Use CDM's preregistered time-diary protocol to measure creator coordination, rights, reviews and reporting before publishing workload benchmarks.
Short answer: answer: CDM cannot give a defensible hours-per-week figure because no prospective time diaries have been collected. This protocol measures active labor by task, role, creator and campaign for 28 consecutive days, then reports distributions only after consent, validation and methods review. Fees alone are not the cost base; coordination, rights and review must be observed before they can be priced.
Creator-budget calculators commonly add creator fees, production and media while treating management as a percentage. That shortcut hides whether a small team spends its scarce hours on relationship work, preventable rework, rights administration or reporting.
CDM’s position is that the hidden operating cost of creator programs sits in coordination, rights and review—not only creator fees—but that claim needs measured time, not a confident estimate. This article preregisters the Creator Operations Time Diary. It contains no CDM results, participant count or observed average. The sample record is illustrative and must never enter analysis.
The Creator Operations Time Diary
The spine has four layers: Event → Activity → Workload unit → Capacity decision. A participant records a work event close to when it occurs; the event receives one primary activity code; hours are normalized to active creators or campaigns; only then may a team model capacity.
The primary outcome is paid or owner labor minutes spent managing creator partnerships per participant-week. Secondary outcomes are hours per active creator-week, hours per campaign, work share by activity, unplanned-work share and rework share. “Waiting for a creator” is not labor time unless a person actively follows up, monitors or resolves the delay. Elapsed campaign time belongs in the production-delay study, not this diary.
Who and what will be sampled
Define a small team as one to ten people who regularly perform marketing work, including founders when marketing is part of their role. Eligible teams must manage at least one compensated, gifted or affiliate creator relationship during the observation month and nominate all staff whose creator work is in scope. Exclude creators’ own production time unless the creator is also part of the participating marketing team.
Recruit through public invitations, professional groups and direct outreach; disclose every channel. Seek variation in country, industry, program maturity, relationship model and agency use. Enroll teams prospectively and include all creator-management activity during 28 consecutive days. Do not invite only teams running a flagship launch or omit a quiet week.
The design target is 50 teams. It is not an achieved sample. Publication as a benchmark requires at least 20 independent teams, 500 valid participant-days, representation from at least three team-size bands, and no organization contributing more than 15% of recorded hours. If those conditions fail, CDM may publish a clearly labelled feasibility report without a market workload number.
Full diary instrument and task taxonomy
Every event has required structured fields:
| Group | Fields |
|---|---|
| keys | participant pseudonym, team pseudonym, event ID, campaign ID, creator pseudonym |
| time | local date, start, stop, duration minutes, UTC offset, contemporaneous or recalled flag |
| activity | primary code, optional secondary code, planned/unplanned, routine/exception, rework yes/no |
| context | active-creator band, campaign stage, organic/paid/affiliate use, synchronous/asynchronous, internal/external |
| role | founder, marketer, creator manager, operations, legal/compliance, analyst, agency or other |
| outcome | completed, handed off, awaiting input, cancelled or incident opened |
| quality | overlap flag, break excluded, source prompt, missingness reason and reviewer status |
The activity codebook is:
- strategy and creator discovery;
- outreach, vetting and selection;
- negotiation, contracting and onboarding;
- briefing and production coordination;
- content review, feedback and revision;
- disclosure, claim and specialist review;
- rights, licence, expiry and asset administration;
- publishing, whitelisting and paid amplification;
- measurement, reconciliation and reporting;
- relationship care and creator support;
- incident, correction and takedown;
- program administration, tools and internal meetings.
Use the activity that best describes the decision being advanced. If a 30-minute call covers contracting and creative, split the event when practicable; otherwise code the dominant activity and mark mixed. Free text is optional, tightly limited and reviewed for names or sensitive information before transfer.
Consent, burden and data protection
Give every participant a plain-language statement covering purpose, fields, expected daily burden, voluntary participation, withdrawal, retention, publication and contact route. Team leaders must agree that individual diaries will not be used for performance management. Where worker monitoring or human-subject rules apply, obtain privacy, employment and ethics review before recruitment.
Prompt participants twice daily and allow same-day correction. Do not install keystroke, screen or location monitoring. The Bureau of Labor Statistics’ American Time Use Survey shows the value of structured diaries, coding lexicons and start/stop activity records, but CDM’s occupational panel is not nationally representative and should not borrow that status.
Store identity keys separately. Publish team-level aggregates only after disclosure testing; suppress small role/geography cells. Pseudonymization reduces exposure but may not make information anonymous, particularly when public campaigns and unusual incidents are recognizable.
Verification and diary quality
Run daily automated checks for negative or excessive durations, overlapping events, missing codes and impossible dates. Participants resolve flags without sending message content. At week end, ask one coverage question: “Was there creator-partnership work you did not record?” Record the estimated omission band as a quality variable, not as retroactively invented events.
A second trained coder independently recodes a random 10% of de-identified event descriptions. Report agreement and adjudication after collection; no reliability result exists today. Compare total diary minutes with participants’ optional calendar totals only as a coverage check, not as proof of task category. Calendar meetings omit asynchronous work and include meetings unrelated to creators.
Exclude sleep, breaks, general social-media consumption and creator production by external creators. Keep zero-activity participant-days: they are essential to a weekly workload denominator. Analyze recalled events separately from contemporaneous entries.
Prespecified analysis and workload calculator
For each team, calculate total creator-operations hours, median participant-week hours, hours per active creator-week and hours per live campaign. Summarize teams with medians, interquartile ranges and selected percentiles. Give each team equal weight in the primary benchmark; show person-event results only as secondary detail.
Report activity shares, unplanned share and rework share with denominators. Stratify descriptively by team-size band, number of active creators, paid-use status and program maturity only when disclosure cells are large enough. Use team-clustered uncertainty intervals if sample size permits. Treat missing days using registered complete-day and sensitivity analyses; do not silently scale partial diaries to full weeks.
The calculator will accept a team’s own active creators, campaigns and local loaded hourly cost. It may apply future study percentiles to show a range. CDM will not convert diary hours to dollars without user-supplied cost assumptions, and it will label all projected values as scenarios rather than study observations.
Illustrative record and calculation—not a finding
Participant P-S04 logs 09:10–09:40, activity 7 “rights and licence administration,” planned, asynchronous, outcome “handed off.” The participant checked two asset-expiry dates and updated the takedown owner for synthetic creator CR-X12. Duration is 30 minutes.
If an invented team recorded 12 such half-hour events in a week, the arithmetic would be six hours. At a user-entered loaded cost of $50 per hour, the calculator would display $300 for that activity that week. Those values demonstrate the instrument and calculation only; they are not a CDM average, quote or recommendation.
Stopping and publication rules
Pause collection for coercion, unexpectedly sensitive notes, identity-key exposure, excessive diary burden or a material instrument defect. Record amendments and seek ethics/privacy/methods review. Stop at the preregistered date; do not extend the diary because an anticipated busy period failed to occur.
Publish results only after minimum coverage, data lock, independent recoding, disclosure-risk review, reproducible analysis and external methods review. Release the frozen protocol, instrument, codebook, synthetic dataset and aggregate tables. Publish inconvenient or null patterns. If quality or diversity is insufficient, publish the recruitment and missingness outcome without calling it an industry time benchmark.
Related guides
Frequently asked questions
How many hours do small teams spend on creator partnerships?
CDM does not have a defensible figure yet. No prospective diary dataset or verified participant sample has been supplied, so an average would be invented. The study will calculate team-weighted distributions after 28-day observation and quality review, including zero-activity days. Teams can use their own diaries now, but should label them local baselines.
The caveat is that workload varies with creator count, paid usage, program maturity and incidents. Any future number must carry its denominator and context rather than becoming a universal staffing ratio.
Why use a diary instead of asking for a monthly estimate?
A prospective diary captures short, fragmented tasks closer to when they occur and can assign them to a consistent taxonomy. A single retrospective estimate is more likely to miss routine messages, rights checks and context switching or to overemphasize unusual incidents. Start and stop fields also make overlaps and impossible totals testable.
The caveat is that diaries create burden and can change behavior. Keep entries short, measure missed coverage, separate recalled events and report participation patterns so apparent precision does not hide recording error.
Does waiting for a creator count as management time?
Only active labor counts in this study. Time spent sending a follow-up, checking a dependency or changing a plan is recorded; passive calendar time while no one works is not. That separation keeps labor capacity distinct from campaign lead time. Record each return to the task if waiting fragments the workday.
The caveat is that on-call attention and interrupted work can impose a real cost. Participants should record the actual monitoring or recovery event and mark it unplanned, while elapsed waiting belongs in the companion workflow-delay dataset.
Will the diary be used to evaluate individual employees?
No; the protocol prohibits performance-management use and is designed for aggregate workload research. Participant identities are separated from analysis keys, individual results are not returned to managers, and small cells are suppressed. Consent must be voluntary under applicable employment and research rules. Withdrawal and complaint routes must bypass the participant’s direct manager where practical.
The caveat is that unusual roles or campaigns can still be recognizable. A privacy reviewer must test disclosure risk, and a team should not participate if it cannot protect staff from adverse use.
How will part-time founders and agencies be handled?
Record actual creator-related events for every in-scope role, then report role bands and team totals. A founder’s creator work counts even if marketing is not the founder’s main job; agency labor counts only when the participating organization has consent and adequate visibility. Normalize hours to participant-days rather than pretending every contributor is full-time.
The caveat is incomplete boundaries: undisclosed subcontractor or creator-side work will make the observed total partial. The final report must state whose labor is included and avoid calling brand-side hours the full economic cost.
When can the workload calculator use real benchmark values?
Only after the registered minimum sample, valid-day threshold, quality checks, data lock, disclosure review and independent methods review are complete. Until then, the calculator should accept only a user’s own hours or clearly labelled illustrative inputs. Its interface must display the source period, sample and uncertainty beside outputs.
The caveat is that even a completed benchmark will describe a sample, not prescribe staffing. Teams should select a relevant percentile and adjust for rights, review intensity and campaign mix rather than treating the median as an efficiency target.
Next decision: How Do You Decide What a Small Marketing Team Should Automate First?
Related reading: How to Write a Creator Brief That Protects Creative Freedom and Brand Outcomes · How to Build a Creator Campaign Dashboard Executives Will Trust · Original Research as Authority Infrastructure: From Question to Citable Asset
Sources and research notes
- US Bureau of Labor Statistics: ATUS Methods Overview — checked 26 September 2026; official documentation for time diaries, coding lexicons, questionnaires and data construction.
- US Bureau of Labor Statistics: ATUS Data Sources — checked 26 September 2026; describes diary collection, activity coding and independent coding/adjudication controls.
- Center for Open Science — checked 26 September 2026; primary source for time-stamped preregistration and linked methods records.
- HHS Office for Human Research Protections: Belmont Report — checked 26 September 2026; ethical principles concerning consent, benefit/harm and participant selection.
- ICO Anonymisation Guidance — checked 26 September 2026; regulator guidance on identifiability, pseudonymisation and governance.
Limitations: CDM has collected no diaries and claims no workload estimate, participant count, activity share or cost benchmark. Design thresholds and the calculation are illustrative protocol choices, not results. Self-selection, diary fatigue, omitted activity, role boundaries, jurisdiction and a four-week window will constrain inference. Replace this protocol with results only after ethical collection and review, while retaining the preregistered record.
This article is editorial guidance. Apply the principles in proportion to your market, evidence, and responsibilities.
