Quick answer
Sales operations turns a sales strategy into a system people can run.
Sales ops owns the operating rules behind selling: process design, CRM governance, routing, planning, measurement, forecasting support, and the removal of avoidable work. It should make the next action clear, the pipeline trustworthy, and management decisions reproducible. It does not replace sales leadership, rep judgment, or a broader revenue operations function.
- Design
- Define stages, ownership, required evidence, and handoffs.
- Operate
- Keep data, routing, automation, and reporting reliable.
- Improve
- Find friction, test a change, and measure the result.
01 / Definition
Sales operations owns the sales system, not every sales decision
The function connects strategy to daily execution across the path from market plan to customer handoff.
Sales operations is the discipline that makes a sales organization easier to run and easier to inspect. It translates a go-to-market plan into territories, stages, definitions, permissions, workflows, dashboards, and review habits. It also maintains the feedback loop: when routing fails, a stage stalls, or a report stops matching reality, sales ops finds the operational cause and corrects the system.
That boundary matters. Sales ops can define what evidence is required to advance an opportunity, but the seller owns collecting and recording that evidence. Sales ops can prepare a forecast and test its assumptions, but the sales manager owns the team commitment. Sales ops can automate a handoff, but the receiving team owns accepting and serving the customer.
- 01
Plan the market
Turn strategy into capacity, coverage, territory, quota, and account-assignment models, with assumptions that leadership can inspect.
- 02
Capture and route demand
Define ownership, deduplication, response paths, consent handling, and exception queues so qualified work reaches the right person.
- 03
Govern the pipeline
Maintain stage definitions, required fields, activity standards, approvals, automations, and the CRM data model that supports them.
- 04
Support decisions
Produce pipeline, conversion, velocity, capacity, and forecast views with documented formulas and known data limitations.
- 05
Close the loop
Coordinate the won and lost handoffs, preserve reason data, compare outcomes with assumptions, and prioritize the next process improvement.
Decision rule: if a problem is about one seller’s judgment or coaching, the manager normally owns it. If it is a repeatable problem in process, data, tooling, or measurement, sales ops should own or coordinate the fix.
02 / Ownership
Use a RACI before work disappears between titles
This is a starting model, not a universal org chart. Keep one accountable owner per row and rename the roles to match your company.
| Operating decision | Sales ops | RevOps | Sales manager |
|---|---|---|---|
| Sales stages, fields, and exit evidence | R | A | C |
| CRM administration, routing, and sales automation | R / A | C | C |
| Territory, capacity, and quota model | R | A | C |
| Weekly deal and pipeline inspection | C | I | R / A |
| Forecast preparation and data-quality checks | R | C | A |
| Rep coaching and deal strategy | I | C | R / A |
| Lead-to-cash process across departments | C | R / A | C |
| Sales-tool evaluation and adoption | R | A | C |
Sales ops versus RevOps
Sales ops concentrates on seller productivity and the sales process. RevOps has the wider cross-functional remit: marketing, sales, customer success, finance, and the connected data from first touch through revenue realization. In a smaller company, one person may perform both roles; the scope still needs to be explicit.
Sales ops versus the sales manager
The manager owns people, deal judgment, coaching, and the forecast commitment. Sales ops owns the consistency of the operating environment and the evidence used in those decisions. A dashboard can reveal a coaching need, but it cannot conduct the coaching.
03 / Stage contract
A stage is a verified customer state, not a confidence label
Each stage needs entry evidence, exit evidence, and a named person responsible for the record.
| Stage | Entry evidence | Exit evidence | Responsibility |
|---|---|---|---|
| Accepted | An assignable record, valid contact path, source, owner, and reason it entered sales. | Fit reviewed; pursue or disqualify with a defined reason. | Rep records the outcome; manager owns exceptions; sales ops owns routing and validation. |
| Qualified | The team’s minimum qualification fields are complete and supported by a buyer interaction. | A documented problem, relevant stakeholders, next action, and mutually understood evaluation path. | Rep owns evidence; manager inspects quality; sales ops monitors missing or contradictory fields. |
| Evaluation | The buyer is actively testing fit against a stated use case or requirement. | Solution fit, decision process, unresolved risks, and next commitment are recorded. | Rep advances the deal; specialists own their assessments; manager approves unusual exceptions. |
| Commercial | The buyer has requested or accepted a commercial step and the approval path is known. | Executed agreement, explicit loss, or documented no-decision; no silent expiry. | Rep owns close data; manager owns concessions; sales ops governs approvals and stale-record rules. |
| Won handoff | A valid agreement and the required customer, product, value, and ownership data exist. | The receiving team accepts the handoff and any open commitments have named owners. | Sales owns completeness; the receiving team owns acceptance; RevOps owns the cross-system contract. |
Implementation test: two informed reviewers should reach the same stage decision from the record. If they cannot, the definition is too subjective or the required evidence is missing.
04 / Metrics
Write the formula before trusting the dashboard
Every KPI needs a cohort, time window, currency rule, inclusion policy, and data owner alongside the arithmetic.
| KPI | Exact formula | Caveat to document |
|---|---|---|
| Win rate | Won opportunities ÷ (won opportunities + lost opportunities) × 100 | Exclude open opportunities. State whether “no decision” is lost, separate, or excluded. Choose one cohort method, such as opportunities created in the period, and do not mix it with a close-date view. |
| Stage conversion | Opportunities from the same entry cohort that entered stage B ÷ opportunities that entered stage A × 100 | Define how skipped stages, reopened deals, backward movement, and duplicate opportunities are treated. Snapshot counts are not a cohort conversion rate. |
| Pipeline coverage | Qualified open pipeline amount for the target period ÷ remaining revenue target for that period | Use the same period and currency on both sides. Label weighted and unweighted coverage separately. If the remaining target is zero or negative, report not applicable rather than infinity. |
| Sales velocity | (Active qualified opportunities × average deal value × win rate as a decimal) ÷ average sales cycle in days | The result is modeled currency per day, not booked revenue. Inputs must describe a compatible cohort and stage boundary; a changing mix can move the result even when execution does not. |
| Average cycle length | Sum of days from the agreed start timestamp to closed timestamp ÷ number of closed opportunities | Name the start event. Report won and lost cycles separately, and show the median when outliers distort the mean. Use one time-zone and pause policy. |
| Forecast error | |forecast amount − actual amount| ÷ |actual amount| × 100 for one period; across periods, MAPE = (100 ÷ n) × Σ |(actual − forecast) ÷ actual| | Freeze the forecast at the same point in every period. Percentage error is undefined when actual is zero; use absolute currency error for that period and report directional bias separately. |
Metric hygiene: keep both the current dashboard and a dated metric dictionary. A renamed stage, changed qualification rule, or backfilled close date can break comparability without changing the chart’s appearance.
05 / Cadence
Run a different control loop at each time horizon
The purpose is to resolve exceptions quickly without turning every day into a strategy meeting.
Daily
Keep work moving
Review failed assignments, unowned records, broken automations, overdue handoffs, sync errors, and urgent data exceptions. Repair the cause where possible; do not make people clean the same symptom forever.
Weekly
Inspect pipeline truth
Reconcile stage evidence, aging, forecast movement, next actions, coverage, and manager exceptions. Review one operational improvement and assign an owner and verification date.
Monthly
Read the funnel as a cohort
Compare source, segment, stage conversion, cycle time, win and loss patterns, forecast error, stack usage, and data debt. Decide what to stop, simplify, test, or train.
Quarterly
Reset the operating model
Revisit capacity, territories, quotas, definitions, handoffs, permissions, tool cost, integrations, and roadmap assumptions. Version changes so historical reporting remains interpretable.
06 / 30 to 60 to 90
Build control before adding automation
The deliverable at day 90 is a repeatable operating system, not a large pile of configuration.
- 0 to 30
Establish the contract
Interview sellers and managers; map the real lifecycle; name object and process owners; publish stage, source, close, loss, and metric definitions; baseline routing failures and data gaps; inventory integrations, permissions, reports, and manual work. End with an agreed backlog and a rollback path for every proposed change.
- 31 to 60
Stabilize the system
Repair assignment and handoff failures; implement the minimum required fields at the point they become knowable; create exception queues; reconcile the forecast view; retire duplicate reports; document one source of truth per business object; and pilot one reversible automation with a named owner, alert, and recovery procedure.
- 61 to 90
Improve and transfer ownership
Compare the pilot with its baseline; train managers on the review cadence; measure adoption through outcomes rather than logins; document recurring controls; prioritize the next quarter by business impact and operational risk; and confirm that another operator can run the system from the written playbook.
Do not automate ambiguity: if people disagree on what a stage, owner, or successful handoff means, automation will distribute that disagreement faster.
07 / Team heuristic
Size sales ops by operational load, not a universal rep ratio
Heuristic: add clear ownership when recurring systems work begins displacing selling or management; specialize when the work has materially different skills and risk.
Named part-time owner
Simple motion, limited change
A capable operator may cover sales ops alongside another role when the pipeline, data model, reporting, and integrations are genuinely simple. Reserve recurring time and publish what is not covered.
Dedicated sales ops owner
Persistent operating workload
Dedicate the role when routing, forecasting, planning, CRM governance, automation, and manager support form a continuous queue or are repeatedly deferred.
Specialized or RevOps team
Cross-functional complexity
Split administration, analytics, planning, enablement, or systems work when one owner becomes a bottleneck. Add RevOps ownership when the governing problem spans marketing, sales, success, finance, or product-to-cash.
Document demand for a quarter: recurring controls, project work, incidents, support, planning, and analysis. Staff against that observed queue and required response time, not a benchmark detached from your process.
08 / Stack
Buy fewer systems with clearer contracts
A tool earns its place when it removes a defined constraint and remains governable after the demo.
Start with the data contract
Name the authoritative system for contacts, companies, opportunities, activities, products, contracts, and revenue. Define identifiers, ownership, permitted writers, sync direction, conflict handling, retention, and export. Integration quantity is less important than knowing what happens when one fails.
Prefer documented APIs, webhooks, scoped permissions, audit history, stable exports, and observable failures. Test a real workflow with disposable data, including a denied action and recovery, not only the happy path.
Price the operating model
Compare license and seat costs with implementation, administration, usage or credit charges, integration maintenance, data storage, enablement, and migration risk. A lower subscription can still cost more to operate; a broad platform can still be wasteful if the team uses only a narrow slice.
Before purchase, name the owner, the workflow being replaced, the success measure, the review date, and the exit plan. If the tool duplicates an existing source of truth, require a reason strong enough to justify reconciliation forever.
Continue the evaluation with the API-first CRM guide, AI-native CRM definition, CRM workflow examples, and CRM ROI model.
09 / Conduyt
Where Conduyt fits, and where it does not
The right CRM should support the operating contract you defined. It cannot define that contract for you.
Worth evaluating
- You want one CRM for contacts, activities, opportunities, tasks, stage rules, and operational reporting.
- You expect more users and want to compare a flat platform price with per-seat economics.
- You want documented programmatic access for your own integrations, automations, or governed AI workflows.
- You value scoped access, visible operating controls, and the ability to test a real workflow before committing.
Probably not the first choice
- Your existing CRM is reliable and the measurable cost of change exceeds the problem.
- You depend on a specific legacy or niche integration that has not been verified in your environment.
- You need a vendor to supply your sales process, management discipline, or data ownership.
- Your workflow requires an incumbent ecosystem or enterprise capability that Conduyt has not demonstrated.
Verify the current product before deciding: review features, the MCP interface, trust and control information, and published pricing. Then test your stage, permission, automation, report, and handoff with disposable records.
Primary source trail
First-party sources reviewed September 20, 2026: Salesforce Trailhead on sales operations responsibilities and the sales ops technology toolbox; Salesforce on RevOps versus sales ops; Salesforce Help on win rate, coverage, and days-to-close definitions; Salesforce on the sales velocity formula; and Oracle on forecast error measures. The RACI, stage contract, cadence, and 30/60/90 plan are practical operating models presented by Conduyt, not universal standards.
Frequently asked questions
What does sales operations do?
Sales operations designs and maintains the system behind the sales team: process, CRM governance, routing, planning, reporting, forecasting support, technology, and continuous improvement. Its job is to make selling work more consistent and management decisions more trustworthy.
What is the difference between sales operations and RevOps?
Sales operations focuses on the sales organization and the path from prospecting through close. Revenue operations has a broader cross-functional scope that connects marketing, sales, customer success, finance, and revenue systems. One person can cover both in a smaller company, but the responsibilities should still be named.
When should a company hire for sales operations?
There is no universal rep-count trigger. Establish a named owner when recurring CRM, routing, reporting, forecasting, planning, or automation work is being deferred or is displacing selling and sales management. Make it a dedicated role when that work becomes a persistent operating queue.
Which sales operations metrics matter most?
Start with metrics that describe flow and decision quality: stage conversion, win rate, pipeline coverage, sales velocity, cycle length, and forecast error. The formula, cohort, time window, and inclusion rules matter more than the number of dashboard tiles.
Can a CRM fix a broken sales process?
No. A CRM can enforce fields, route work, record evidence, automate defined steps, and expose exceptions. It cannot resolve disagreement about qualification, ownership, stage meaning, coaching, or customer handoffs. Define the operating contract first, then configure the CRM to support it.