Skip to content
Compliance Automation Tools

Home / Guides / Measure compliance work in hours per control, per audit cycle and per questionnaire, with the same clock each time

Guide, read 13 Nov 2026

Measure compliance work in hours per control, per audit cycle and per questionnaire, with the same clock each time

Framework Coverage MatrixAll guides

Time three units and keep the definitions fixed

Hours saved is the number every vendor quotes and few teams measure. To measure it yourself, pick three units. Hours per control per cycle: the time to collect, check and record one control's evidence. Hours per audit cycle: everything from kickoff to the final report, split by phase. Hours per questionnaire: from receipt to the sent answer. Fix the start and stop rule for each, write it on one page, and do not change it mid-year.

Count person-hours, not elapsed days. A questionnaire that sits for six days and takes 90 minutes of work is a 1.5 hour item with a long wait. Both numbers matter, but they are different numbers. Track elapsed time as well and label it as elapsed.

What to time and how

Use a simple time log for four weeks before any tool change. Ask each control owner to log minutes in a shared sheet with the control ID, the activity type and the date. Sampling works if logging every task is too heavy: log two weeks per quarter and scale. State the method with the number, for example 'based on 2 weeks of logging in each of 4 quarters'.

Split each unit into activities so you see where hours go, since a tool may move one activity and not others. The usual split is find the evidence, get the evidence, check it, record it, chase the owner, and explain it to the auditor.

UnitStart and stopLog by activityCount it with
Control testRequest issued to result recordedFind, pull, check, record, chaseControl ID in the log
Audit cycleKickoff to final reportScoping, evidence, sampling, questions, fixesAuditor request list timestamps
QuestionnaireReceived to sentTriage, answer, review, approvalTicket created and closed times
Monitoring alertAlert raised to closedTriage, fix, retestTicket times, see the [remediation](/glossary/remediation/) term
Policy reviewReview opened to signedEdit, route, signDocument version dates

Control count sets the size of the problem

Hours per control only make sense next to the count and the cadence. SP 800-137 says a single control can need different frequencies for different parts. Its example is AC-2 Account Management, with ten requirements (a through j) and four enhancements in the SP 800-53 text it cites. So ask how you counted. A team that counts 200 controls and a team that counts 200 control-parts are measuring different things.

A rough test of your own load is controls times tests per year times minutes per test. A control tested quarterly for 45 minutes costs 3 hours a year. That is the figure to put beside a vendor's claim. The current catalog counts differently again: in NIST's OSCAL release 5.2.0, AC-2 has parts a through l and 13 enhancements. Fix your counting unit before you compare to anyone.

Then split the load by who does it. Of the 3 hours a year in the quarterly example, the owner may spend 2 finding and pulling and the compliance lead 1 checking and filing. A tool that automates the pull moves 2 hours and leaves the check. If your log shows the check is the larger share, automation of collection will not give the saving you expect, and you should know that before you buy.

Questionnaire hours depend on how much you can reuse

For questionnaires, the single most useful split is new answers versus reused answers. Tag each question when answered: written fresh, adapted from a prior answer, or pasted unchanged. Log minutes per question in each class. A tool that helps with reuse will show up as a falling share of fresh answers, long before total hours change.

Also record the question count and the format, since a 400-question spreadsheet and a 40-question web form are not the same size of job. Hours per question, and not hours per questionnaire, is the number that survives comparison. Report the median and the longest, because one outlier can double an average. See the security questionnaire term for scope.

Audit cycle hours: split by phase and by who

For an audit cycle, record hours per phase for your team and separately for others, such as engineering and legal. Engineering hours are the ones most often missed, since they are not logged to compliance. Ask for them in a short form at close: which tickets were raised for the audit, and how long each took.

Keep the retention duty in the count. HIPAA 45 CFR 164.316(b)(2)(i) requires 6 years retention of required documentation, and 164.308(a)(8) requires a periodic evaluation, so cycles do not end when the report is issued. Time the filing and retrieval as well. A platform that reduces collection hours but leaves retrieval unchanged shows up in this split, and not in a single total.

Reading vendor hour claims against your own

This guide quotes no vendor hour figures, because none were fetched for this edit. When you see one, label it as the vendor's, as in 'Vendor X says.', and ask four things. What was the baseline? Which unit? How many customers? Over what period? Without those, it is a marketing figure.

NIST's own OSCAL site says OSCAL, through automation, dramatically reduces audit durations from months to minutes. That is NIST's statement about the data format and not a measured result for a tool. Treat both the same way. Run your four-week log, switch one framework or one questionnaire stream first, and compare the same unit after one cycle. CA-7 in SP 800-53 gives a model for the reporting side: it asks for metrics to be defined, frequencies to be set, and status to be reported to named recipients. Hours per control are a metric of that kind, so give the number an owner and a reporting date. Then use the method page and the platform matrix to see what we compare, and continuous monitoring for the tests that cut collection time.

Where this shows up in the index

1 of 16 tools score 6 or more on pricing transparency, which means a price figure is published on the vendor's own pages; the rest ask for a quote, so hours saved cannot be priced from public pages.

Sources

  1. nvlpubs.nist.gov/nistpubs/Legacy/SP/nistspecialpublication800-137.pdf, read 12 Nov 2026
  2. raw.githubusercontent.com/usnistgov/oscal-content/main/nist.gov/SP800-53/rev5/json/NIST_SP-800-53_rev5_catalog.json, read 12 Nov 2026
  3. ecfr.gov/api/versioner/v1/full/2026-11-01/title-45.xml?part=164&subpart=C, read 12 Nov 2026
  4. pages.nist.gov/OSCAL, read 12 Nov 2026

Last reviewed 13 Nov 2026.