Discovery-to-Demo Conversion Rate

• A useful denominator is unique opportunities or accounts with a completed, eligible discovery, not every dial, meeting invitation or CRM record. Salesforce defines discovery as an early fit-and-motivation conversation✓. 4
• A booked demo, held demo and qualified demo are different outcomes. Salesforce explicitly lists a demo alongside other possible next actions after discovery✓. 4
• Keep opportunity-stage forecasts and follow-up-call correlations out of the discovery-to-demo benchmark cell✓. 35

1. Define discovery and demo before calculating a rate

Salesforce calls discovery an early prospect conversation intended to assess fit and motivations✓. It also notes that a complex sale may require multiple discovery calls with multiple stakeholders✓. That means a simple count of discovery calendar events is not necessarily a count of distinct eligible opportunities. A single opportunity may generate several meetings while still being in discovery. 4

Choose an entity and a start event. For an opportunity-level measure, count each opportunity once when its first completed, substantive discovery meeting occurs. Exclude no-shows, internal preparation meetings and duplicate events. Record the prospect, account, owner, date and source. Decide whether existing customers and new-logo prospects belong in the same cohort. If not, split them before reporting the rate. This definition is an editorial measurement recommendation, not a number supplied by a vendor report.

Define the destination just as tightly. A demo booked is a scheduled future event; a demo held has actually occurred; a qualified demo has met your predefined fit criteria. Salesforce describes a demo as one possible next step, alongside a technical conversation or follow-up material✓. Therefore a low demo-booking rate can reflect a different appropriate route, not necessarily a failed discovery call. If your motion often uses technical workshops, track those as a separate valid next step. 4

2. Use a cohort formula that cannot quietly change

A practical booked-demo rate is: unique eligible opportunities with a demo booked within a stated follow-up window, divided by unique opportunities whose first eligible discovery was completed in the cohort. A held-demo rate substitutes a demo that actually occurred. Both use the same discovery cohort as the denominator. Write the window into the dashboard, such as 14 or 30 days, and let the entire cohort mature before finalizing the number.

For example, if 100 eligible opportunities completed discovery in September and 42 booked a demo within 30 days, the booked-demo rate is 42%. If 35 of that same group attended, the held-demo rate is 35%. Those are hypothetical arithmetic examples, not external benchmarks. The gap between booking and attendance is operationally important. Track cancellations, reschedules and no-shows separately instead of silently converting a booking into a held meeting.

If your process skips a demo for some valid opportunities, publish a second measure: the share with any agreed, qualified next step. Do not change the main denominator to only those who were offered a demo after seeing the results. That would select easier cases and make the reported conversion rate look better without improving the underlying workflow.

3. Why large sales reports are not a universal demo benchmark

Pavilion and Ebsta say their 2024 benchmark drew on 4.2 million opportunities, $54 billion in revenue, 530 companies and more than one million hours of conversations✓. That is a substantial dataset, but scale does not answer a different question. The cited figures describe the dataset, not a published rate of completed discovery opportunities that went on to hold a demo. 1

The report associates a canceled meeting with lower stage progression and two canceled meetings with a larger drop✓. It also reports that 29% of opportunities skip a stage and that skipped stages are associated with lower close likelihood✓. These are useful warnings about stage hygiene, but they are not discovery-to-demo conversion rates. They do not tell us the eligible discovery count or the number of those opportunities with a subsequent demo. Nor do they prove that eliminating cancellation or a skipped stage would cause the reported outcome. 2

The study's population is also not your pipeline. Its companies, opportunities and deal types may differ from yours✓. Use the report to challenge definitions and investigate friction, not to declare that your team is above or below a nonexistent universal discovery-to-demo benchmark. 1

4. A follow-up call is not necessarily a demo

Gong says its discovery analysis used 803,402 recorded web-conferencing sales meetings✓. It reports a negative correlation between slides in discovery and the likelihood of earning a follow-up call✓. It specifies that this association concerns the first call, which it labels discovery✓. None of those statements says that the follow-up was a product demo. 3

A next conversation may be another discovery meeting, a technical session, a pricing review or an executive alignment call. Turning 'follow-up call' into 'demo' would change the outcome being measured. Likewise, a negative correlation is not evidence that slides themselves caused fewer follow-ups. Sellers may choose slides for harder, more complex or less-qualified opportunities. If you want to test a new discovery format, compare similar audiences and predefine whether the outcome is a booked demo, held demo or qualified next step. 3

5. Do not turn a forecasting illustration into a rate

HubSpot's forecasting guide introduces its stage table with 'For example, your team may have established' stage figures✓. Its surrounding explanation says opportunity-stage forecasting uses a stage-specific likelihood to close✓. That is not the same as the probability of moving from discovery to demo. Even if a table lists a discovery stage and a demo stage, counts in a hypothetical forecast are not a measured cross-company transition rate. 5

A common mistake is to divide the demo-stage count by the discovery-stage count from a snapshot of an open pipeline. Such a ratio is affected by stage duration, stalled records, different entry dates and opportunities that skipped stages. It is not a cohort transition rate unless the records are explicitly linked and observed over a defined follow-up period. Build the rate from event history rather than a current-stage bar chart.

6. Segment and audit the result

Split the metric by inbound versus outbound, new business versus expansion, geography, company size and deal complexity. Salesforce's note about multiple stakeholder discovery in complex deals is one reason a single meeting count can mislead✓. A longer enterprise discovery path is not automatically worse than a shorter transactional one. 4

Audit a sample of CRM records every month. Verify that discovery really occurred, a demo invitation had an external attendee, the held event actually happened and the stage timestamps are ordered. Report both numerator and denominator beside each percentage. Flag small cohorts instead of celebrating unstable changes. Keep a versioned definition so a change in process or CRM automation is not mistaken for a performance shift.

Finally, follow the same cohort to opportunities won and revenue. An increase in demo bookings can be negative if it comes from sending poor-fit buyers to demonstrations. Conversely, a stricter discovery process may reduce the demo rate while increasing meeting quality. Those possibilities require your own data to adjudicate; the observational studies cited here do not settle them.

FAQ

How we researched this

We reviewed Pavilion/Ebsta's 2024 study overview and report, Gong's discovery-call analysis, Salesforce's discovery guidance and HubSpot's forecasting guide on October 1, 2026. The published sample sizes and examples are identified by source and date. We did not infer a discovery-to-demo rate from stage snapshots, follow-up calls or forecast probabilities. The absence of a matching numerator and denominator is an evidence limitation, not evidence that the true rate is zero.

Build a measurable handoff

Modern Inbound can help define the audience, discovery handoff and next-step reporting for a research-led outbound motion. Discuss a measurable sales process, or email [rish@moderninbound.com](mailto:rish@moderninbound.com).