HubSpot reports a 20% cross-industry sales close-rate benchmark, but its article does not publish the sample, denominator, or calculation method needed to reproduce the figure. Treat 20% as a directional reference, not a universal target. Your first measurement decision is not the percentage. It is deciding which opportunities belong in the denominator.
A team that wins 15 of 50 qualified opportunities has a 30% qualified-opportunity close rate. The same 15 wins divided by 200 deals created produces 7.5%. Both calculations can be correct because they answer different questions. A useful report names the metric, counts, cohort, pipeline, and observation window together.
This guide explains what the published benchmark means, how to create a denominator contract, how to calculate a mature cohort, and how to investigate a change without allowing qualification rules or AI scoring to silently alter the metric.
What is the average close rate for sales?
HubSpot publishes a 20% cross-industry figure, along with Software at 22%, Finance at 19%, and Biotech at 15%. The article attributes these values to HubSpot data and a 2024 survey, but the reviewed source does not expose the underlying dataset, sample, questionnaire, denominator, or industry methodology. These are therefore published company-reported figures, not independently reproducible industry norms. See HubSpot’s published close-rate benchmark with that limitation in mind.
Do not assume that exceeding 20% proves strong performance or that falling below it proves a problem. Sales motion, deal size, qualification rules, cycle length, lead source, and opportunity mix can all change the result. Compare your team with its own historical cohorts first, using one stable definition. Use an external benchmark only when its population and method are sufficiently similar to your business.
A benchmark is useful only after you can state who was counted, what event started the clock, and which outcome ended it.
Define the denominator before calling a metric close rate
Close rate is a ratio. The denominator determines what the ratio describes. A lead-conversion measure may count leads entering a funnel. A qualified-opportunity rate may count opportunities entering an agreed qualified stage. A deals-created rate may count every deal created during a declared cohort. Name the metric accordingly.
HubSpot’s own materials illustrate why this matters. Its editorial article calculates close rate as wins divided by qualified opportunities. Its current Sales ROI documentation defines deal close rate as closed-won deals divided by deals created. The current ROI methodology also describes aggregated Sales Hub customer data and controls for its own dataset. Do not assume those controls or that denominator apply to the separate 20% editorial benchmark.
Put the denominator in the metric name. “Qualified-opportunity close rate” and “deals-created close rate” should never appear as the same unlabeled “close rate.”
Before building a dashboard, sales operations and sales leadership should approve a metric contract containing:
- Numerator event: the definition of closed-won and the date that determines when the win occurred.
- Denominator event: deal creation, entry into a qualified stage, or another specific funnel event.
- Qualification and scope: required criteria, pipeline, segment, deal type, and any excluded categories.
- Cohort window: the dates used to select opportunities and the time allowed for them to mature.
- Exception rules: treatment of duplicates, reopened deals, repeated opportunities, missing dates, and multiple opportunities for one company.
- Version information: source system, export or calculation version, and calculation timestamp.
Save this definition with the CRM report or reporting table. A dashboard label alone can be interpreted differently by sales, marketing, and finance. A consistent CRM systems consulting approach can help document field ownership and reporting rules without changing the evidence behind the metric.
How to calculate sales close rate
Use the general formula below after filtering to eligible records:
Close rate = eligible closed-won opportunities divided by eligible opportunities in the declared denominator, multiplied by 100
For a hypothetical January cohort, suppose 15 of 50 opportunities entered the qualified stage during the month and all 50 have had the agreed time to mature. The qualified-opportunity close rate is 15 divided by 50, multiplied by 100, or 30%. Report the counts beside the percentage. Three wins from 10 opportunities is also 30%, but the smaller cohort gives a less stable trend.
A practical opportunity record usually needs:
- Opportunity ID and creation date.
- Qualification or qualified-stage entry date, if applicable.
- Closed-won status and close date.
- Pipeline, owner, lead source, segment, and deal type.
- Currency and deal amount when revenue analysis is also required.
Count-based close rate and revenue-weighted performance answer different questions. Keep deal amount as a separate measure unless the organization explicitly defines a revenue-weighted rate. Do not average individual representative percentages to produce a team rate. Calculate the team result from the combined eligible wins and denominator records.
Choose exception rules before calculation. Decide whether a reopened deal counts once or as a new selling cycle, whether duplicate opportunities are excluded, and how records with missing lifecycle dates are classified. Retain the exclusion count and reason so a data-handling change is not mistaken for a performance change.
Keep source data at the correct grain. Use one row per opportunity in the opportunity table, one row per opportunity-stage transition in stage history, and one row per aggregate metric in the benchmark summary. A summary key should include metric name, denominator definition, segment, period, pipeline scope, and source version. Do not use only segment and date, because different definitions or reporting runs can collide.
Make sales-cycle timing and segmentation comparable
A close-date report and a cohort report answer different questions. Counting deals closed during April describes wins recorded in April. Following opportunities qualified in January describes how the January cohort converted. In a long sales cycle, many January opportunities may still be open in April.
For a 90-day sales cycle, a qualification-month cohort should remain labeled immature until the observation window closes. Pending opportunities should not be treated as losses simply because the report was run early. Choose the date basis according to the decision:
- Close-date reporting: useful for understanding the wins recorded during a calendar period.
- Qualification-date cohorts: useful for evaluating opportunities that entered the funnel together.
Do not mix these bases in one trend line. Report wins, eligible opportunities, close rate, median days to close, and exclusions together. Segment by pipeline, lead source, deal type, or representative only when each segment has enough observations to interpret. A small segment can move sharply when one deal changes outcome.
Diagnose a low rate before changing the sales process
First check measurement quality: consistent qualification, complete lifecycle dates, correct attribution, and mature cohorts. Then inspect stage conversion, lead source, deal type, time to close, and recorded loss reasons. These views can distinguish poor-fit opportunities from a qualification rule that is too loose, a bottleneck in one stage, or a handoff that lacks a clear next step.
Ebsta and Pavilion’s 2024 B2B Sales Benchmark Report analyzed 4.2 million opportunities and reported that top performers were 412% more likely to have a next step defined, 843% more likely to overcome objections, and 218% more likely to self-source pipeline. These are relative-likelihood comparisons in a B2B study. They are not close-rate gains, percentage-point increases, or evidence that top performers close a stated multiple of deals. Use them as prompts for operational investigation, not as targets for a close-rate dashboard.
Choose one intervention tied to a confirmed issue. If qualified opportunities from one source repeatedly fail at the same stage, review the qualification criteria or handoff for that source. If opportunities stall without a dated next action, coach representatives to record one. Evaluate the change on a comparable, sufficiently mature cohort rather than reacting to one month’s percentage.
Use qualification rules and AI without corrupting the metric
Keep qualification and close-rate measurement separate. A score can prioritize review, but it should not silently change which records count as qualified or mark an opportunity closed-won. HubSpot documents fit, engagement, combined, and deal scoring in specified contexts, with availability dependent on subscription and object. Its lead-scoring documentation and AI scoring documentation describe particular capabilities, not proof that a score establishes sales readiness.
A bounded qualification workflow can use this sequence:
- Apply deterministic gates. Check required fields, valid contactability, consent where applicable, eligible geography, allowed values, hard budget limits, and a stable source ID. Use deterministic rules for duplicate detection and other non-negotiable constraints.
- Use AI only for ambiguity. AI may classify free-text use cases, summarize intent, or categorize objections. It should not decide hard eligibility or directly alter the close-rate denominator.
- Require a structured result. Store the source record ID, allowed classification, confidence, evidence reference, model or prompt version, review status, and reviewer decision when applicable.
- Route through an approved rule. A human-approved classification may feed a documented routing decision. Preserve human-edited fields and compare the current value and modification time before any automated write.
For example, this illustrative classification record has one row per classification attempt, not one row per aggregate metric:
{
"classification_attempt_id": "class-2026-04-02-9001",
"source_record_id": "web-7781",
"classification": "forecasting_consolidation",
"confidence": 0.86,
"evidence_reference": "source_record_id:web-7781",
"model_or_prompt_version": "qualification-v1",
"review_status": "pending",
"reviewed_at": null
}Validate the classification against an allowed-value list before routing. For CRM updates, use a stable source ID and an idempotency key that includes the source system, object type, account or tenant scope, and source record ID. Read the current CRM value and last-modified time, reject stale source versions, and avoid overwriting a newer human edit.
A lookup followed by a create is not sufficient duplicate protection when concurrent workers can process the same record. Use a database-enforced uniqueness constraint or transactional upsert in the integration layer where supported. Record every write attempt and send ambiguous matches to human review. HubSpot documents specific deduplication behavior and limitations, including for API-created companies, as well as unique-value property validation. Those controls should not be described as a guarantee of race-safe external writes.
- The source record ID, object type, tenant scope, and idempotency key are present and stable across retries.
- The classification is an allowed value and meets the documented confidence or review threshold.
- The current CRM value and last-modified time show that no newer human edit would be overwritten.
- The integration layer enforces uniqueness or uses a transactional upsert for concurrent runs.
- Stale versions, ambiguous matches, and failed writes are logged and routed to an exception queue.
Teams planning qualification workflows or controlled CRM write-backs can review HubSpot systems consulting. For a narrowly scoped AI classification or review workflow, AI agent services may be relevant. Neither service link is evidence that AI increases close rates.
Build a close-rate review that leads to a decision
Publish a compact report containing the close rate, numerator, denominator, metric definition, cohort or period, pipeline, segment, median days to close, and excluded-record count. Sales operations owns definitions and data checks. Sales leaders decide process changes. Representatives keep opportunity details and next steps current.
Review meaningful cohort conversion, stage progression, lead-source quality, and loss reasons monthly or quarterly according to the sales cycle. When a trend shifts, verify eligibility and maturity first. Then identify the largest confirmed friction point, assign one intervention, and test it on a comparable cohort.
A useful summary record might include the following fields:
- Metric: qualified-opportunity close rate or deals-created close rate.
- Counts: eligible wins, eligible opportunities, and excluded records.
- Scope: cohort dates, pipeline, segment, owner or source dimension.
- Timing: maturity status and median days to close.
- Provenance: source version, calculation timestamp, and definition version.
Keep benchmark summaries separate from citations. A summary identifies the aggregate metric; citation records identify the sources supporting it; calculation runs identify each extraction or reporting attempt. This separation prevents a rerun or second source from overwriting the identity of an earlier result.
Sources and measurement note
HubSpot’s 20% and industry figures are published claims whose underlying methodology was not available for independent validation in the reviewed article. HubSpot’s current Sales ROI documentation uses a deals-created denominator and a separate customer-data methodology. That difference is why the denominator contract should be stored with every internal result.
The Ebsta and Pavilion report is useful for its B2B scope and relative comparisons, but it is not evidence for a universal close-rate benchmark. Product availability and scoring options can depend on subscription and object, so check current plan documentation before configuring a workflow. The calculation examples and implementation fields in this article are illustrative operating guidance, not a claim that a vendor publishes the exact proposed schema.