Decisions that made the business smaller on paper and sounder in fact

Restating our customer count before the lenders asked

Ridgeburn's draft lender report for the end of 2021, my first year as CFO, showed 357,200 broadband subscribers and 1,080,000 homes passed. When my team reconciled those figures against the billing system, our installation records and county property data, both turned out to be inflated. The subscriber figure included 9,800 apartment units billed to landlords at a bulk rate, about 4,500 non-paying storm-relief and staff accounts, and several hundred business accounts counted once per site. The passings included 61,000 addresses we could not connect at our standard cost: empty lots, farmhouses far from the road and buildings lost to a 2019 tornado.

The stakes were not only presentational. Granmoor's valuation of its holding, our lenders' view of our growth headroom and the management incentive plan were all built on those figures, and the excluded accounts never churned, so they flattered our retention as well as our count.

I recommended restating the year-end 2021 figures to 342,900 customer relationships and 1,019,000 serviceable locations, and reporting bulk and non-paying accounts on separate lines. Our chief executive disagreed. He argued that competitors would keep publishing the larger figures, that we would look weaker beside them, and that a restatement in the year before a refinancing would invite questions. Granmoor's operating partner shared his concern about the valuation. The audit committee chair supported the restatement on condition that we restated two prior years as well, which took my team four months.

We gave up a cleaner story. The incentive plan was rebased, including my own award, and Granmoor marked its holding down by an amount it did not disclose to us.

The value showed in 2023, when fixed-wireless competition cost us 9,400 customers in a year. Because the base had been cleaned, the losses reached the board at their true size in the quarter they happened, concentrated among short-tenure customers on slow tiers. That evidence shaped the second fibre tranche and the 2024 refinancing, where every credit committee asked whether our figures could be trusted and received three years of reconciled history.

What I would do differently: I would have brought the regional general managers into the reconciliation from the start. They learnt of the restatement from a memo, and two of them read it as an accusation.

Releasing the fibre programme one tranche at a time

In mid-2022 our chief technology officer and chief executive proposed upgrading the 41 towns with our oldest coaxial plant, 410,000 locations, to fibre to the premises within three years at a cost of $520 million. Fixed-wireless offers were spreading and a fibre overbuilder backed by an infrastructure fund had announced plans in two of our larger towns. Their case for speed was strong: crews are cheaper in volume, and a town upgraded after a competitor arrives is harder to win back.

My concern was that the plan assumed a construction cost of $1,050 a location and a take-up curve we had never tested, while net leverage was heading towards six times EBITDA. If costs ran 20 per cent over across the whole programme, refinancing our 2025 maturity would become far harder.

I proposed releasing the programme in four tranches. Each would proceed only when the previous towns met three tests: construction cost within 10 per cent of plan, net new customers equal to 4 per cent of locations within a year, and monthly churn below 1.5 per cent. The chief technology officer disagreed at the board, and Granmoor's operating partner supported him. The board approved the gates, but at its insistence we committed the first two tranches to contractors together to protect crew pricing.

The first tranche of 118,000 locations failed the cost test at $1,240 a location, mostly because electric cooperatives' make-ready work took longer and cost more than our engineers had assumed. The second tranche was delayed by five months while we renegotiated pole arrangements and changed our design standards. In that time the overbuilder reached two towns ahead of us, and we lost about 2,600 customers there that we have not regained.

The second tranche came in at $990 a location. By December 2025, 236,000 locations were on fibre at an average of $1,115, monthly churn in upgraded towns had fallen from 1.9 to 1.2 per cent, and revenue per customer there had risen by $6 a month. Every capital programme at Ridgeburn now carries release tests written before the first dollar is spent.

What I would do differently: I would have negotiated make-ready terms with the cooperatives before the first tranche, not after it failed. The two towns we lost were the price of learning that late.

Letting the video business shrink until it paid its way

By 2022 Ridgeburn's video service had 129,000 customers, down from 171,000 two years earlier, and earned $139 million of revenue. When I allocated its full costs, programming, set-top box depreciation, the care calls video generated and its share of billing, it lost about $4 million a year. Programming costs per customer were rising by about 9 per cent annually, led by retransmission fees for local stations and a regional sports network that charged $5.40 a month for every basic customer.

The commercial case for keeping video as it was rested on retention: customers who bundled video with broadband churned at a third of the rate of broadband-only customers. Our chief revenue officer believed that ending promotional bundles would accelerate broadband losses just as fixed wireless arrived.

I asked our analysts to test the claim on matched cohorts of customers with similar tenure, town, tier and payment history. Most of the difference disappeared. Bundled customers churned less mainly because they were older, longer-standing and more often homeowners; the bundle itself explained about a quarter of a percentage point of monthly churn. Priced at that effect, the loss on video was larger than the retention it bought.

I recommended four changes in 2023: stop buying set-top boxes and move customers to an app-based service on their own streaming devices, end promotional video bundles for new customers, decline to renew the regional sports network, and show local broadcast fees as a separate line on the bill. The chief revenue officer opposed the bundle change and the sports decision. The board approved all four, and I recommended writing off $9 million of box inventory immediately.

The cost was real. About 4,200 customers left within six months of the sports network's removal, and we estimate that 1,700 of them also took their broadband elsewhere. Some town councils that grant our franchises were unhappy, and they were entitled to be. Video customers fell to 84,000 by December 2025.

Video contribution moved from a $4 million loss in 2022 to an $11 million profit in 2025, and $17 million a year of box spending moved to fibre.

What I would do differently: I would have written to the town councils before the sports network left, not after. They heard from residents first, and we spent a year repairing that.

Refinancing a 2025 maturity without a dividend recapitalisation

By early 2024 Ridgeburn had $1.24 billion outstanding on a term loan due in October 2025, the remainder of the financing used for Granmoor's 2018 buyout. Net leverage had peaked at 5.9 times EBITDA in late 2023, fixed-wireless losses were still fresh, and the fibre programme needed another $300 million over three years. If the loan was still outstanding when our 2024 statements were issued, management and the auditors would have had to assess whether its maturity raised substantial doubt about our ability to continue, a disclosure that would have damaged us with suppliers, town councils and grant agencies.

Two routes reached the board. Granmoor's deal team favoured an amend-and-extend in the syndicated loan market, at a margin of about 4.50 per cent over the benchmark rate, with no maintenance covenant, and an incremental $175 million to fund a distribution to shareholders. A club of five private-credit lenders led by Ellisand Credit Partners offered a $1.25 billion unitranche at 5.25 per cent over the benchmark, with a $200 million delayed-draw tranche for fibre and grant matches, and a net leverage covenant starting at 6.75 times and stepping down to 6.0 times by 2027.

I recommended the unitranche without a distribution. The syndicated route was cheaper on paper, but it depended on a market window, left fibre unfunded and added $175 million of debt at the moment our EBITDA was weakest. Granmoor's deal partner argued that private credit was expensive and that the covenant would restrict us in a bad year. Both points were fair. The audit committee chair asked for covenant projections under a downside case with 4 per cent fewer customers, which showed headroom of at least 0.6 times throughout.

We closed in May 2024, seventeen months before the maturity. The higher margin costs about $8 million a year, and the original issue discount cost $24 million. Granmoor forwent the distribution, and two of our managers left the incentive plan earlier than they would have.

Net leverage was 5.5 times at December 2025, the delayed-draw tranche has funded the second fibre tranche, and our lenders now receive the same monthly pack as our board.

What I would do differently: I would have started the private-credit conversations six months earlier, when we had more choice of lenders and terms.