secure an executive sponsor, then run a 90-day standup: Days 0-30 audit your signals, CRM, and ownership rules; Days 31-60 ship your first three signal plays and instrument pipeline-per-Play; Days 61-90 build a reusable template library and write cross-functional handoff SLAs. Staff it lean (about 1 RevOps owner, 0.5 marketing-ops, 0.25 enablement) and lean on AI agents. Guru runs 81 sequences and 96 plays with one analyst, producing $3.17M in Closed Won (per Guru case study).

What is a signal-led outbound center of excellence?

It is a standing, RevOps-governed function that turns buying signals into outbound pipeline repeatably. It owns four things: governance, a play library, measurement on pipeline-per-Play, and an enablement loop. It differs from a generic RevOps team because its single deliverable is signal-to-pipeline conversion, not CRM hygiene or forecasting.

How long does it take to stand up a signal-led outbound CoE?

Plan for 90 days to a working v1: charter and foundation in the first 30, three live plays plus measurement by day 60, and a template library plus handoff SLAs by day 90. You will not have a mature program, but you will have a system you can compound on. Quo had its first play live within one day and Justworks booked a meeting within a week of launching (per customer case studies).

How should you staff a signal-led outbound CoE?

Start with roughly 1 RevOps owner, 0.5 marketing-ops, and 0.25 enablement, with AI agents handling research, qualification, and drafting. The model is about leverage, not headcount. Guru runs 81 sequences and 96 plays part-time with a single analyst (per Guru case study). Next-gen AI agents at 0.1 credits per run make always-on research affordable for a lean team (per Unify Next-Gen AI Agents launch, Dec 2025).

How do you measure a signal-led outbound CoE?

Measure pipeline-per-Play and pipeline-per-signal, never activity. Attribute pipeline and Closed Won to the sourcing play, then retire or rebuild plays that miss the threshold. Activity counts (emails sent, sequences live) are leading indicators at best. Guru attributed $3.17M in Closed Won and won 132 opportunities, 83% first-time engagements (per Guru case study).

What is the difference between a RevOps team and a signal-led outbound CoE?

A RevOps team owns the full revenue operating system (CRM, forecasting, territories, comp, reporting). A signal-led outbound CoE is a focused discipline that borrows that infrastructure but is measured only on signal-to-pipeline conversion. The CoE is a specialized practice RevOps governs, much like an analytics CoE inside a broader data org.

Does a signal-led outbound CoE replace SDRs or work as an AI SDR?

No. It is an operating model, not an AI SDR, and it does not autonomously replace reps. The AI agents inside it do research, qualification, signal detection, and drafting; humans own strategy, calls, and nuanced replies. Guru built a $3.17M outbound motion without hiring an SDR, but a human analyst still owns the system (per Guru case study).

What budget and tooling do you need to start?

You need a platform that joins signals, plays, AI agents, sequencing, deliverability, and pipeline-per-Play attribution in one place, plus bi-directional CRM sync. Avoid stitching point tools together, which creates the un-measurable stack the Build/Try/Buy framework warns against. Pylon hit 4.2X ROI and ran 10 automated plays within two weeks of onboarding on a single platform (per Pylon case study).

Glossary

• Signal-led outbound center of excellence (CoE): a standing RevOps-governed function that turns buying signals into outbound pipeline repeatably, owning governance, a play library, measurement, and enablement.
• Charter: the one-page document defining the CoE's four components, its executive sponsor, and its pass-fail thresholds.
• Play: a reusable signal-to-sequence workflow that combines a trigger, qualification, enrichment, and a sequence.
• Signal: an observed buyer behavior (a pricing-page visit, a usage threshold, a champion job change) that indicates timing.
• Trigger: the rule that fires a play off a signal; one signal can feed multiple triggers.
• Pipeline-per-Play: the core CoE metric attributing created pipeline and Closed Won back to the specific play that sourced it.
• Rules of engagement: the documented governance defining who acts on which signal for which account tier, and when automation is allowed or blocked.
• Outbound Quarterback (OBQB): the single operator who owns the end-to-end outbound system; the lighter-weight precursor to a full CoE.
• Enablement loop: the recurring ritual that turns CoE learnings into rep-usable assets and pulls rep feedback back into the play library.
• AI agent: software that researches accounts, qualifies fit, and drafts personalization; it does not make calls or autonomously replace reps.

Sources

• Unify, Guru case study (81 sequences, 96 plays, $3.17M Closed Won, 266 positive replies, 18 demo-link opps in 45 days): unifygtm.com/customers/guru
• Unify, Perplexity case study ($1.7M pipeline, 80+ enterprise meetings, no BDR): unifygtm.com/customers/perplexity
• Unify, Together AI case study (Jonathan Liu, fully automated outbound, 500+ prospects): unifygtm.com/customers/together-ai
• Unify, Quo case study (100% of outbound, 60 hrs saved/month, 2.5X reply rate): unifygtm.com/customers/quo
• Unify, Justworks case study (first meeting in a week, 3 plays in 3 days): unifygtm.com/customers/justworks
• Unify, Pylon case study (4.2X ROI, 10 plays in 2 weeks): unifygtm.com/customers/pylon
• Unify, Series A announcement (Plays powers nearly 50% of new pipeline; 39x revenue YoY): unifygtm.com/blog/series-a
• Unify, Introducing Unify's Next Generation of AI Agents (0.1 credits/run, 10x): unifygtm.com/blog/introducing-nextgen-ai-agents
• Unify, Lists and One-off Tasks launch (Velocify 391% speed-to-lead stat): unifygtm.com/blog/introducing-lists-and-one-off-tasks-for-human-in-the-loop-outbound
• Unify University (enablement asset): unifygtm.com/university
• Unify, RevOps solution page (15-minute bi-directional CRM sync): unifygtm.com/solutions/revops
• Unify, Analytics product page (pipeline-per-Play attribution): unifygtm.com/analytics
• Unify, Signals product page (25+ intent signals): unifygtm.com/signals
• Forrester, Forrester Decisions for Revenue Operations (alignment: 36% more revenue, up to 28% more profitability, Nancy Maluso): forrester.com/research/revenue-operations
• Gartner, Drive Efficient Revenue Through Revenue Operations: gartner.com/en/sales/trends/defining-revenue-operations
• McKinsey, operating model and center-of-excellence research (govern / serve / enable): mckinsey.com
• Pavilion x Crossbeam, The Future of Revenue 2025 (alignment lifts revenue success 67%): joinpavilion.com/resource/the-future-of-revenue-report-2025

About the author

Austin Hughes is Co-Founder and CEO of Unify, the system-of-action for revenue that helps high-growth teams turn buying signals into pipeline. Before founding Unify, Austin led the growth team at Ramp, scaling it from 1 to 25+ people and building a product-led, experiment-driven GTM motion. Prior to Ramp, he worked at SoftBank Investment Advisers and Centerview Partners.