The Booked-Meeting Acceptance Standard: Accept, Remediate, or Reject
Grade any booked meeting Accept, Remediate, or Reject with one reason code from attached evidence, and prove two graders agree on at least 18 of 20.
Key takeaways
- A booked meeting should earn credit only when it is held, defined as the prospect present for at least 15 minutes with the status logged in the CRM within 24 hours.
- Acceptance benchmarks only compare when the denominators match: the Bridge Group's derived 60% is measured on held meetings, while the second-hand RepVue figure of 56-67% is measured on meetings booked.
- Account criteria must be checked against the ultimate parent matched by web domain, because a converted lead becomes a fresh account in the CRM unless something matched it to the existing record first.
- A rejection window only works if silence counts as acceptance, because the AE controls the CRM record and the SDR otherwise loses every dispute by default.
- Two graders agreeing on 18 of 20 meetings is not enough on its own: at 90% agreement, kappa is about 0.79 on a 14/3/3 grade split but only about 0.46 on an 18/1/1 split.
- In Refolk's index, 45,735 US profiles and 5,793 UK profiles carry SDR or BDR titles, and the naming splits differently in each market, so policy should define the booker by how they are paid, not by job title.
This standard sets the bar for when a meeting booked by an SDR or a pay-per-meeting agency counts. It is for the SDR lead and AE lead who need to grade the same meeting the same way, and for the founder or partnerships lead writing an agency contract. Each criterion is defined by the evidence that proves it on both the account side and the attendee side. The standard then fixes the review windows, a closed list of eight reason codes, and a calibration test that two graders must pass before anyone is paid on it.
What counts as an accepted meeting?
An accepted meeting is a held meeting where the account passes three account criteria and the attendee passes three attendee criteria. Each one has to be proved by evidence attached to the CRM record before the grade is given. Every booked meeting ends with exactly one of three grades, and a grade with no evidence behind it does not count as a grade.
- Accept. All six criteria have evidence behind them. The meeting counts toward quota or the invoice once it is held.
- Remediate. The account passes, but the attendee evidence has a gap the booker can close before the meeting date. Remediate is only available before the meeting. If the gap is still open when the meeting starts, the grade becomes Reject with the matching code.
- Reject. The meeting fails, and the grader records one reason code from the closed list.
Credit is for held meetings, not booked ones
The Bridge Group labels its SDR quota "Stage 0 held", and its 2025 median is 10 held meetings a month. Cymate is blunter: "A no-show is a booked meeting that wasn't held, so it shouldn't count toward the number in your contract." This standard uses Skipcall's working definition of held: the prospect is "actually present for at least 15 minutes, status logged in the CRM within 24 hours". A rescheduled meeting counts once.
Define the booker by function, not by title
The policy should name what the booker is paid for, not what they are called, because the job title changes from team to team and market to market.
| Market | Total profiles | SDR:BDR naming (n=25 sample) |
|---|---|---|
| United States | 45,735 | 14:7 |
| United Kingdom | 5,793 | 10:10 |
| US/UK multiple | 7.9x (derived) | - |
In Refolk's index of professional profiles, US titles lean toward "Sales Development Representative", while the UK sample splits evenly with "Business Development Representative". If the policy says "SDR-booked meetings", a BDR team, or an agency rep with neither title, can argue it does not apply to them. Write "any meeting booked by a person or vendor paid wholly or partly on meetings" instead. One limit: a third index query, for US Account Executives under a software industry filter, returned 23 records. That count was an artefact of the filter, so I discarded it and do not publish an SDR to AE ratio from the index.
Which acceptance benchmarks can you trust?
Published figures put AE acceptance at roughly 50 to 70 percent. They only compare when the denominator matches, because acceptance of held meetings and acceptance of booked meetings are different numbers. Fix the denominator in the policy before you compare your team to anyone.
| Source | Acceptance figure | Denominator | No-show or show |
|---|---|---|---|
| Bridge Group 2025 | 6 of 10 = 60% (derived) | Held meetings | Not reported |
| Cleverly | 50%+ target | Meetings attended | 65-90% show, by team stage |
| Smarte | 60-70% | Unstated | 75-80% show |
| RepVue via WinsAbove | 5-6 of 9 = 56-67% (derived) | Qualified meetings booked | 22% no-show, enterprise |
| Bluebird Talent | Below 70% = problem | Unstated | Above 20% = problem |
The strongest source is the Bridge Group's 2025 report. It covers 351 B2B companies, 83% of them SaaS and 78% based in North America. Its median quota of 10 held meetings a month is down 40% since 2018, and 60% of SDRs reach it. The RepVue figure is quoted second-hand and has not been verified at the primary source. Quota also moves with the meeting model: 16 a month for introductory-meeting models, against about 9 to 10 for qualified models. A team that books intro calls will therefore show lower acceptance by design.
Skipcall claims that paying on booked meetings produces 35-50% no-show rates, but it gives no methodology. Treat that claim as a hypothesis to test against your own data, not a fact.
Account criteria: what proves ICP fit and no collision?
The account passes when its ultimate parent, matched by web domain, sits inside the ICP schedule, is not an existing customer, and has no open opportunity anywhere in the corporate family. The ultimate parent is the top company in a corporate family. Checking the parent rather than the account name is what catches a subsidiary that looks like a new logo.
| Criterion | Evidence that proves it | How it lies |
|---|---|---|
| A1 ICP fit | Booked entity's size and industry matched to signed ICP schedule | Parent fits, booked division does not |
| A2 Not a customer | Customer status checked on the ultimate parent | Subsidiary converted as a fresh account |
| A3 No open opportunity | Open-opportunity report run on the ultimate parent | Opportunity sits on a sibling or parent record |
Why the ultimate parent, and why a report. Traction Complete notes that a converted lead "becomes a fresh Account unless something matched it to the existing record first". The record in front of the grader can therefore look clean while the parent is a customer. Native Salesforce does not help: LeanData describes the Account Hierarchy view as "read-only and non-configurable", and native users cannot see open opportunities across the hierarchy. The grader runs a report on the ultimate parent. Looking at the account page is not enough.
What A1 proves, and what it does not. Nousu Collective names the failure "Right person, wrong entity": the contact fits, but works at a subsidiary or division outside the ICP. A1 therefore applies to the entity the meeting is actually with, not to the parent's headline headcount.
Ownership routing must be written down. Kubaru points out that a subsidiary lead "might go to the parent account owner, the subsidiary account owner, or follow your territory assignments." This standard does not pick one. The signed criteria must, because a collision with no written routing rule turns into a fight over credit.
The evidence package, outermost first
- Signed criteriaICP schedule, title schedule, routing rule, windows, reason codes, arbiter
- Account evidenceUltimate parent by domain, customer status, open-opportunity report
- Attendee evidenceDated employment capture, named authority source, dated trigger
- Meeting evidenceHeld for 15 minutes, logged within 24 hours, notes or recording link
Attendee criteria: what proves the right person, now?
The attendee passes when three things are on the record: a dated capture verifying current employment at the booked entity, a named source for their authority beyond their title, and a dated trigger that explains why they are taking the meeting now. All three are about the attendee, so a gap in any of them can be fixed before the meeting date.
| Criterion | Evidence that proves it | How it lies |
|---|---|---|
| B1 Current employment | Dated profile or company-page capture naming the booked entity | Stale profile, or employer changed after booking |
| B2 Authority | Named source: attendee's own words in notes, or a public record of their remit | Right title, no budget |
| B3 Dated trigger | Event with a date and a source link | Undated "interest", or a trigger at another entity |
B1 is a local agreement, not an industry norm. Contractable lists validation evidence as "CRM fields, meeting notes, call recording link". No public source treats a profile link or headcount record as accepted proof. This standard does accept a dated capture, and the contract has to say so explicitly. Capturing employment is the slow part of booking. Refolk cuts it to one plain-English question: ask for the attendee by name and employer and you get their current public profile back across LinkedIn, GitHub, and the open web, ready to attach with the date.
B2: a title alone never satisfies authority. Coldcallme describes the trap: "the title is right but the person has no budget or authority for what you sell." SaasHero's rejection list includes "Wrong title or no decision-making influence." The authority claim needs a named source. If the only evidence is the title, B2 fails.
B3: the trigger needs a date. Sparkle's rejection reasons are "no intent, wrong person, no need, or no timeline". A trigger proves timing only if it has a date and a source, and the signed criteria should set the maximum age the team will accept. A trigger lies when it is old, or when it happened at the parent rather than the booked entity.
Job changes between booking and meeting. No public guidance exists for this case, so this is my rule. Re-check B1 at the pre-meeting review and again the business day before the meeting. If the attendee has left, grade it Remediate and swap in a verified attendee at the booked account, or reject it as NOT-CURRENT. A move to another ICP account counts as a new booking and is graded from scratch.
Account evidence against attendee evidence
Why does one standard work for SDR comp and agency contracts?
It works for both because both disputes come down to the same question: who decides that a meeting counts, on what evidence, and by when. Answer it once, in writing, and the same text serves as the comp plan clause and the agency qualification clause.
The record owner wins by default. ORM Tech warns that the SDR loses disputes "by default because the account executive controls the record." Without a window and a default, every AE who doesn't get around to grading becomes an AE who rejects.
Paying on bookings can reward the wrong SDR. Revenue Best Practices gives a worked example.
The stakes are real: the Bridge Group's median SDR OTE is $80,000, of which $25,000 is variable pay.
Agencies write their own definition unless you write it first. Pay-per-appointment pricing runs from $30 to $1,700 a meeting depending on segment, and Virtustant notes that "the provider writes the definition of a qualified meeting". Abstrakt's model clause supplies two useful terms: a no-show is replaced at no charge, and either party may flag a meeting "within five business days, with a written reason." Two things are not established publicly: late-cancellation billing terms for B2B meetings, and profile links as proof. Write both into the contract yourself.
If you are choosing an agency, start with vendors that already bill on the same unit this standard credits.
Keep pipeline clawbacks separate. Apollo's sample plan reverses credit if an opportunity is disqualified within 30 days. That clause is about pipeline quality, not meeting acceptance. If you use it, keep it as a separate clause so the meeting grade stays final once its window closes.
Reason codes and review windows
Every Reject carries exactly one code from a closed list of eight, picked by a fixed order of precedence. Every grade also has a window, and once it closes, silence counts as acceptance. ORM Tech's requirement fits: "A closed list, entered in the CRM, reviewed monthly."
The closed list, in precedence order
- DUPLICATE: the same account and attendee have already been credited. A reschedule counts once.
- CUSTOMER: A2 fails. The ultimate parent is an existing customer.
- OPEN-OPP: A3 fails. An open opportunity exists in the corporate family.
- OUTSIDE-ICP: A1 fails, including right person, wrong entity.
- NOT-CURRENT: B1 fails. Employment is not verified at the booked entity.
- NO-AUTHORITY: B2 fails. Title only, or no decision-making influence.
- NO-TRIGGER: B3 fails. No dated trigger, need, or timeline.
- NOT-HELD: no-show, or the prospect was present for less than 15 minutes.
The grader goes down the list and records the first code that applies. Nobody has published a list this short that is proven mutually exclusive, and I do not claim these categories never overlap. The precedence order is what makes every grader resolve an overlap the same way. Codes 1 to 4 can only be Reject. Codes 5 to 7 can be Remediate before the meeting. Code 8 earns no credit, and an agency replaces the meeting at no charge. "Booked without pre-qualification", from SaasHero's list, is not a separate code: missing evidence before the meeting is Remediate, and at the meeting it takes the code of the first missing criterion.
What the published windows say
| Source | Window | Starts at | If AE is silent |
|---|---|---|---|
| SalesHive | 1 business day | Booking | Unstated |
| ORM Tech | 48h | Meeting | Meeting stands |
| Revenue Best Practices | 2 business days | Unstated | Accepted |
| Contractable clause | 5 business days | Unstated | Accepted |
| Abstrakt / Xynario agency | 5 business days | Unstated | Unstated |
The windows this standard sets
- Pre-meeting review: 1 business day from booking, and always before the meeting.
- Held status: logged within 24 hours.
- Post-meeting grade: 2 business days from the held meeting. A post-meeting Reject is only allowed as NOT-HELD, or with notes or a recording link from the meeting that contradict the evidence package.
- Dispute: RevOps arbitrates within 5 business days. If you prefer Skipcall's model, a neutral Head of Sales settles disputes within 48 hours.
- Agency flag: 5 business days with a written reason, so the internal 2-day window fits inside it.
1. Unit. A Qualified Meeting is a meeting held with a prospect present for at least 15 minutes, logged in the CRM within 24 hours. A rescheduled meeting counts once. A no-show is not a Qualified Meeting and is replaced at no charge. 2. Booker. This clause applies to any meeting booked by a person or vendor paid wholly or partly on meetings, whatever their title. 3. Account criteria, checked on the ultimate parent matched by web domain: (A1) the booked entity is inside the attached ICP schedule; (A2) the ultimate parent is not an existing customer; (A3) no open opportunity exists anywhere in the corporate family. 4. Attendee criteria: (B1) current employment at the booked entity, shown by a dated capture; (B2) authority shown by a named source, never by title alone; (B3) a trigger with a date and a source link. 5. Grades. Each meeting is graded Accept, Remediate, or Reject. Remediate applies only before the meeting date and only to B1, B2, or B3. Every Reject carries one code, chosen in this order: DUPLICATE, CUSTOMER, OPEN-OPP, OUTSIDE-ICP, NOT-CURRENT, NO-AUTHORITY, NO-TRIGGER, NOT-HELD. 6. Windows. Pre-meeting review within 1 business day of booking and before the meeting. Post-meeting grade within 2 business days of the held meeting. A meeting not graded within its window is deemed accepted. 7. Disputes. Either party may dispute a grade in writing. RevOps decides within 5 business days on the evidence attached to the record. 8. Versioning. Changes take effect only in a new signed version, and only for meetings booked after its date.
Paste into the SDR comp plan or the agency MSA unchanged, attaching your own ICP schedule, title schedule, and routing rule.
How to run the standard, step by step
The procedure runs from signed criteria to weekly review, with a named owner and a deadline at every step. Sources disagree on timing: SalesHive grades before the meeting, while ORM Tech and Scalelist grade after it, and Scalelist gives the AE "a fixed window, such as 2 business days, to reject a held meeting". This standard does both. The pre-meeting review catches problems the booker can still fix, and the post-meeting grade is limited to what only the meeting itself could reveal.
Booked-meeting acceptance procedure
- Write and sign the criteriaSales leader, SDR lead, and RevOps write the six criteria, eight codes, windows, and named arbiter, with ICP, title, and routing schedules attached. Done when the document is signed and versioned.
- Book with the evidence package attachedThe SDR or agency fills every required field, attaches account and attendee evidence, and sets the status to Pending AE review. Done when every criterion has a value and a link or note behind it.
- Run the pre-meeting reviewWithin 1 business day of booking and always before the meeting, the AE checks the ultimate parent, collisions, and the three attendee criteria. Done when the record shows Accept, Remediate, or Reject with one code.
- Log held or not held within 24 hoursThe AE logs the meeting as held only if the prospect was present for at least 15 minutes. A no-show gets NOT-HELD and earns no credit.
- Give the post-meeting gradeWithin 2 business days of a held meeting, the AE confirms the grade or rejects it, citing notes or a recording link. If the AE says nothing, the meeting is deemed accepted.
- Arbitrate disputesEither side disputes in writing, and RevOps decides within 5 business days using only the evidence on the record. Done when the final grade and a one-line reason are on the record.
- Review weekly and recalibrate quarterlyThe SDR manager or RevOps reports accepts and rejects by code plus window breaches every week, and reruns the 20-meeting calibration test every quarter. Done when the report has been sent and any changes have gone into a new version.
Booker (person or vendor): Booked entity and web domain: Ultimate parent (matched by domain): A1 ICP schedule row matched: A2 Customer status on ultimate parent (date checked): A3 Open-opportunity report on ultimate parent (date run, result): Attendee name and title: B1 Employment capture (link, date captured): B2 Authority source (who said it, where, date): B3 Trigger (event, date, source link): Pre-meeting grade / code / AE / date: Held (yes / no, minutes present, logged at): Post-meeting grade / code / evidence link / date: Dispute (raised by, date, arbiter decision, date):
Create these as required CRM fields so the record cannot reach Pending AE review with any of them empty.
How this standard goes wrong
Most failures pass a grader's quick look: the meeting appears to meet the standard, and the problem only surfaces in a dispute or an audit. Each failure below comes with the check that catches it.
- A subsidiary passes as a new logo. The converted lead became a fresh account and nothing linked it to the parent customer. Check: match on domain and ultimate parent, never on account name.
- An open-opportunity collision is missed. The opportunity sits on a parent or sibling record, and native hierarchy views do not show it. Check: run the open-opportunity report on the ultimate parent before grading. If the report has not been run, A3 has no evidence.
- Right title, wrong power. The attendee's title matches the schedule, but they have no budget or authority for what you sell. Check: require a named source for B2. A title is a hypothesis, not proof.
- The AE games the rejection. Skipcall warns that "the AE rejects hard meetings at the end of the month". Check: every quarter, compare each AE's rejection rate against their peers, and look at how rejections bunch up by date within the month.
- Calibration agreement looks better than it is. A sample that is mostly Accepts can hit 18 of 20 while the graders barely agree on the hard cases. Check: seed the sample with Remediate and Reject cases, and report kappa alongside raw agreement.
- The attendee changes jobs. The profile was current at booking and out of date by the meeting date. No public procedure covers this. Check: re-run the B1 capture the business day before the meeting. A fresh search in Refolk takes seconds and gives the reviewer a new dated link to attach.
- A no-show is counted as qualified. The meeting was booked, accepted before the date, and never happened. Check: credit only on the held status, which needs 15 minutes of presence logged within 24 hours.
An AE who can ignore the standard without consequence makes it optional. Silence has to count as acceptance.
How to run the calibration test and keep the standard current
Before go-live and every quarter after, two graders independently grade the same 20 booked meetings. The standard passes only if they agree on at least 18 and the reported kappa is at least 0.61 on a seeded sample. Raw agreement alone can pass a standard that two graders do not actually share.
What kappa measures. Cohen's kappa measures the agreement between two raters who each classify N items into C mutually exclusive categories, corrected for the agreement you would expect by chance. The Landis and Koch bands rate 0.61 to 0.80 as substantial and 0.81 to 1 as almost perfect. Landis and Koch supplied no evidence for those bands, so treat 0.61 as a working floor, not a proven threshold. CASRAI's advice applies: always report raw agreement and kappa together.
Why the sample must be seeded. Both cases below have 90% observed agreement:
| Grade split (both graders) | Chance agreement | Kappa |
|---|---|---|
| 14 Accept / 3 Remediate / 3 Reject | 0.535 | about 0.79 |
| 18 Accept / 1 Remediate / 1 Reject | 0.815 | about 0.46 |
The same prevalence problem appears in the general literature: at a fixed 85% raw agreement, kappa is 0.699 when the categories are evenly split but -0.081 when one category covers 85 of the items. Build each calibration sample with at least 6 meetings that the SDR lead expects to be Remediate or Reject, and include at least one collision case and one job-change case.
Why 20 is fragile. Each disagreement moves raw agreement by 5 points, and one study with 7 raters found that a single rating change moved its weighted statistic by roughly 0.13. No sales-specific sample size has been published. If a test passes or fails by one meeting, run a second sample before acting on the result. With a median of 1 SDR to 2.4 AEs, each SDR's meetings are graded by more than one AE, so rotate the AE in the grader pair each quarter.
Keeping it current.
- Review reason-code counts monthly and window breaches weekly.
- Compare AE rejection rates quarterly.
- Re-sign the criteria whenever the ICP schedule, title schedule, or routing rule changes, and apply each version only to meetings booked after its date.
Before the standard governs pay or invoices
- The criteria, ICP schedule, title schedule, and collision routing rule are signed and carry a version number.
- The CRM blocks Pending AE review until every booking field has a value and a dated link or note.
- The reason-code field is a required picklist that accepts only the eight codes.
- A report shows open opportunities and customer status rolled up to the ultimate parent.
- Window timers run automatically, and an ungraded meeting flips to accepted when its window closes.
- Held status requires minutes present and a logged-at timestamp.
- Two graders agreed on at least 18 of 20 seeded meetings, and kappa of at least 0.61 was reported alongside.
- The agency contract, if any, contains the same clause text as the comp plan, word for word.
Questions practitioners ask
What is a good SDR meeting acceptance rate?
Published figures cluster around 50 to 70 percent, but only compare figures that use the same denominator. The Bridge Group's 10 held meetings with 6 reaching Stage 1 gives a derived 60% of held meetings. Cleverly targets 50% or more of meetings attended. Smarte says 60 to 70 percent without naming a denominator. No primary source breaks acceptance out by segment, so measure your own held-meeting baseline before you set a target.
How long should an AE have to reject an SDR-booked meeting?
This standard gives the AE 1 business day from booking for the pre-meeting review and 2 business days after a held meeting for the final grade. After that, silence counts as acceptance. Published windows run from 24 hours to 5 business days. For agency contracts, Abstrakt's model clause gives either party 5 business days to flag a meeting with a written reason, so the shorter internal window fits inside it.
Should a no-show count as a qualified meeting in a pay-per-appointment contract?
No. A no-show is a booked meeting that was not held, so it should not count toward the contracted number. Abstrakt's model clause says a meeting the prospect does not attend is replaced at no charge, and Cymate says a rescheduled meeting counts once. No public source sets late-cancellation billing terms for B2B meetings, so write your own into the contract rather than assuming a default.
What happens if the attendee changes jobs between booking and the meeting?
No public guidance covers this, so the standard sets its own rule. Re-verify current employment at the pre-meeting review and again the business day before the meeting. If the attendee has left, the grade is Remediate: swap in a verified attendee at the booked account before the meeting date. If nobody is swapped in, the meeting is rejected as NOT-CURRENT. A move to another ICP account is a new booking and gets graded from scratch.
Why not just use BANT as the qualified meeting definition?
BANT labels say what to look for, not what evidence proves it, so two graders reading the same record will disagree. This standard defines each criterion by the evidence attached to the record, on both the account side and the attendee side. It also adds collision checks against the ultimate parent, a closed list of reason codes with a fixed order of precedence, and a calibration test. BANT covers none of these.
How many meetings do I need for a grader calibration test?
This standard uses 20, with at least 6 that should not be Accepts, and requires both 18 of 20 raw agreement and a reported kappa. No sales-specific sample size has been published. With 20 meetings, each disagreement moves raw agreement by 5 points, so treat a narrow pass or fail as a reason to run a second sample, not as a verdict.
Try it on the search you came here for
Stop building boolean strings. Just describe the person.
Type one sentence. I plan the search, read GitHub, public LinkedIn and Crunchbase records, and the open web as it is right now, and hand back a ranked list with the reason next to every name.
01Describe them
One plain sentence. Role, city, stack, stage, whatever matters to you.
02I read the web live
GitHub, public LinkedIn and Crunchbase records, the open web. Not a database that went stale last quarter.
03You read the shortlist
Ranked, with the reasoning under every name. Open a profile, ask a follow-up, narrow it down.
- Staff backend engineers in NYC who shipped Rust in production
- Series A fintechs in SF under 50 people, growing headcount this year
- Maintainers of fast-growing Rust web frameworks on GitHub
- No boolean, no filters, no seat to buy. One box.
- Read at search time, so a profile updated yesterday counts today.
- Every step visible as it runs, every name with its reason.
500 free credits on sign-up. No card, no demo call. See real searches.