Coverage is a map, not a ratio

Abstract

Headcount ratios are easy to quote and poor at explaining coverage across multiple sites and time zones. A leadership team needs a map of where work arrives, when it needs a response, and which absences break the schedule. This review uses public occupational data as context and sets a bounded local method for sizing administrative coverage. It also explains how to test the model against a real month before changing headcount [1][2][3].

Methodology

The review treats public occupational descriptions and wage data as role-design context, not as a staffing formula. The local method maps each site, principal, and coverage window, then separates planned work from reactive work. It models planned leave and peak periods, and it tests the result against a representative month. Results are reported as coverage gaps and options rather than a single headcount target [1][2][4].

The question and why headcount ratios fail

A common shortcut is to size support coverage with a ratio, such as one assistant for every two executives. The ratio is easy to quote and hard to defend because it ignores the work, the locations, and the time zones. Two offices with the same headcount can have very different coordination loads [1][2]. A better question is what coverage the office needs across sites and hours, and what capacity each role provides. That question requires a map rather than a ratio. The map can still use public data as context, but it starts with local demand [3][4].

Ratios also hide the difference between coverage and capacity. One person may be present but fully occupied, while another has slack. The ratio counts people, not available hours. A coverage map should show both the presence and the usable capacity. Capacity should be measured in available hours after planned work, not in headcount alone. The map should be simple enough to update each quarter.

What public occupational data can show

The Bureau of Labor Statistics describes the occupation and reports employment and wages by area. O*NET describes the tasks and requirements, including coordination and information management. Those sources help define the role and compare it with the local market [1][2]. They do not report how many assistants a multi-site leadership team needs. That number depends on the principals' schedules, the number of sites, and the coverage model. Public data can inform the role design but not the headcount decision [2][3].

Use the public data to describe the role and the local market, not to justify a headcount. If leadership asks for a benchmark, show the source and explain why it does not answer the coverage question. That honesty prevents a false precision. If the public figure is used at all, label it as context rather than a target.

Map coverage by site and time window

List each site, the principals based there, and the hours when coordination is needed. Include the overlap windows with other sites and with clients. Mark the periods that currently have no coverage or depend on one person [1][3]. The map will show whether the demand is concentrated or spread across the day. A site with a short, intense window may need a different schedule from one with steady requests. Coverage should follow the demand pattern rather than a uniform shift [2][4].

Include the informal coverage that already exists, such as a site lead who answers questions when the desk is closed. That support is real capacity, but it may be invisible and unsustainable. Record it and decide whether to formalize it. Hidden coverage is a risk, not a plan. Formalizing it may be cheaper than adding a new role, but it needs a defined scope and backup.

Separate demand from presence

Presence is the number of hours someone is at a desk, while demand is the work that arrives. A site can have full presence and still miss deadlines because the work clusters at the end of the day. Track the arrival pattern and the response time by window [1][2]. Separate planned work from reactive work. A desk that spends most of its time on reactive requests has little capacity for board preparation or travel planning. The mix affects how many people are needed, not just when they are present [3][4].

Measure the arrival of requests by hour and by day, not only by total volume. A desk can look busy overall while missing a critical two-hour window. The pattern should drive the shift design. A total without a distribution is hard to act on. A weekly view can reveal a pattern that a monthly total hides.

Model absences and peak periods

A coverage model that assumes everyone is present every day will fail in the first holiday season. Build in planned leave, sick days, and training. Estimate the peak periods, such as board cycles, budget season, and major travel [2][4]. Test the model against a real month rather than an average. Count the requests, the hours, and the days when the desk was short. The gap between the model and the month is the useful information [1][3].

Model the peak period with the actual calendar, including board meetings, travel, and holiday closures. A generic busy season will miss the specific weeks that matter. Ask the principals which periods are immovable. Those dates anchor the model. They also reveal where a deputy is needed even if the total headcount looks adequate.

Test the model against a real month

Choose a representative month and reconstruct the coverage day by day. Note the days when a site had no support, when a request waited, and when someone worked late. Compare those events with the model's assumptions [1][4]. Adjust one variable at a time, such as a shift start or a shared coverage window. Do not redesign the whole model from a single month. The goal is a realistic schedule that the team can sustain [2][3].

Compare the model with the desk's own experience. Ask where the schedule felt tight and where it had slack. The people doing the work often know which window breaks first. Use their observation to refine the assumptions. Their answers can expose a peak that the calendar alone does not show.

What to report to leadership

Report the coverage map, the demand pattern, the absence assumptions, and the gaps. Add the public occupational context with its source and date. Recommend one change, such as a shared coverage window or a deputy at each site [1][2]. Give the change a cost, an owner, and a review date. Do not present the model as a precise staffing formula. It is a way to make the trade-offs visible [3][4].

Show the options rather than a single answer. A shared coverage window, a deputy at each site, and a temporary increase during peaks have different costs and risks. Leadership can then choose the trade-off. A model that offers one number invites an argument about the number. The comparison should include the cost of each option.

Limits and responsible use

The model cannot capture every informal coordination task or the value of local knowledge. It depends on records that may be incomplete and on assumptions that may change. It also cannot resolve employment or legal questions about schedules [1][3]. Review the model when the sites, the principals, or the service scope change. Keep the assumptions visible so a later reader can judge the result. Use it to support judgement, not to replace it [2][4].

Document the assumptions in a short appendix so the model can be updated rather than rebuilt. Record the date, the sites, the principals, and the peak periods. A later review can then see what changed. The model is a living planning tool. Update it when a site opens, closes, or changes its leadership team.

Key findings

  • Headcount ratios hide differences in demand, sites, time zones, and principal support loads [1][2].
  • Public occupational data helps define the role but does not set multi-site headcount [1][3].
  • Map coverage by site and time window before deciding how many people are needed [1][4].
  • Model planned leave and peak periods, then test the model against a real month [2][4].
  • Report gaps and one recommended change rather than a single staffing formula [1][3].

Sources

  1. Occupational Outlook Handbook: Secretaries and Administrative Assistants, U.S. Bureau of Labor Statistics
  2. Occupational Employment and Wage Statistics, U.S. Bureau of Labor Statistics
  3. O*NET OnLine Summary Report 43-6011.00, O*NET OnLine
  4. Job Openings and Labor Turnover Survey, U.S. Bureau of Labor Statistics

Return to the Research Content library