The Fibre Uptake Gap: Which SA Networks Actually Turn Coverage Into Customers

South Africa's fibre industry spent a decade measuring itself by homes passed - how many properties an operator's cable physically reaches. It is the number that gets announced, the number in the investor deck, and the number that decides whether you can order fibre at your address at all. But it is not the number that decides whether the industry is working.

The number that does is homes connected: how many of those passed properties actually have a paying service. The distance between the two is the fibre uptake gap, and on the most recent disclosed figures for the eight networks we track, it is enormous.

Roughly 3 341 910 households in this country can order fibre today, with no civils work, no waiting list and no coverage question - and have not. That is not a coverage problem. Every rand of trenching for those homes is already spent. It is a price, awareness and affordability problem, and it is the single biggest commercial fact in South African fibre right now.

A caveat that matters: these totals sum figures with different as-of dates (see the per-network table below), so the blended rate is an indicative industry picture, not a same-day census. Networks that have not disclosed a connected count at all are excluded from the ratio entirely rather than assumed.

Here is every South African fibre network operator that discloses both figures, ranked by uptake rate. The pale bar is homes passed; the solid bar is the share of those homes actually connected.

Openserve leads on conversion at 53.9%, and it is worth being precise about why that is impressive rather than just large: Openserve is not a small network cherry-picking dense estates. It passes 1 565 750 homes and still converts more than half of them.

At the other end, fibertime converts 26.0% of the homes it passes. The 27.9-percentage-point spread between top and bottom is the real headline here. These operators are selling a broadly comparable product - a fibre line to a house - into the same economy, and one is roughly twice as effective at closing the sale as the other.

The obvious read is that the low-uptake networks are worse. That is mostly wrong, and the reason is worth understanding before you use any of this to choose a network.

Uptake rate is not a measure of network quality. It is mostly a measure of where a network built, and secondarily of how long it has been there. Three factors dominate:

1. Where the cable went. A network that built in high-income suburbs where almost every household already wanted fibre converts quickly. A network that pushed into lower-income areas as a deliberate strategy - which several South African operators have, correctly, done - passes homes where a R500-a-month line is a genuine budget decision. Slower conversion there is the predictable cost of expanding access, not evidence of a bad network.

2. How recently it was built. Homes passed count from the day the cable goes in. Connections accumulate for years afterwards. A network in a heavy build phase is constantly diluting its own uptake percentage with fresh, not-yet-sold coverage. This is precisely why Vumatel's figure deserves context rather than a verdict: Remgro's FY2025 results state that Vumatel limited homes-passed growth to 2% while subscribers grew 18.3% - a deliberate pivot from building to filling.

3. How many ISPs sell it. A network is only as visible as its retail channel. An operator with dozens of competing ISPs on it gets marketed constantly by all of them; one with a handful does not.

Put together, these explain most of the spread. A network passing newly-built coverage in price-sensitive areas through a thin ISP channel will post a low uptake rate no matter how good its fibre is. That is a commercial position, not a fault.

Here is the part that affects what you pay. An operator with a large base of passed-but-unconnected homes has already spent the capital on those streets. Every month a home stays unconnected, that spend earns nothing. The marginal cost of connecting one more household on cable that already exists is small; the marginal revenue is a full monthly subscription for years.

That asymmetry is why the aggressive offers in South African fibre cluster where uptake is weakest: free installation, waived router fees, first-month-free promotions, and entry tiers priced well below what the same speed costs on a mature network. The operator is not being generous. It is converting a sunk asset.

Against the live deals we track, here is what each network's uptake position looks like next to the cheapest per-megabit price actually available on it today:

Uptake and homes-not-connected computed from the footprint figures above (as-of dates differ per network). Price per megabit is the cheapest download-Mbps rate across every deal we currently list on that network, recomputed at each deploy, and counts each operator's retail sub-brands with it (MetroFibre includes Metro Nexus and Metro Nova; Vumatel includes Vuma Reach). "No deals tracked" means we do not currently list retail packages on that network - fibertime, for instance, sells prepaid direct to households rather than through the ISPs we track - not that none exist.

Read the table as leverage, not as a ranking. If you live somewhere a low-uptake network reaches, you are exactly the customer that network most wants and has the most room to discount for. Ask directly about waived installation and promotional pricing before signing - and check our fibre coverage map to see which networks actually reach your address, since in much of South Africa you will have two or three options rather than one.

Uptake rate should be a tiebreaker in your decision, never the deciding factor. Price for your real household usage, and the network's track record where you live, both matter more. Use it like this:

One thing the uptake gap genuinely does tell you: an operator sitting on 185 000 unconnected homes, as fibertime is, has every incentive to keep entry pricing keen and to keep expanding its ISP channel. That is a reasonable thing to factor into a two-year decision.

And the industry-level read is the encouraging one. 3 341 910 passed-but-unconnected homes is a large number, but it is a number that gets worked down by competition on price rather than by more capital expenditure. The most useful thing that can happen to South African fibre pricing over the next two years is exactly the fight the operators are now having over those homes.