Financial analytics · Singapore
Smartdatafinance builds and runs the reporting, forecasting and analytics layer for companies that have outgrown spreadsheets but do not need a full in-house finance analytics team. Monthly packs that arrive on time, forecasts that reconcile, dashboards people actually open.
Practice
We work with owner-managed groups and mid-market companies, typically between SGD 5 million and SGD 150 million in turnover, including multi-entity operators based in Singapore and trading across the region. Finance teams in this band already close the books and file what must be filed. What they lack is a numbers layer that is timely enough, and structured enough, to steer the business between board meetings.
The starting problem is usually familiar. The monthly close takes three weeks, so the pack arrives after the next month is already in motion. The cash forecast lives in one person’s workbook and cannot be reconciled to the bank. Margin by product, customer or outlet is guessed from a pivot that nobody quite trusts. Pricing decisions rest on last year’s list plus a conversation.
Smartdatafinance puts a working layer in place: a close calendar and a management pack that lands in the first ten days, a 13-week cash view rebuilt from bank and receivables data, models that hold together when a driver is changed, and dashboards that read from a cleaned chart of accounts. We stay to run the cycle, or we hand it to your team with the documentation they need.
Services
Reporting, forecasting, commercial analytics and data work are specified separately so you can see the catalogue clearly. In practice they share a chart of accounts, a close calendar and a review rhythm. The six services below are the ones most engagements start with.
Monthly packs built around the decisions your board and operators actually take: flash P&L, cash, working capital and a short commentary. Designed to land in the first ten days after month-end, with a stable pack structure from one period to the next.
Read moreA 13-week cash view rebuilt each week from bank balances, receivables ageing and known payments. Receipts are dated from collection history rather than invoice dates, so the forecast can be compared with what actually cleared.
Read moreAn annual budget that can be loaded into the same structure as the monthly pack, then a quarterly reforecast that updates volume, price, labour and overhead without rebuilding the model from a blank sheet each time.
Read moreThree-statement models and scenario work for expansion, a new outlet, a pricing change or a refinance. Drivers sit in one place, statements articulate, and the outputs can be dropped into a board paper without a late-night rebuild.
Read morePower BI or Looker Studio views fed from a warehouse or a disciplined extract, so the same figure appears in the pack, the forecast and the screen. Pipelines are documented so your team can refresh them after handover.
Read moreContribution and fully loaded margin by product, customer, outlet and channel, with the allocation rules written down. The point is to make the choices visible, so a debate about rent or delivery cost is a debate about a method, not a feeling.
Read moreEngagement
Most work follows the same sequence. The diagnostic is short and written. The build is time-boxed. The run phase is where the calendar becomes a habit. You then keep us on a retainer or take the file in-house.
We interview finance and the people who use the numbers, list every source system and spreadsheet, and walk a recent close. You receive a findings note, a data map and a prioritised build plan with an indicative fee. Nothing is built in this phase beyond samples needed to test a method.
Access is arranged through your systems. We test whether the chart of accounts can support the pack you want, whether entities can be consolidated without manual journals that disappear, and whether bank and subledger extracts can be scheduled. Gaps are written into the plan before the model is locked.
Pack templates, cash models, dashboards and allocation files are built against the agreed scope. Each artefact has a review cycle with your finance lead. We do not declare a build finished until a live month has been run through it and the figures have been reconciled to the ledger.
We operate the calendar with you: flash numbers, pack assembly, the weekly cash call and a short commentary. Issues that only appear in a live close are fixed in the file. This is also when operators are shown how to read the new views so the pack is used, not merely filed.
You receive a handover file: data map, pack calendar, model notes, refresh instructions and a list of known limitations. Some clients take the layer in-house at this point. Others keep us as a retained analytics partner to run the cycle each month and to handle reforecasts.
Sectors
Outlet contribution, labour as a percentage of sales, waste, delivery mix and the rent line that decides whether a site is earning its keep.
Load profitability, lane margin, detention and demurrage leakage, and a cash view that respects the lag between shipment and collection.
Utilisation, realisation, work-in-progress ageing and partner-level contribution, so pricing conversations rest on recovered hours rather than hope.
Gross retention, net retention, cost-to-serve by cohort and a cash forecast that treats annual invoices as a collection profile, not a single receipt.
Gross margin after freight and returns, inventory days by category, and a production-cost view that can be reconciled to the ledger.
Contract contribution, recoverable versus non-recoverable cost, and a working-capital view across deposits, sinking funds and billed arrears.
Work
The examples below are illustrative composites of typical engagements, not named client accounts.
The monthly pack arrived on day 19. Outlet managers saw a P&L once a quarter. Labour cost was known in total and guessed by site. The cash forecast was a twelve-month sheet updated when someone remembered.
A close calendar with a day-6 flash and a day-8 pack, outlet contribution after labour, waste and delivery commission, and a 13-week cash view rebuilt weekly from bank and receivables extracts. Rent and shared kitchen cost were allocated on a written rule, not a late journal.
Close moved from day 19 to day 7. Two sites that had looked acceptable on a fully loaded P&L were shown to be below contribution after labour and delivery. The weekly cash call replaced an ad-hoc request to the bookkeeper when a supplier asked to be paid early.
Jobs were profitable in the operations system and thin in the ledger. Lane margin could not be shown without a three-day extract. Cash was managed from the bank login because the 13-week view used invoice dates for receipts that routinely cleared 20 days later.
A job-to-ledger bridge, lane and customer contribution after agent cost and detention, and a cash forecast that applied collection curves by customer group. Multi-currency entities were translated on a documented rate policy so the pack and the cash view used the same method.
The 13-week cash view was rebuilt weekly from bank and AR data and began to match the following week’s cleared funds within a range the FD would sign. Two lanes that had been defended on volume were repriced after three months of contribution reporting.
Utilisation was reported from the time system. Realisation was a conversation. Work-in-progress older than 90 days sat in a spreadsheet that was not tied to the ledger. The board pack was a set of charts pasted from last month’s file.
A monthly pack that joined time, billing and the ledger: utilisation, realisation, WIP ageing and partner contribution. A short pricing appendix showed discounting by client tier. The dashboard refreshed from the same extract as the pack, so the two views could not drift.
WIP older than 90 days was reviewed on a fixed calendar instead of when cash tightened. Two client relationships that looked large on fees were shown to be below target realisation after write-offs. The pack landed on day 8 for four consecutive months.
Next step
A two-week diagnostic produces a findings note, a data map and a build plan you can take to your board. It is scoped work with a written fee, not an open-ended conversation.
If the plan is useful, we can start the build. If it is not, you keep the note.