The cost of sales rep admin is the loaded labor cost of observed administrative work plus the cost of correcting errors and redoing failed handoffs. It is not automatically equal to lost revenue. Build the business case from a local time audit, explicit formulas, and a conservative view of what reclaimed capacity can become.
Direct answer: annual direct admin cost = admin hours per rep per week × active reps × active weeks × loaded hourly cost. Add measured rework separately. Show low, base, and high cases, then test the workflow that produces the cost before buying software.
External surveys help establish why the question matters, but they do not answer it for your team. Salesforce's 2026 State of Sales surveyed 4,050 sales professionals in 22 countries. Its disclosed sample is broad context, not your task-level ledger. Use the sales rep time audit for deeper observation guidance.
Define sales admin before pricing it
Write a mutually exclusive task taxonomy. Without one, every team labels inconvenient work “admin” and inflates the business case. Separate direct buyer work, preparation and strategy, administration, internal coordination, learning, and unavailable time.
| Bucket | Examples | Count as admin? |
|---|---|---|
| Buyer interaction | discovery, demo, negotiation, useful follow-up | No |
| Decision preparation | account evidence, call plan, deal strategy | Not automatically |
| Record keeping | CRM notes, activity logging, field updates | Yes |
| Handoff work | copying transcript, reformatting notes, duplicate entry | Yes |
| Internal coordination | forecast meeting, approvals, specialist requests | Classify by purpose |
| Correction | fixing wrong records, duplicate tasks, stale fields | Yes, as rework |
Call prep illustrates the boundary. Reading a relevant buyer email and choosing discovery questions may advance the sale. Copying the same email into three tools is admin. A CRM update that creates useful shared state is necessary administration; re-entering a field because sync failed is avoidable rework.
Define the unit and period: minutes per task, tasks per day, per rep, over two representative weeks. Exclude holiday, onboarding, quarter-end, or outage periods unless they are the subject of the study. Preserve segment—AE versus SDR, new versus tenured, SMB versus enterprise—because averages can hide different workflows.
Run a two-week time audit
Use direct observation or prompted self-logging with clear definitions. Passive telemetry can count app activity but often cannot distinguish useful research from duplicate work. Self-reports capture purpose but introduce recall error. A practical audit triangulates task logs, system timestamps, and a small observed sample.
- Freeze the taxonomy. Give reps examples and a rule for ambiguous tasks.
- Sample representative days. Include meeting-heavy and prospecting-heavy work.
- Log start, end, task, system, record, and outcome. Avoid rounding everything to 15 minutes.
- Handle concurrent time once. A meeting plus note-taking cannot create 90 minutes inside a 45-minute interval.
- Tag rework and errors. Record why the task repeated and which handoff failed.
- Calibrate reviewers. Have two people classify a sample and adjudicate differences.
Microsoft's 2023 Work Trend Index surveyed 31,000 people and analyzed aggregated Microsoft 365 signals; it reported broad pressure from communication and information search. Treat that as knowledge-work context. Your audit must isolate actual sales tasks.
Report median and distribution, not only mean. A small group of reps with complex territories or broken permissions can dominate the average. Record task frequency and time separately; a rare 40-minute task may cost less than a two-minute task repeated 30 times.
Calculate labor and rework cost
Use loaded labor cost, not salary alone. Include employer taxes, benefits, and other labor costs according to the finance team's standard. Do not add software or manager cost twice if the model treats them separately.
Direct annual admin cost = H × R × W × C, where H is observed admin hours per rep per week, R is active reps, W is active weeks per year, and C is loaded hourly labor cost.
Worked example: a 12-rep team observes a median 6 admin hours per rep each week. Finance uses 46 active weeks and a loaded hourly cost of $62.
6 × 12 × 46 × $62 = $205,344 per year.
Now add measured rework. Suppose the team logs 30 CRM corrections per week, with a median eight minutes of rep or operations time per correction:
30 × 8 ÷ 60 × 46 × $62 = $14,260 per year.
The base observed cost is $219,604. That is an estimate under stated assumptions, not a promise that software will save the full amount. Build low and high cases from observed confidence intervals—for example, 5, 6, and 7 hours—rather than arbitrary percentages.
Salesforce's older fifth-edition survey summary reported that reps spent 28% of their week selling across a survey of more than 7,700 professionals. Do not classify the other 72% as removable admin. It includes many legitimate activities and a different sample and period.
Separate capacity from revenue
Three values are commonly confused:
- Labor cost: what the observed admin time costs.
- Reclaimable capacity: time a tested workflow actually removes after review and exceptions.
- Cash or revenue impact: spending avoided or additional contribution generated.
Reclaimed capacity becomes cash savings only if staffing, contractor spend, overtime, or hiring plans change. It becomes additional selling time only if the team has enough qualified work and the recovered slots are deliberately reassigned. It becomes revenue only through a separate funnel model with response, meeting, opportunity, win, timing, and gross-margin assumptions.
Use a conservative realization factor derived from a pilot. If a workflow removes two measured hours but reps reallocate one hour to buyer work, the realized selling capacity is one hour. Do not use 100% without evidence.
The sales productivity ROI guide covers business-case discipline. Present annual cost, addressable share, pilot reduction, review time added, exceptions, software and implementation cost, and realized allocation separately.
Prioritize the workflow to fix
Rank task families by annual cost, avoidability, error severity, and feasibility. Start with high-frequency, rule-bounded work where source and destination are clear. Avoid automating ambiguous commercial judgment merely because it consumes time.
| Candidate | Good automation fit | Required control |
|---|---|---|
| Post-call note draft | source transcript and fixed schema | identity, claim support, rep review |
| Follow-up task draft | explicit buyer commitment | owner and due-date confirmation |
| Activity logging | stable event and destination | idempotency and reconciliation |
| Stage suggestion | documented criteria | protected field and approval |
| Deal strategy | decision support only | human ownership and uncertainty |
Run a matched pilot. Freeze task definitions, compare before and after for the same workflow, and measure admin minutes, review minutes, correction minutes, critical errors, adoption, and failure handling. A faster workflow that creates wrong CRM state is not an improvement.
Keep tool subscription, implementation, administration, training, and exception-handling costs outside the labor numerator and add them once in the business case. Record whether the intervention shifts work from reps to RevOps or managers; transferred labor is not eliminated cost.
The admin-reduction guide provides broader implementation steps. Keep the local audit as the baseline for renewal and rollback decisions.
Where Gangly fits
Gangly's Workflow Sequencer connects signal, outreach draft, call prep, live coaching, post-call notes, and CRM suggestions. The rep reviews outbound messages and CRM changes. Gangly is a workflow layer, not the CRM.
Use the calculator to identify the relevant step, then pilot that step. For example, test Post-Call Notes on a labeled set and measure draft review time plus corrections against the current manual note. Do not claim the survey average or this worked example as your savings; report observed local results and limitations.