Know who has room before you promise the work
FigsFlow is capacity planning for accounting firms — every job, task and deadline your team already works from, totalled by person and by week. An assignment shows what it costs someone before you make it, and a week that will not fit is visible while there is still time to move it.
Partner or owner, start with the capacity board in the tour below. Manager or scheduler, go straight to where the plan stops matching the week.
The six-week average is 78%, which looks fine. The two weeks that matter are both over.
- Capacity 37.5h · contracted, less time off
- Assigned 44.5h · across 9 open tasks
- Gap 7 hours over · flagged Monday
- Cause Two 31 December year ends landed on the same due date
- Action One moved to Dan · 6.0h · reviewer unchanged
the week got fixed on Monday, not explained on Friday
Add work to a week and watch it tip
This is one senior with 37.5 hours in the week of 21 September, and 22 already committed. Every job an accounting firm takes on is a weight going onto one side of a scale nobody usually gets to see. Put a couple of returns on it and watch what a partner's reasonable-sounding promise actually costs.
15.5 hours of headroom. This is the only moment where moving the work is cheap.
Nobody decided to overload her. Four systems each said yes on their own.
It is the third week of September. A partner takes a call, agrees to a new set of statutory accounts, and tells the client it will be ready before the deadline. That promise is completely reasonable — and nobody in the conversation has any way of knowing whether it is true.
The job list says what is due. It does not say how long any of it takes. The capacity spreadsheet says who is free, but it was built in April and three engagements have changed since. The timesheets know exactly what happened, and they know it four weeks late. So the promise gets made from memory, which is the only system available in the moment.
On Friday the senior working that return is nine hours down and has stopped replying to Teams. The overload was not a decision anyone made — it was assembled from four separate yeses, each defensible on its own, none of them able to see the other three.
The same work, counted once — as hours somebody has to find
This is what separates capacity planning inside a practice management system from a resourcing tool bolted on beside it. The job, the budgeted hours, the assignment, the logged time and the firm's utilisation are five views of one record — so the plan cannot drift away from the work, because it is the work. A capacity tool that has to be updated by hand is a spreadsheet with a login screen.
A signed engagement creates the job and its task list. Nobody opens a planner to tell it the work is coming — it is already there.
Tasks come off a template with budgeted hours already on them, so a job has a size the moment it exists rather than after someone estimates it.
Hours total by person and by week against real availability. Overload shows before the assignment is confirmed, not after the deadline moves.
A timer runs on the same task that was planned, so budget and actual are two columns of one row instead of two systems to compare.
Utilisation, recoverability and where the season actually bit come out of the records that captured them. No month-end rebuild.
- Budgeted hours on tasks — the size of the work comes from the template that built it, not from a separate estimate
- Deadlines and review steps — a week is loaded by when work is due, including the reviewer's share of it
- Working patterns and time off — contracted hours per person, less approved leave, so availability is not assumed to be five full days
- Logged time to date — what is already spent on a job adjusts what is left, mid-week rather than next month
- The assignment decision — you see the cost of a job to someone's week before you hand it to them
- Deadline risk — a due date with no room in front of it is a warning, not a surprise
- Utilisation reporting — planned against actual by person, team and service, from the same records
- Hiring and pricing conversations — whether the firm is short of people or short of price, backed by hours the system captured itself
The plan and the week that actually happened
Planning the work is one side. Doing the work is the other. They are supposed to meet on the same task — and in most firms they don't, so somebody clears the difference by hand every Monday. What you are clearing depends on the size of your firm, and so does how wide the gap is.
Judge capacity planning software on that gap rather than on a feature list. The question is not how many views a planner has — it is whether the hours it plans with are the same hours the work is measured in, and whether the plan updates itself when a client sends the records three weeks late. A resource planner that has to be maintained by hand is a spreadsheet that costs more.
One task. The budgeted hours, the assignment, the logged time and the utilisation number are the same object — so the plan cannot go stale without the work going stale with it.
Two systems. The gap is where scope changes, late records and unbudgeted chasing live — and somebody rebuilds the plan on Monday morning to bridge it.
A plan built on job counts assumes every set of limited company accounts is the same size. They are not, and the firm knows they are not — the knowledge just lives with whoever did it last year. Without budgeted hours on the task itself, a capacity plan is a headcount divided by a deadline, and it will be wrong in the direction that hurts.
Leave, study days, a compressed four-day week and the four hours a week that go on nothing chargeable. These are the difference between a plan that works and a plan that assumes everyone is at a desk for five full days. Most planning happens against a theoretical week nobody in the firm has ever worked.
When a week does not fit, something moves — and in most firms that move leaves no trace. By June nobody can say whether the season was hard because the firm is short of people, short of price, or short of a client who sends their records on time. Those are three completely different decisions, and the data to tell them apart was never captured.
Every size of firm, and the week that sends each one looking
Capacity planning stops being optional at the point where one person can no longer hold the next six weeks in their head. That point arrives at a different headcount for every practice — and the first thing a firm should switch on is different every time too.
Same choice as the comparison above — pick once, both sections follow
Short list, honestly labelled
Most of what capacity needs is already inside FigsFlow — the jobs, the hours and the time. The connections that matter are the ones that tell it when somebody is not at their desk. Everything else on this page is what is not connected yet, because finding that out in month two is worse than reading it now.
This is the connection that does the work, and it is not an integration — it is the same database. Budgeted hours arrive on the task when the template builds it, so the plan is populated by work being created rather than by anybody maintaining a planner.
- BUDGETED HOURS FROM THE TEMPLATE
- DEADLINES AND REVIEW STEPS
- OWNER AND REVIEWER
Logged time lands on the same task the plan was built from, which is what makes planned-against-actual a column rather than a project. A job running long adjusts what is left in the week while there is still a week left.
- PLANNED VS ACTUAL PER TASK
- REMAINING HOURS UPDATED LIVE
- UTILISATION FROM ONE SOURCE
Sign in with the identity your firm already administers, and see the calendar commitments that are competing with the week you are planning. Booked meetings are visible beside the board rather than discovered when somebody misses an afternoon of work.
- ENTRA ID SSO
- CALENDAR VISIBILITY
- GRAPH
Your existing capacity plan imports as a starting point, and everything the board holds exports in full. If you leave, you leave with the data — and if a partner wants to model something the reports do not cover, the export is the answer rather than a support ticket.
- CSV / EXCEL IMPORT
- FULL EXPORT BACK OUT
- NO LOCK-IN ON YOUR OWN HOURS
No sync with BrightHR, BambooHR, Breathe or the rest. Leave is requested and approved inside FigsFlow, which means a firm running an HR system will be approving it in two places until this lands.
Phase 2 — not committed to a dateThe board shows an overloaded week, the cause, and what each fix would cost someone else. It does not solve the schedule and hand you a finished plan. A person decides, every time.
Deliberate, not a gap we are closingSuggested owners rank on who has room and who has done that service before. There is no competency model behind it, and no rule that stops work being assigned to someone who is not qualified for it.
Not in this releaseNo integration with the tax and accounts production tools a UK firm files from. Capacity plans the work being done; it does not read progress out of your tax software.
Not in this release