WHITE PAPER · OCTOBER 2026
The Cost of Answering Questions in a State VR Agency
What an agency spends to answer its own internal questions, built from cited federal wage data and labeled estimates, with three recommendations that do not require asking fewer questions.
THE BASE CASE
$359,000
per year to answer 37 questions a day from 300 users.
502 hours
of staff time per month, about 2.9 full-time positions
57%
of the cost is indirect and appears in no system
$38.54
per question, 38.75 minutes of agency time
PURPOSE AND SCOPE
What this paper measures
This paper estimates what a state vocational rehabilitation agency spends to answer its own internal questions. The model covers questions originating with VR counselors and escalating to supervisors, program managers, and policy managers. It does not cover questions from consumers, vendors, or federal partners.
Wage and compensation inputs are drawn from published federal data and cited. Time inputs are modeled. The distinction is marked throughout.
Basis of estimate
Each input carries one of three labels.
Sourced. Published data, cited in the Sources section with a link.
Modeled. An Outlook Insight estimate, with the reasoning stated.
Agency input. A figure that must come from the agency rather than from this model.
This convention is applied because most published analysis of interruption cost does not distinguish between the three, and the resulting figures cannot be verified. Two of the most frequently cited statistics in this subject area do not survive a check against their stated source. Both are addressed below.
EXECUTIVE SUMMARY
Most of the cost is invisible
A VR agency fielding 37 questions per working day spends approximately $29,900 per month, or $359,000 annually, to answer them. This represents roughly 502 hours per month of counselor, supervisor, and manager time, equivalent to 2.9 full-time positions.
Remediation efforts directed at response speed address only the direct portion and are capped at 43% of the total. The three recommendations in this paper target the indirect portion.
THE QUESTION PATH
One question, three paychecks, no record
A counselor meeting with a consumer needs to confirm whether a specific assistive technology purchase can be written into the IPE or must first be run through comparable benefits. The counselor is reasonably confident but will not commit the answer to a case file without confirmation.
01
Counselor searches
Policy manual, shared drive, prior email. No conclusive answer.
02
Counselor asks
Messages a supervisor who is mid-case-review.
03
Supervisor escalates
Consults state policy, remains uncertain, forwards to the program manager.
04
Answer returns
The following afternoon.
No portion of that sequence is recorded anywhere. The cost is distributed across three paychecks in increments small enough to escape notice, and the sequence repeats dozens of times per day.
LABOR COST BASIS
What an hour of agency time costs
| Role | Annual | Basis |
|---|---|---|
| VR Counselor | $60,920 | Sourced. BLS mean, Rehabilitation Counselors (SOC 21-1015), State Government excluding schools and hospitals [1] |
| VR Supervisor | $86,700 | Sourced. BLS mean, Social and Community Service Managers (SOC 11-9151), State Government excluding schools and hospitals [2] |
| Program Manager | $98,740 | Sourced. BLS 75th percentile, SOC 11-9151 [2] |
| Policy Manager | $110,000 | Modeled. Positioned between the BLS 75th percentile ($98,740) and 90th percentile ($127,550) for SOC 11-9151 [2] |
BLS does not separate state VR management into supervisor, program, and policy tiers. Three of the four figures are published means or percentiles. The fourth is placed within a published range.
Burden rate
BLS reports that wages and salaries account for 61.5% of total compensation for state and local government workers, with benefits accounting for the remaining 38.5% [3]. This yields a burden multiplier of 1.63.
General business literature commonly applies a multiplier of 1.30 to 1.45. That range reflects private-sector compensation structures and understates public-sector cost by approximately 20%. The difference is driven primarily by defined-benefit pension obligations, which have a limited private-sector equivalent. Agencies costing internal staff time at 1.35 are understating it.
$47.62
per hour for a VR counselor, or $0.79 per minute
$72.86
per hour for the blended responder, or $1.21 per minute
The responder blend assumes 60% of answers originate with supervisors, 25% with program managers, and 15% with policy managers. Modeled. Escalation patterns vary by agency and should be measured locally.
Annual salary is divided by 2,080 hours. Actual productive hours for state staff, net of holidays, annual leave, sick leave, and training, are closer to 1,840. Using 1,840 would increase every figure in this paper by approximately 13%. The more conservative denominator is applied throughout.
FINDINGS
What one question costs
Direct cost
| Activity | Minutes | Charged to | Cost |
|---|---|---|---|
| Counselor searches for the answer independently (policy manual, shared drive, prior email, case notes) | 5.0 | Counselor | $3.97 |
| Counselor composes the question | 2.0 | Counselor | $1.59 |
| Supervisor or manager reads, researches policy, and responds | 7.0 | Responder | $8.50 |
| Counselor reads the response and poses one clarifying question | 3.0 | Counselor | $2.38 |
| Subtotal | 17.0 | $16.44 |
Indirect cost
| Activity | Minutes | Charged to | Cost |
|---|---|---|---|
| Responder regains task context following interruption | 9.0 | Responder | $10.93 |
| Counselor blocked time, unable to advance the case | 8.0 | Counselor | $6.35 |
| Escalation to a second responder | 1.0 | Responder | $1.21 |
| Rework following an incorrect or incomplete answer | 3.75 | Both | $3.61 |
| Subtotal | 21.75 | $22.10 |
Per question: 38.75 minutes of agency time at a cost of $38.54.
All four direct durations are modeled. Search and research time exceeds what a general office model would allow. VR questions routinely touch Title I, WIOA, state policy, the fee schedule, and the case record itself. The source material is distributed across more systems, and retrieval time scales accordingly.
The first line item represents five minutes of unsuccessful search occurring before any question is asked. Where the answer existed and was not located, that time is a total loss. The closest published reference point is the McKinsey Global Institute estimate that interaction workers spend nearly 20% of the workweek locating internal information or identifying colleagues who can assist [4]. That figure is a share-of-week measure rather than a per-question measure and is not used as a model input. It is cited only to establish that the order of magnitude is reasonable.
TWO STATISTICS, CHECKED
Two numbers everyone cites, and why this model does not
The 23-minute recovery figure
The figure of 23 minutes and 15 seconds to recover from a workplace interruption appears throughout vendor material and trade press. It is typically attributed to Mark, Gudith, and Klocke, The Cost of Interrupted Work, CHI 2008.
That attribution is incorrect. The number does not appear in that paper, which reported a laboratory study of 48 students performing email tasks. The 23:15 figure originates in Gloria Mark’s field observation work and is stated in a 2006 Gallup interview describing observation of 36 information workers [5]. The associated peer-reviewed paper, No Task Left Behind? (Mark, Gonzalez, and Harris, CHI 2005), reports 25 minutes and 26 seconds with a standard deviation of nearly 55 minutes.
The measure also does not describe idle time. The interval includes handling the interruption and completing an average of two intervening tasks before returning to the original work.
The figure is therefore not used in this model. This model applies 15 minutes of degraded productivity to the 60% of questions assessed as genuine interruptions, averaging 9 minutes across all questions. Modeled, and materially below what the common citation would support.
The “9.3 hours per week searching” figure
This figure is attributed to McKinsey and does not appear in the McKinsey report cited as its source. The supportable statement in The social economy (McKinsey Global Institute, July 2012) is the approximately 20% of working hours referenced above, and that estimate traces through a footnote to IDC data rather than original McKinsey fieldwork [4].
The 20% figure is cited in this paper as context. No model input is built on it.
CONSERVATIVE ASSUMPTIONS APPLIED
Three assumptions set below their defensible maximum
Blocked time is limited to unrecoverable time
Response latency in most agencies exceeds one hour, but counselors redirect to other cases while waiting. Eight minutes of genuinely unrecoverable stall time is charged rather than the full wait interval. Modeled.
Duplicate questions are not given a separate line item
The working estimate is that approximately 30% of internal questions already have a documented answer. That estimate is modeled, not sourced, and is identified as such because it circulates in vendor material without attribution. Duplication is excluded from the cost tables in any case, because the associated minutes are already captured in the line items above. A separate row would double-count. The 30% is a remediation target, not an additional cost.
Rework is priced; its downstream consequence is not
When an incorrect answer reaches a case file, the exposure extends beyond the redo to case review findings, corrective action, and RSA monitoring items. This model charges 3.75 minutes of rework and nothing for compliance exposure, because that exposure cannot be priced reliably from available data.
RANGE OF ESTIMATE
The variable is not the questions
| Scenario | Per question | Per month | Per year |
|---|---|---|---|
| Low. Smaller agency, consolidated policy source, predominantly asynchronous questions | $22 | $17,100 | $205,000 |
| Base | $38.54 | $29,900 | $359,000 |
| High. Larger agency, senior staff responding, predominantly live interruption, distributed documentation | $62 | $48,200 | $578,000 |
The spread between the endpoints is approximately $373,000 annually against an identical question volume. The variable is not the questions. It is the agency’s capacity to answer them.
RECOMMENDATIONS
Three changes that target the indirect cost
R-1
Convert live interruptions to asynchronous channels
$4,200 per month
Reducing the interrupt share from 60% to 30% saves approximately $4,200 per month. Question volume is unaffected. This is the single highest-leverage change available and requires no technology investment.
R-2
Establish an authoritative, searchable source for recurring policy questions
$4,500 per month
Resolving half of the estimated 30% duplicate volume saves approximately $4,500 per month. As an external reference point, McKinsey estimates that a searchable record of institutional knowledge reduces information search time by as much as 35% [4].
R-3
Reduce the first-answer error rate
The highest-cost minute
Rework is the highest-cost minute in the model. The agency pays twice for the same answer, and in a VR context the second payment carries compliance exposure that this model does not attempt to quantify.
Taken together, R-1 through R-3 support a 30% to 40% reduction target without reducing the number of questions staff are permitted to ask.
LIMITATIONS
Internal questions are not waste
Internal questions transfer institutional knowledge, support counselor onboarding, and surface errors before those errors reach a case file. An agency recording zero internal questions would be in a worse position than one spending $29,900 per month on them.
This analysis addresses the avoidable portion of the cost, defined as duplicate questions and interrupt-driven delivery. It does not recommend reducing question volume, and it should not be presented as support for doing so.
APPLYING THIS MODEL
Four inputs, all from your agency
01
Volume
Sample one week of Teams channels, the policy inbox, and supervisor queues. Add an allowance for in-person and desk-side questions, which leave no record and are routinely undercounted.
02
Loaded rates
Actual salary bands multiplied by 1.63, calculated separately for counselors and for escalation recipients.
03
Interrupt share
The proportion arriving live versus batched. This is the largest single lever in the model.
04
Duplication rate
Review 50 recent questions and count those with an existing documented answer.
SOURCES
Where the numbers come from
- U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics: 21-1015 Rehabilitation Counselors, current release. Industry detail, State Government excluding Schools and Hospitals. bls.gov/oes/current/oes211015.htm
- U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics: 11-9151 Social and Community Service Managers, current release. Industry detail and national percentile estimates. bls.gov/oes/current/oes119151.htm
- U.S. Bureau of Labor Statistics, Employer Costs for Employee Compensation, March 2026 (USDL-26-0827, released June 12, 2026). State and local government total compensation $66.41 per hour worked; wages and salaries 61.5%, benefits 38.5%. bls.gov/news.release/archives/ecec_06122026.htm
- McKinsey Global Institute, The social economy: Unlocking value and productivity through social technologies, July 2012. Interaction workers spend nearly 20% of the workweek seeking internal information; a searchable knowledge record can reduce search time by up to 35%. mckinsey.com/featured-insights/mckinsey-digital/the-social-economy
- Gloria Mark, interviewed in Too Many Interruptions at Work, Gallup Business Journal, 2006. Source of the 23 minutes 15 seconds figure. The associated peer-reviewed paper is Mark, Gonzalez, and Harris, No Task Left Behind? Examining the Nature of Fragmented Work, Proceedings of ACM CHI 2005, reporting 25 minutes 26 seconds. news.gallup.com/businessjournal/23146/too-many-interruptions-work.aspx
Federal wage and compensation data accessed October 2026. BLS issues revised figures periodically; verify against the current release before citing these numbers externally. Modeled inputs are Outlook Insight estimates and are labeled throughout. Figures support planning and prioritization and are not intended for financial reporting.
ABOUT OUTLOOK INSIGHT
We built our work around the five-minute search
Outlook Insight, LLC has worked with state agencies since 1999 and vocational rehabilitation agencies since 2009. The company builds software for environments where locating the correct answer quickly is a compliance requirement and a service delivery obligation.
The focus of that work is the five-minute search identified at the top of this model: making policy, procedure, and institutional knowledge retrievable at the point of need, with the source attached, so that the search produces a citable answer rather than an escalation. That is what SixDot Knowledge Interpreter does.
Outlook Insight will review this model with agency staff using local salary bands and question volume on request.