The Challenge
A fast-growing UK internet service provider had reached a critical juncture. Monthly net cash outflow was running at £1.9M — but the board couldn't tell which part of the business was driving it. Their existing spreadsheet model was a patchwork of assumptions built over three years by four different people. Actuals were never reconciled to the model. Scenarios were guesswork.
With a fundraise approaching and 14 months of runway left, the founding team needed to know: where exactly is the money going, and what levers actually move the needle?
Our Process
We started by deconstructing the existing model rather than discarding it — preserving the institutional knowledge baked into old assumptions while rebuilding structure around it.
Revenue architecture. We separated subscriber revenue into five cohorts by acquisition channel and tariff tier, each with its own ARPU, churn rate, and marginal cost of service. This alone revealed that the two fastest-growing segments were also the two lowest-margin ones.
Infrastructure cost mapping. Backhaul, peering, and last-mile costs were rebuilt on a per-node, per-region basis and linked to subscriber density assumptions. For the first time, the model could show the true unit economics of expanding into a new postcode area.
Headcount and OpEx waterfall. Every cost line was mapped to a driver — headcount to subscriber support ratios, sales costs to conversion funnels, G&A as a function of revenue scale. Nothing was left as a flat assumption.
Three integrated statements. We built a fully integrated P&L, balance sheet, and cash flow statement, with a dynamic working capital model that flagged the precise month each covenant threshold would be breached under each scenario.
Key Outcome
The model surfaced a structural problem that had been invisible in the old spreadsheet: the company was pricing its residential broadband tier at a level that covered direct costs but not an accurate allocation of network infrastructure depreciation. At scale, this tier was quietly destroying margin.
Armed with this insight, management repriced the tier, accelerated churn in unprofitable postcode areas, and redirected sales effort toward SME customers — which the model showed were contributing 3.1x the lifetime value per acquisition cost.
Results
- ·£720K reduction in annualised burn within two quarters of implementing model recommendations
- ·Fundraise closed at a £12M pre-money valuation — the model was cited by the lead investor as "the clearest financial story we've seen at this stage"
- ·14 months runway extended to 22 months without additional external capital
- ·Board reporting cycle reduced from 3 days to 4 hours using the model's integrated actuals-vs-budget dashboard