# CFO Brief — DataRite Consulting Ltd

**Prepared:** 28 August 2026 · **Currency:** GBP
**Statement period:** 1 March 2025 – 27 August 2026 (545 days)
**Forecast horizon:** week commencing 31 August 2026 → 29 November 2026

---

## 1. Where the cash stands today

Closing bank balance is **£2,001,620**. Over the statement period the business
was **cash generative**, adding an average of **£41,977 a month**.

That last point matters for how you read the rest of this brief: there is no
burn rate, so **cash runway is NOT AVAILABLE** — a runway figure only means
something against a net outflow. Reporting one here would be inventing a
number.

### Working capital

| Metric | Value | Formula and inputs |
|---|---|---|
| **DSO** | **61.3 days** | (avg AR 6,308,964 ÷ credit sales 37,545,856) × 365 |
| **DPO** | **53.1 days** | (avg AP 3,389,276 ÷ COGS 23,278,431) × 365 |
| **DIO** | **30.2 days** | (avg WIP 1,928,750 ÷ COGS 23,278,431) × 365 |
| **CCC** | **38.4 days** | DSO 61.33 + DIO 30.24 − DPO 53.14 |

Every figure above is recomputable from the inputs shown. Sources: AR/AP
ageings and the financials summary in `/temp/input/`, all SHA-256 recorded in
`temp/output/04_working_capital.json`.

**Read on the cycle:** you collect in 61 days and pay in 53, so you are
financing roughly **8 days of the operating cycle** yourself, plus 30 days of
unbilled work in progress. The gap between DSO and DPO is the lever with the
most cash behind it — every day taken off DSO releases about **£103,000**
(£37.5m ÷ 365).

---

## 2. The 13-week picture

Opening **£2,001,620** → closing **£2,694,449**. The low point is **week 1 at
£2,018,538**; the balance rises from there.

**No week closes below zero.**

**The minimum-cash breach week is NOT AVAILABLE.** `min_cash_threshold` is
still `null` in `/temp/resources/config.json`. CashLens will not assume a cash
floor on your behalf. Set one number in that file and the next run reports the
first breaching week automatically.

### Confidence

- **Weeks 1–4** rest on invoices already issued with known due dates.
- **Weeks 9–13** rest on behavioural projection.
- **99.7% of projected receipts** come from customers whose own payment history
  is `HIGH` confidence (39 of 40 customers have six or more settled invoices).
  That is unusually strong; treat weeks 1–8 as reliable.

### A limitation you must read before acting on weeks 11–13

**Weeks 12 and 13 show zero receipts, and this is an artefact, not a forecast.**
The AR ageing contains only invoices *already raised*. Work you will invoice in
September and October is not in the file, so it cannot appear in the forecast.
The same applies to payments: the AP ageing holds only bills already received.

The practical effect is that **the tail of this forecast is light on both sides
and the closing balance of £2.69m is not a reliable estimate of where you will
actually be on 29 November.** Weeks 1–8 are sound. Beyond that, the forecast
tells you what happens to *today's* book, not what happens to the business. To
close that gap you would need a sales pipeline or billing forecast as a fourth
input.

---

## 3. What changed since the last run

**NOT AVAILABLE — this is the first run against this dataset.** No prior
snapshot exists in `/temp/output/history/` to compare against. Every subsequent
run diffs DSO, DPO, DIO, CCC, closing balance, total AR and high-risk count
against its predecessor automatically.

---

## 4. The three actions that matter this week

**1. Call Marlowe Toolworks about INV-10695 (£138,777).**
Highest risk score on the book at 63.9, largest exposure at £769,075 (13.4% of
total AR). This invoice fell due on 9 August and, critically, has now passed
*its own customer's* median payment lag of 20 days — it should have been paid
on 29 August. Marlowe pays late as a habit, but this is late even by Marlowe's
standards, which is what makes it worth a call today rather than a statement.

**2. Chase Harkness Industries (£577,958, 31–60 day bucket).**
Second largest exposure, 10.1% of the book, and the slowest payer of any
material account at a median **48 days beyond terms**. Nothing here has broken
its pattern yet — this is simply a large, chronically slow balance. Ask for a
dated commitment on the 31–60 day tranche.

**3. Set the minimum cash threshold in `/temp/resources/config.json`.**
Until it holds a number, the single most useful output of this forecast — the
week you first breach your floor — cannot be produced. It is one line of
config, and it makes every future run answer the question automatically.

Together, actions 1 and 2 address **£1.35m, 23.5% of the receivables book**.

---

## 5. What this brief cannot tell you

Stated plainly so nothing is inferred that the data does not support:

- **Cash runway** — not applicable; the period was cash generative.
- **The minimum-cash breach week** — no threshold configured.
- **Change since last run** — first run, no baseline.
- **Anything about revenue not yet invoiced** — outside the three inputs.

### Data quality notes

The following were corrected before any figure above was calculated:

- **6 duplicate rows** removed from the bank statement (double-posting
  artefacts).
- **2 duplicate rows** removed from the AR ageing. The deduplicated AR total of
  £5,739,933.91 reconciles to the financials summary **exactly (variance
  £0.00)**.
- **Three spellings of one customer** — "Northgate Precision Ltd", "Northgate
  Precision Limited" and "NORTHGATE PRECISION" — folded into a single account.
  Left unfolded, that customer's payment history would have been split three
  ways and its risk understated.
- **One credit note (CN-10477, −£18,450)** treated as a negative open item that
  reduces AR. It is not counted as an overdue receivable.
- **10 transactions have blank descriptions.** They are included in balances
  but excluded from recurring-pattern detection.
- **Two date formats** (DD/MM/YYYY and YYYY-MM-DD) appear in the bank
  statement. Each is individually unambiguous and was parsed on its own shape;
  neither was guessed.
- **Corporation tax was not projected forward** — it appears only once in the
  statement, below the three-occurrence threshold for establishing a pattern.
  It is identified but not forecast, so a payment may fall in the horizon that
  this forecast does not show.

---

*Figures derive solely from the files in `/temp/input/`. Every metric has its
formula and inputs recorded in the JSON envelopes in `/temp/output/`.
Projections reflect observed payment behaviour and are estimates, not
commitments. This brief recommends operational actions only.*