Working Capital Dashboards: Weekly Metrics, Real Trade-offs

Most working capital dashboards don't fail because the numbers are wrong. They fail because nobody agreed on what "cash" means at 9am on a Monday. The finance lead at a ₹40-150 crore B2B company gets a board that reconciles perfectly on day one, then drifts: a bank feed that lags two days, three entities with different chart-of-accounts habits, and a sales head disputing an overdue figure that came from the CRM instead of the books. By week six, half the team is back in the spreadsheet they had before.

The cause is almost never the tooling. It's a set of trade-offs nobody puts in the scope document.

The trade-off between freshness and trust

Bank feeds, Tally or Zoho Books connectors and payment gateway APIs all promise "real-time". In practice, real-time cash and correct cash are not the same number. UPI and card settlements land T+1 or T+2 depending on your PSP. Cheques you've deposited sit uncleared for days. A NEFT initiated at 6pm may post tomorrow. Your accounting ledger, meanwhile, only knows about it when someone posts the entry.

So the honest design question: is this board a cash position or a cash proxy? If it's a proxy, say so on the screen — "as of 4:10 pm, bank feeds last synced 3:00 pm". A visible timestamp does more for adoption than another chart.

A practical rule: sync everything daily, but only review weekly. Daily sync keeps the underlying data honest. Weekly review stops the board from becoming a stock ticker nobody acts on.

What earns a place on the weekly board

Weekly suits anything you can still change before month-end. Six candidates:

• Cash and bank by account, net of uncleared items. One line per account, plus total available versus total book balance. They will differ, and that difference is useful information.
• Receivables ageing with names attached. Not only 0-30/31-60/61-90 buckets — the top 10 overdue invoices with the sales owner and last contact date. Section 15 of the MSMED Act requires payment to micro and small suppliers within 45 days, so track how many of your own invoices cross 45 days with larger buyers.
• Payables due in the next 14 days, including TDS, GST and advance tax, not just vendor bills. Discovering on the 8th that TDS wasn't on the board is an expensive way to learn this.
• Committed monthly outflows. Payroll, rent, EMIs, and annual SaaS renewals spread into a monthly figure. That's your floor.
• DSO and DPO, computed identically every week so the trend line means something.
• Bookings-to-cash lag — days from delivery or milestone to invoice raised. In services businesses the leak is usually here, not in collections.

Six candidates, but most teams should start with three.

Granularity is a maintenance cost

Every extra dimension is a mapping someone has to keep alive. Invoice-level ageing across 3,000 open invoices is manageable. Adding "sales rep" looks trivial until the CRM and the accounting system disagree on account ownership, and territories get reshuffled twice a year.

Take a ₹40 crore distributor wanting weekly invoice-level ageing by customer, by rep, by region. The first build is roughly two to three weeks of work. Keeping the rep mapping current is about two hours a week, every week — and the moment it slips, the sales head stops trusting the board and returns to his own file. The two hours is the real cost, not the build.

The rule: add a dimension only when someone will make a different decision because of it. Margin by customer changes pricing conversations. Margin by pincode rarely does. The same logic governs access — reps respond well to seeing their own overdue list and badly to seeing company cash. If customer personal data is on the board, DPDP Act obligations apply: collect what you need and restrict who can export it.

Weekly numbers and the monthly close are not the same thing

Trying to make a weekly dashboard match the statutory books exactly is a known dead end. Weekly numbers are provisional — accruals are estimates, one entity's GST gets posted late, a few invoices sit unapproved. That's fine, provided the board is labelled provisional and the month-end close stays the source of truth.

Build a small variance bridge instead. When the close lands, show the delta between the last weekly figure and the final one. After two or three cycles you'll know which lines to trust fully and which to read with a pinch of salt — and you'll stop having the same argument on the 5th of every month.

Where to start

Pick three numbers: available cash net of uncleared items, top 10 overdue receivables with owners named, and 14-day payable outflow including statutory dues. Put one person's name against each, review in a fixed 30-minute Monday slot, and give it six weeks before adding anything. If you want help pulling data from Tally, Zoho Books, bank statements and your CRM into a board your finance team will actually open on a Monday, that's the kind of work Maysan Labs does.