Design SDR training around the work a representative must perform, then increase independence only when a manager can verify the required skills in realistic exercises or permitted customer interactions. A new rep might complete product and CRM practice in the foundation phase, for example, but move to independent outreach only after demonstrating accurate ICP selection, sound qualification, and reliable CRM notes.
A calendar date or activity count is not proof of readiness. HubSpot presents a 30/60/90-day structure as a way to organize foundation, skills, and performance training, but its meeting-count example is illustrative rather than a universal benchmark. This guide turns that outline into an operating model that separates learning progress, execution quality, and sales outcomes.
The practical decision is simple: define the capability required at each stage, collect traceable evidence, apply a consistent rubric, and record a manager-approved result before changing a rep’s phase status or quota ownership.
How do you design SDR training that shows whether a rep is ready?
Start by defining the tasks an SDR owns and the evidence that demonstrates competence. Teach and rehearse those tasks in stages, use a consistent rubric, and record a pass, revise, or incomplete result before increasing independence.
Track three distinct things: learning activity, such as completed role-plays; execution quality, such as ICP fit and CRM completeness; and sales outcomes, such as held meetings or accepted opportunities. Activity can show effort, but it cannot establish conversation quality or pipeline value by itself.
A calendar marks elapsed time; observable evidence demonstrates readiness.
Define readiness before choosing training activities
SDR and BDR responsibilities vary by company. One organization may assign inbound qualification to SDRs and outbound prospecting to BDRs, while another may use the titles differently. Document role boundaries, handoff expectations, territory assumptions, and systems before building the curriculum.
For every competency, specify the expected behavior, evidence source, reviewer, and pass-or-revise criteria. Core competencies commonly include:
- Offer and ICP: explain approved product claims, identify the intended buyer, and distinguish a good-fit prospect from a poor-fit one.
- Prospecting and research: build an appropriate list and connect relevant account or persona information to an outreach reason.
- CRM practice: find and update the correct records, follow data conventions, and record an actionable next step.
- Conversation skills: open a relevant conversation, ask useful questions, respond to an objection, and request a clear next step.
- Qualification and handoff: apply the approved framework and document the evidence supporting a handoff.
Use current product claims, persona definitions, qualification criteria, and messaging in exercises. A mock pitch based on outdated positioning can produce a passing score for the wrong skill.
Define funnel events separately in the CRM or training system. A meeting booked is scheduled; a meeting held took place; a qualified opportunity meets stated qualification criteria; and pipeline accepted means the receiving sales team accepted it under the documented process. Assign each event a system-of-record source and date rule. A booked-meeting count is not a proxy for qualification quality.
Build ramp gates around a 30/60/90-day learning sequence
Use the 30/60/90-day structure as a planning sequence, then set locally appropriate review dates and thresholds. For each phase, document the outcome, practice, evidence, reviewer, threshold, and next action. Timing may change with role complexity, territory, prior experience, and manager capacity.
Foundation: learn the offer, buyer, and tools
Teach product and ICP knowledge, CRM use, prospect-list building, approved messaging, and the qualification approach. Use tool walkthroughs, shadowing, list-building exercises, messaging workshops, and supervised mock calls. The gate is demonstrated accuracy: the rep can explain the offer using approved claims, identify fit, use the CRM correctly, and complete a supervised customer-facing task to the agreed standard.
Skills: practice guided execution
Move into guided live outreach, multichannel execution, qualification practice, and regular call or exercise review. Use a versioned rubric with observable criteria such as relevance, question quality, response to an objection, qualification evidence, and next-step clarity. Require a consistent result across multiple examples, not one strong call. Coach the observed gap and reassess with new evidence.
Performance: test sustained quality and independence
Increase independence when the rep consistently meets the role’s quality standards across an explicitly defined measurement window. Review pipeline quality, self-monitoring, CRM hygiene, and handoff outcomes alongside activity and quota progress. Discuss quota ownership or a career path using evidence and approved targets. Neither should become automatic on day 61 or day 90. Keep manager support responsive to observed needs rather than reducing check-ins solely because a date has passed.
If leaders set a meeting or activity target, work backward only after defining the target and funnel semantics. Start with the approved qualified-meeting goal, apply the organization’s observed qualification and held-meeting rates, then estimate earlier conversations using observed conversion data. Label the source cohort, measurement window, target version, and assumptions. Recalculate when the data changes. Do not adopt an example such as 15 qualified meetings per month as a universal quota.
Make practice and call review traceable
Use one rubric version for a defined assessment period. Record the exercise or call reference, reviewer, observation, rubric version, review date, and final result. A manager reviewing objection handling should cite the relevant call or role-play and describe the observable response instead of recording only “good call.” Keep the manager’s interpretation separate from any AI-generated summary or suggested category.
HubSpot documents review features for eligible call recordings and transcripts, including associated call information. Transcription and analysis depend on the calling source, account configuration, permissions, and qualifying seat or subscription. Before recording or sharing calls, confirm applicable consent requirements and restrict access to people who need it. HubSpot’s call recording and transcript review documentation and call-recording consent guidance describe these product and privacy considerations. The guidance is not legal advice.
A practical review chain is: permitted recording or mock exercise, verify the source and association, optionally draft a summary or coaching category, check the cited evidence against the rubric, then save the approved observation with its source reference. If consent is uncertain, the transcript is missing, or the call is associated with the wrong record, route it to a manager or designated privacy owner instead of scoring it.
Track training events separately from dashboard metrics
Model durable representative status, individual practice events, and calculated performance summaries as different data. A phase-status record answers “where is this rep now?” An attempt record answers “what happened in this exercise?” A dashboard metric answers “how did the defined group or period perform?” Combining those grains makes duplicate attempts difficult to detect and scores difficult to audit.
The following is an illustrative design, not a HubSpot standard schema. The event identifier is unique for one attempt, while the business key includes the attempt number and observed timestamp so repeated attempts are not collapsed into one row.
{
"training_event_id": "evt-illustrative-0042",
"sdr_record_id": "123456",
"ramp_phase": "skills",
"exercise_type": "objection_roleplay",
"attempt_number": 2,
"observed_at": "2026-10-13T14:30:00Z",
"rubric_version": "v1.2",
"source_record_id": "mock-789",
"reviewer_id": "manager-45",
"result": "revise",
"reviewed_at": "2026-10-13T15:10:00Z"
}Define one row as one attempt at one exercise. Keep call observations at call level, activity facts at rep, activity type, and reporting-date level, and quota summaries at rep, period, target-version, and measurement-window level. If transcript citations are stored, keep each citation or timestamped passage as a separate observation linked to its call and review.
Give each event a stable identifier. If an integration or orchestration layer may process events concurrently, enforce uniqueness in the database or use a transactional upsert where supported. A lookup followed by create is not race-safe because two workers can both find no record and then create duplicates. When the source has no event ID, define a business key that matches the intended grain, such as the SDR, exercise, attempt number, and observed event time. Replays should reuse the same identity; a genuine new attempt should receive a new identity.
For review observations, use a separate identity from the underlying attempt when multiple reviewers or multiple cited passages are valid. A proposed observation key might include the source system, source record ID, rubric version, reviewer, and citation start time. Treat this as an implementation recommendation, not a vendor-defined key.
HubSpot dashboards can centralize configured reports and control access, but they are views over available data, not a complete standardized SDR ramp system. HubSpot documents a limit of 100 values for a report’s “Break down by” property in dashboards, which is another reason to aggregate rather than display every raw event. See dashboard management and dashboard customization. For organizations reviewing where operational records should live, CRM systems and data structure is a relevant resource.
Use the dashboard for the answer and the training ledger for the evidence. A monthly completion rate can be a useful summary, but it must remain traceable to attempt-level records, rubric versions, reviewers, and source references.
Use rules for readiness decisions and AI for bounded assistance
Use deterministic rules for checks with explicit answers: required evidence present or missing, correct measurement window, valid consent state, correct SDR assignment, allowed rubric value, and existing source record. These checks are easier to explain and audit than asking a model to infer whether a rep is ready.
AI can draft a coaching note, summarize a call, suggest an objection category, or highlight possible qualification evidence from supplied material. Require source references, validate required fields and allowed values, and have a manager review the result before any formal readiness, quota, promotion, or performance-management status changes. A model suggestion is a prompt for review, not proof of competency.
HubSpot documents Prospecting Agent functions that include account research, email generation from research, contact enrollment, and outreach strategies. Availability and use depend on product edition, credits, permissions, and feature status. Buying Signals and Contact Sourcing may be identified as beta functionality in the documentation. The documentation does not establish a performance uplift, so confirm the live account configuration before making a feature a training dependency. See Prospecting Agent documentation.
| Trigger and input | Task | Validation and destination | Human fallback |
|---|---|---|---|
| Phase review with exercise, rep, rubric, and source | Rules check required evidence; AI may draft coaching notes. | Validate rubric version and source; save the attempt and phase status separately. | Manager resolves missing or contradictory evidence and approves the gate. |
| Permitted recording or mock call with call ID | Optional summary or suggested coaching category. | Check consent, permissions, association, and cited passage; save a call-level observation. | Manager or privacy owner handles uncertain consent, missing transcript, or disputed interpretation. |
| Approved target and observed funnel rates | Deterministic quota-backsolving calculation. | Validate definitions, cohort, period, and target version; save assumptions with the plan. | Sales operations checks the calculation and leadership approves target changes. |
For administrative reminders, HubSpot workflows support configured enrollment triggers and re-enrollment. Because re-enrollment can run actions again, treat it as repeat execution. Add a completion marker or stable event-key safeguard before sending notices or writing records. See workflow creation and re-enrollment behavior.
Review the ramp system and improve the next cohort
Review learning progress, execution quality, funnel outcomes, and activity as separate metric groups. If reps complete exercises but struggle with accepted opportunities, inspect qualification and handoff evidence. If activity is high but meetings are not held, examine targeting, messaging, and booking quality. If CRM records are incomplete, address the workflow or system behavior instead of assuming more outreach practice will solve it.
Compare cohorts only when role, territory, ICP, target version, and measurement windows are sufficiently comparable. A before-and-after change can show where to investigate, but it does not prove that a training change caused the result. Preserve the rubric and target versions used for each cohort.
HubSpot’s Sales Training Plan is a gated downloadable framework. Its public page describes onboarding, training, technical setup, communication, 100-day goals, development planning, and feedback, but it does not expose the complete editable asset for inspection. Treat it as a starting resource, not evidence of a particular CRM schema, scoring rubric, or automated readiness process. View the Sales Training Plan resource.
- Booked, held, qualified, and accepted events have separate definitions, sources, and date rules.
- Each gate has observable evidence, a threshold, a rubric version, and a named reviewer.
- Call review has a consent and access check, with a route for uncertain cases.
- Attempts, observations, and citations have identities that distinguish repeats and multiple passages.
- Concurrent writes use a database-enforced unique key or atomic upsert where supported.
- Dashboards show calculated summaries while source evidence remains traceable.
- Cohort comparisons preserve role, territory, target, rubric, and measurement-window context.
At a defined review interval, sales leadership should decide whether to adjust instruction, rubric, target assumptions, or manager capacity. For organizations implementing the record model and operational review around HubSpot, HubSpot systems support may be relevant. The ramp should keep improving, but readiness remains tied to documented capability, not elapsed days alone.