Revenue operations

Revenue Operations

Marketing, sales, and customer success each optimize their own metric while the customer experience and the revenue number fall through the seams between them: marketing is measured on leads so it sends volume, sales is measured on closes so it cherry-picks, CS is measured on renewals so it inherits bad-fit customers. Each team is locally optimal and globally broken. RevOps fixes the seams - and the single highest-leverage fix is one funnel with one set of definitions the whole org signs.

Operating procedure

Definitions come first because every metric, SLA, and forecast downstream inherits them; instrumenting an undefined funnel just automates the disagreement.

Step 1: Gather inputs

1. The current funnel stages as each team describes them - collect all three versions; the deltas are the diagnosis.
2. The CRM and data stack, and where shadow spreadsheets live.
3. Current targets: revenue, pipeline, renewal.
4. Known seam failures - lead disputes, stalled handoffs, surprise churn. Label anecdotes as anecdotes.

Step 2: Define one funnel, org-wide

Write the full funnel end to end and get all three leaders to sign the same document:

Lead -> MQL -> SQL -> Opportunity -> Closed-Won -> Onboarded -> Renewed/Expanded

The funnel-definition rule: exactly one definition of each stage exists org-wide, written in observable criteria, owned by RevOps, and changed only by agreement of all three leaders. Most pipeline disputes are definitional, not real - two teams counting different things and arguing about the difference. A definition that requires judgment ("seems ready to buy") is not a definition; require observable facts (fields filled, actions taken, meetings held).

Step 3: Set the hand-off SLAs

Each seam gets a written SLA - what the sending team owes, what the receiving team owes, and the clock:

• Marketing -> sales: an MQL arrives with required fields complete; sales touches it within 24 hours (speed-to-lead decays fast - same-day contact converts several times better than next-week).
• Sales -> sales: an SQL is accepted or rejected with a reason within 48 hours; rejects route back with the reason coded, so definitions can be tuned on data.
• Sales -> CS: a closed deal arrives with a completed handoff doc (use-case, promises made, stakeholders) before kickoff; no doc, no kickoff scheduled.

Step 4: Establish one source of truth

• A single CRM/data model all teams use; shadow spreadsheets are found and retired, not tolerated.
• Governed data: deduped accounts, consistent required fields, enforced hygiene - bad data quietly corrupts every report and forecast built on it.
• Instrument the funnel so every stage conversion and stage velocity is measurable without manual assembly.

Step 5: Install shared metrics

Replace siloed vanity metrics with numbers that cross the seams:

• Pipeline coverage: open pipeline versus target. Healthy default is 3-4x for the current quarter (calibrate to your actual win rate: coverage needed ≈ 1 / win rate). Below 3x, escalate top-of-funnel now; padding above 5x usually means junk pipeline, not safety.
• Stage conversion rates and velocity - where deals stall and for how long.
• CAC and payback - ties marketing and sales to efficiency, not volume.
• Net revenue retention - ties CS and sales to expansion and to deal quality.
• Shared goals at the seams: marketing compensated partly on pipeline-sourced revenue, not just leads; sales partly on the quality of accounts passed to CS (e.g., 90-day retention of new logos).

Step 6: Build the weighted forecast

• Forecast from pipeline data weighted by historical stage-conversion rates, not from rep optimism.
• Run one revenue forecast that all three leaders sign - not three forecasts reconciled in the CFO's inbox.
• Track forecast accuracy every quarter and tighten the weights; a mature RevOps forecast lands within ±10% of actuals. Persistent misses in one direction mean a stage definition or weight is wrong - fix the model, not the narrative.

Step 7: Remove friction and run the cadence

• Map lead-to-cash and remove manual handoffs; automate routing, enrichment, and alerts so reps sell instead of data-entering.
• Keep the stack consolidated - tool sprawl recreates the silos RevOps exists to remove.
• Operating cadence: weekly cross-functional pipeline review; monthly funnel-health review (conversion, velocity, leakage); quarterly planning that allocates targets across the unified funnel.

Worked artifact: stage-definition table

Copy and fill; this one page ends most pipeline arguments.

STAGE DEFINITIONS - signed by [FILL: CMO] / [FILL: CRO] / [FILL: VP CS] - v[FILL]

Stage        Entry criteria (observable)                       Owner      Exit SLA
Lead         Known email + company, source recorded            Marketing  -
MQL          Fit score >= [FILL] AND [FILL: action, e.g.       Marketing  Sales touch
             demo request or pricing-page visit x2]                       within 24h
SQL          Discovery call held; need + timeline + budget     Sales      Accept/reject
             fields complete in CRM                                       within 48h
Opportunity  Named buyer engaged; next step with date in CRM   Sales      -
Closed-Won   Signature received; handoff doc complete          Sales      CS kickoff <=5
                                                                          business days
Onboarded    Activation milestone hit: [FILL: product event]   CS         -
Renewed      Renewal signed / expansion booked                 CS         -

Change control: definitions change only at quarterly planning, all three signatures.

Deliverable

Produce a RevOps blueprint: the unified funnel with the signed stage-definition table, hand-off SLAs with clocks, the single-source-of-truth data model and hygiene rules, the shared-metric set with the 3-4x coverage bar, the stage-weighted forecasting method with an accuracy target, and the weekly/monthly/quarterly operating cadence.

Do NOT

• Do not let any team keep a private metric or spreadsheet that contradicts the shared dashboard - one contradiction re-legitimizes all the silos.
• Do not optimize lead volume over lead quality; volume compensation is how marketing floods sales with junk MQLs.
• Do not forecast from rep gut; weight the pipeline by measured stage conversion.
• Do not write stage definitions that require judgment calls - "feels qualified" guarantees the dispute continues under a new name.
• Do not celebrate coverage above 5x; audit it for zombie deals instead.
• Do not add tools to fix process problems; sprawl recreates the seams.

Quality bar

• Every stage has observable entry criteria, one owner, and all three leaders' signatures.
• Every seam has an SLA with a clock (24h MQL touch, 48h SQL disposition, 5-day CS kickoff) and a coded reject path.
• One forecast exists, stage-weighted, with a tracked accuracy target of ±10%.
• Pipeline coverage is calibrated to actual win rate, not a borrowed multiple.
• The cadence calendar names owners for the weekly, monthly, and quarterly reviews.