Small-business outsourcing research

How long does month-end close take with a virtual assistant?

A protocol for measuring close duration by task, dependency, and rework instead of a single number at the end of the month.

How long does month-end close take with a virtual assistant?

Research question and scope

This protocol asks how long the month-end close takes when a virtual assistant supports the accounting work. The unit is one close cycle for one business, measured from the first close task to the last approved adjustment. The study separates task time, waiting time, and rework. It does not judge the assistant's accounting skill or the accuracy of the financial statements.

Set the scope to one business, one accounting system, and one close checklist. Record the tasks in scope, the approvers, and the cut-off for source documents. If the checklist or the system changes during the window, keep the earlier definition for the full sample or analyze the change as a separate period. A new approval step can look like a slower assistant when it is only a different control.

Define the population as every close cycle completed during the observation period, including cycles that were closed late or reopened. The unit is the cycle, not the task, because tasks overlap and a delay in one can block several others. A cycle that is reopened after a material correction is a new unit linked to the original. Test the definition on two recent close records before collection.

Definitions that make the measure reproducible

Define the close start as the first recorded task after the period cut-off, not the date someone first mentioned close. Define task complete as the point when the work and its evidence are attached to the checklist. Define waiting as time when a task is blocked by a missing document, a decision, or an approval. Define closed as the date the owner approves the final reconciliation. Each definition needs a positive and a boundary example.

Do not count a task that was started but not finished as complete. Do not count a review comment as an approval unless the record shows the owner accepted the result. A task that is reopened keeps both the first completion and the final completion date. Record both so the rework is visible.

Record the task category, the owner, the system, and the dependency for each task. These strata matter because a bank reconciliation and a payroll accrual have different inputs and approvers. Include tasks that were added mid-close, because an unplanned task is part of the cycle. Use one category list for every close.

Evidence standard and data dictionary

For each task, capture the start time, the completion time, the waiting intervals, the owner, the dependency, and the evidence link. Add the number of review rounds and whether the task was reopened. Keep the checklist export and the supporting documents together. A second reviewer should be able to rebuild the critical path from the record.

Also capture the date the source documents were available. A close that waits for a supplier statement is not the same as a close that waits for an internal approval. Record the source and the date. That field separates an external dependency from an internal one.

If the business uses a shared inbox for documents, record when the document arrived rather than when it was filed. A document that sat unread for three days is an internal delay. The protocol should not attribute that wait to the supplier. Separate the arrival date and the filing date in the data dictionary.

Sampling and collection

Draw the sample from the close checklist and the accounting system rather than from a list of difficult months. Record the export date, the filters, the row count, and the selection rule. Include every close cycle in the window so a single bad month does not dominate. If the sample is small, describe the cycles instead of computing a rate.

Collect the checklist as it existed at close, not after a later cleanup. If a note was added later, record the addition date and analyze the original wait separately. Do not infer a completion from a journal entry that was posted without evidence. Note the collection date separately from the close date.

A useful additional measure is the number of tasks that were not on the original checklist. An unplanned task extends the close and often reveals a missing template. Record the task, the reason, and the date it was added. That view helps the owner improve the checklist rather than asking the assistant to work faster.

Analysis and reporting

Report the number of close cycles, the median duration, the median task time, the median waiting time, and the median rework time. Show the counts by task category and dependency. Present the share of cycles that missed the target date and the tasks on the critical path. Use the same denominator for every rate.

Inspect the longest cycles and describe which dependency dominated without naming the employee. A wait for a supplier statement may point to a reminder process, while a wait for owner approval may point to a review calendar. Compare the result with the task list and the staffing. Report reopened cycles separately because they carry a different cost.

A useful additional view is the time from the last task completion to owner approval. A close that is technically finished but waits days for sign-off still delays reporting. Show that interval beside the task time. The owner can then decide whether to schedule a review slot or delegate a subset of approvals.

Limitations and sources of error

The record cannot show whether the financial statements were accurate or whether the assistant made a good judgment on an accrual. It can show only the task sequence and the recorded waits. A fast close is not proof of quality, and a slow one is not proof of poor work. The protocol also cannot capture work done outside the checklist.

Accounting systems can timestamp a posting at the end of the day or merge a correction into the original entry. A task may appear complete before the evidence was reviewed. Record those cases and their treatment instead of dropping them. If the same person prepares and approves a task, note the independence limitation and analyze that task separately.

A useful check is whether a new assistant can follow the close checklist and name the owner for each task. Ask them to walk through the first five tasks and record where they hesitate. That test measures the checklist, not the person. Fix the unclear owner before adding another control.

Review cadence and next action

Give the result to the owner and the accounting lead and ask them to choose one observable change. The change might be a document reminder, a pre-close checklist, a named backup approver, or a smaller review batch. Record the owner, the effective date, and the measure that will show whether the change worked. A change without a measure becomes an opinion at the next review. Avoid changing the checklist and the software at the same time.

Repeat the protocol with the same definitions after the change and compare the same fields. If task time fell but waiting time did not, the bottleneck may be a dependency rather than the assistant. If waiting time fell but rework rose, the faster path may be skipping evidence. Close the review when the owner accepts the finding and the next cycle is scheduled.

A useful additional view is the share of close tasks that were ready before the period cut-off. A task prepared early is less likely to delay the cycle, while a task that starts after the cut-off adds pressure to the review. Record the preparation date and the owner. That view helps the team move work earlier without adding a new approval.

A useful additional view is the number of tasks that were completed before the source document arrived. A task marked complete without evidence may be a placeholder, not a finished item. Record the evidence date beside the completion date. That view prevents a fast close from looking better than it is.

Ask the owner which approval took the longest and why. The answer may point to a decision that could be delegated with a written rule. Record the decision and the amount band. A small delegation can shorten the close without removing oversight from the items that need it.

Sources

Frequently asked questions

Does a favorable sample prove the routine caused the result?

No. The design is observational and supports a bounded operating decision, not a universal causal claim.

What should an owner review first?

Start with high-consequence defects, missing denominators, exceptions, and cases that required the owner to reconstruct the evidence.

Related research