Broadband take rate calculator
One neighborhood, two ways to fill it.
Fiber take rate, also called penetration, is the share of homes a network can serve that pay for it: active subscribers divided by serviceable passings, times 100. It decides whether a build pays back, because a passing costs the same to build whether or not the home subscribes. This model runs the number week by week for a build sold door to door against the same build sold through your other channels.
- Not subscribed
- Already subscribed
- Other channels
- Door to door
- Organic, from the D2D wave
Each home is 1% of the market.
The model adds 0.5 sales per rep per day for every $50 above $350, starting from 0.5.
Sales per rep per day
2.5
Campaign length
5.8 months
2 reps, 1,317 addresses
What a higher acquisition cost buys.
Set the per-subscriber rate above. The maps show how fast a neighborhood fills in; the numbers below compare 4 years of revenue and acquisition spend against the channels you already run.
An illustrative model. Every assumption is adjustable, and the defaults are calibrated to one real 1,317-address deployment.
Four years, side by side
Time to 40% penetration
4.8 months
Not in 4 years without D2D
Gross revenue, with D2D
$2.02M
$877k without
Acquisition spend, with D2D
$256k
$95k without
Net after acquisition, with D2D
$1.77M
$783k without, +$983k
Earlier subscribers, more months of revenue.
Drag the marker to read any month. Door to door costs more per subscriber; the revenue view shows what that spend buys, and when.
Month 0
Plus door to door2.4%
Other channels only2.4%
What a wave of installs does that a mailer can't.
Three effects of a D2D wave that no invoice captures. The defaults are estimates; set each to what you would expect in your market.
Sign-ups nobody invoices for
A rep knocks, nobody answers, the badge shows up on the doorbell camera, and the household signs up online that night. Installs on one street start the same conversation next door. Those come through your own channel, and Knoqly is never paid for them.
The ones who come back
Some households hear the pitch, ask their questions, and say not yet. A share of them sign up over the following months, already sold on fiber and already clear on what install day looks like.
Doors no mailer reaches
Some households will not sign up from an ad, a mailer, or a billboard at any price. A rep on the porch running a live speed test and walking through the install converts a share of them that no other channel touches.
Every other assumption
Your market
Most builds plan around 30 to 45% after 2 to 5 years. Industry figure, unverified.
Door to door
6 knocking days a week. A crew of 2 works a market over months; the 1,317-address deployment used a blitz crew of 8 and finished in weeks. The crew stays until the territory is within half a point of its ceiling, or a year, whichever comes first.
Other channels
Digital, mailers, events, whatever you run today. Fiber providers we have worked with land between $150 and $300 per subscriber outside door to door; published estimates run higher once installs are counted.
At this pace they reach 40% in Not in 4 years.
How this model works
- Each rep closes 0.5 sales a day at $350, plus 0.5 for every $50 above that, 6 days a week.
- A crew's weekly yield slows as the territory fills, by the square root of the share still unsold.
- Other channels add a fixed number of penetration points a year until they reach the ceiling, in both scenarios.
- Revenue counts every active subscriber at the monthly rate. Acquisition spend is each channel's per-subscriber cost times the subscribers it produced. Organic and later sign-ups cost nothing.
- No churn, no price changes, no chargebacks, in either scenario. 4 years simulated, week by week.
- With a blitz crew of 8, the model lands at 21.0% on day 14 against one real deployment's 19.74%: 1,317 addresses, 2.43% to 19.74% in 14 days. The page defaults to a crew of 2.
Take rate explained
How to calculate fiber take rate
Take rate is active subscribers divided by serviceable passings, times 100.
In one Knoqly deployment, 1,317 addresses gained 228 new subscribers in 14 days [Knoqly deployment data]. 228 new subscribers divided by 1,317, times 100, adds 17.31 points: 2.43% before, 19.74% after. For your own level, divide active subscribers by serviceable passings.
How the model works
Stated so you can argue with it. Each rep closes 0.5 sales a rep-day at $350 per subscriber, rising 0.5 for every $50 above that, six days a week. Weekly yield scales by the square root of the unsold share, so the first pass of an empty territory runs near full speed and the tail slows as it fills. Other channels add a fixed number of penetration points a year until they reach the same ceiling. Revenue counts every active subscriber at their monthly rate; acquisition spend is each channel's cost per subscriber times the subscribers it produced, and organic and later sign-ups cost nothing. Four years, week by week, with no churn, no price changes, and no chargebacks.
The calibration check: a crew of eight on the 1,317-address market reaches 21.00% on day 14 against the 19.74% that deployment actually hit, and 218.5 paid installs against 228 [Knoqly deployment data]. The model runs about 6% hot on penetration and about 4% light on installs at the anchor.
Homes passed, serviceable, unique, and connected
Four denominators appear in published take rates, and the numerator can be gross or net.
- Homes passed counts every address the plant reaches, connectable or not. Neither Cable One nor WideOpenWest files that count; both 10-Ks define the figure they report as serviceable addresses, quoted under serviceable passings below.
- Serviceable passings counts only the addresses that can order service today. Cable One reports passings, "which we previously referred to as homes passed," as "the number of serviceable and marketable homes and businesses passed by our active plant" (Cable One 10-K, 2024); WideOpenWest reports homes passed as "the number of serviceable addresses, such as single residence homes, apartments and condominium units, and businesses passed by our broadband network" (WOW 10-K, 2024). Pulse Fiber in Loveland, Colorado ages the count and includes only "fiber-enabled premises available in the Pulse network 90 days or more" (Pulse Fiber, 2020).
- Unique passings is the RVA LLC count of each home once, however many fiber providers pass it: 84.6 million unique homes marketed against total passings at the 100 million mark, with 16% of homes holding more than one fiber passing (Fiber Broadband Association, April 2026). The FBA headline average uses this denominator.
- Homes connected counts addresses with a drop installed, active account or not, so it runs above active-subscriber take rate. RVA reports connections as a count separate from passings and puts the time for most homes passed to become connections at about five years (Fiber Broadband Association, April 2026).
- Net vs gross take rate turns on the numerator. Gross counts every order written; net counts active subscribers after cancellations, disconnects, and chargebacks. The formula at the top of this page is net.
What a good take rate looks like
The US average fiber take rate is 46.5% on unique passings and 40% once homes with two or more fiber providers are counted, per RVA LLC data the Fiber Broadband Association published in April 2026. That average blends networks that opened last quarter with networks that have sold for a decade; in the same release RVA's Mike Render puts the time for most homes passed to become connections at about five years. For the number that matches your build's age, use the take rate benchmarks by year since launch, which lines up Frontier's 12 and 24 month targets and the NRECA and NRTC co-op survey by build age.
How to raise fiber take rate
Six levers an ISP controls, in launch order.
- Pre-register before the plant is lit, and build where the registrations are. American Fork, Utah runs a no-cost municipal registration and states that "neighborhoods that have the largest pre-registrations will be the first to get services": 76 planned areas, 17,111 registered as of 2026-09-07 (American Fork; registration map). From pre-registration to activation covers converting the list.
- Put an offer behind the order. Glo Fiber gives pre-registrants free wall-to-wall WiFi for 12 months, which it values at $120 (Glo Fiber, as of 2026-09-07). Ezee Fiber charges $10 to pre-register and returns the first month free, up to $109 by plan, plus priority installation (Ezee Fiber, as of 2026-09-07).
- Knock every serviceable address. A trained rep at the door with the offer and an install date [field experience]. One 1,317-address market went from 2.43% to 19.74% in 14 days, 228 new subscribers, in one door-to-door wave [Knoqly deployment data]; door-to-door sales for ISPs covers the crew.
- Book the install on the porch and keep the gap short. Orders that sit between signature and install are where sold subscribers disappear; track cancellations between order and activation as their own metric [field experience].
- Sign access agreements before knocking multi-dwelling units. FCC 22-12, adopted February 11, 2022, prohibits exclusive and graduated revenue sharing agreements with building owners and requires a disclaimer on marketing materials where an exclusive marketing arrangement exists (FCC 22-12). Unit access still runs through the property manager, and knocking without an agreement produces trespass complaints and no activations [field experience].
- Pair the fiber offer with mobile. AT&T's plan for the fiber markets it bought from Lumen, at about 25% penetration against about 40% in its legacy footprint, is more marketing spend plus convergence: 42% of legacy AT&T fiber subscribers take its mobile service against under 20% of Lumen's, and its CFO said in March 2026 that "a big part of our efforts will be focused on increasing penetration" (Light Reading, March 2026).
Questions about take rate
Five questions ISPs ask
Further reading
What is a good fiber broadband take rate? Benchmarks by year since launch
Fiber broadband take rate benchmarks by months since launch, sourced to FBA/RVA, NRTC, NRECA, and named operators, plus a diagnostic for a stalled build.
Fiber subscriber acquisition cost by channel: door to door, digital, direct mail, events
What a fiber subscriber costs to acquire by door to door, digital, direct mail, events, referral, and call center, with payback and chargeback exposure.
BEAD subscriber take rate milestones: turning letter of credit reductions into a subscriber plan
How BEAD letter of credit reductions tie to buildout for fiber and to 25% and 50% take rate for LEO, with a $20M worked example and a field plan table.
Run it with your own numbers, on a call.
20 minutes, your market, your current cost per subscriber.