Badly, unless you drive. Half the working population gets into a car. Nearly a third never leaves the house to work at all. The station runs steam trains for people on holiday. Watchet is reachable rather than connected, and those are two very different words.
Rail here is a visitor attraction rather than a transport option. The line through the town is a heritage steam operation, and the share of working residents who commute by train rounds to nothing. Buses do more of the real work: the area has 123 stops, and about 3% of workers use them to get to their jobs.
What actually moves people is the car and the spare bedroom. About 52% of working residents drive to work and roughly 30% work from home, which between them account for the overwhelming majority of weekday behaviour. Around 8% walk and about 2% cycle. Roughly 15% of households have no car at all, which in a place shaped like this is less a lifestyle choice than a daily negotiation, while about 42% run two or more.
The Department for Transport's modelled journey times put the nearest town centre at about 27 minutes by car, roughly 60 minutes by public transport, around 58 minutes by bicycle and just under two hours on foot. That walking figure is not a quirk of the modelling. It is the finding. These are national statistics modelled for the area rather than a route from any particular front door, but the shape of them holds.
The same statistics put a doctor's surgery at about 13 minutes by car, roughly 37 by public transport and around 56 minutes walking. A hospital is about 50 minutes by car and roughly 88 minutes by public transport, which is the number worth thinking hardest about before buying here. A food store is around 11 minutes by car and about 44 minutes on foot. Employment centres sit at roughly 59 minutes by car and about 51 minutes by public transport. Those journey times were modelled in 2019, and while the distances to fixed destinations have not moved, the public transport figures have aged worse than the car and walk times as rural services have thinned.
Current bus service data covers 51 of the stops, and across those there are 73 scheduled departures on an average weekday, with roughly 5 in the peak hour. About 80% of stops have some weekend service. That is a genuine service rather than a token one, though departures at a stop are not the same as routes going where you need to go.
Broadband is the modern version of the same problem. Superfast is available to about 88% of premises, which is respectable, but gigabit and ultrafast both reach only around 13%. Under 1% of premises fall below the universal service obligation, so almost nobody is stranded, but anyone whose living depends on a stable connection should check the specific address rather than the postcode average. Working from home is the dominant commute here, and the infrastructure supporting it is adequate rather than good.
Mobile coverage is better. At least one operator provides fourth or fifth generation service at nearly 100% of premises, and all operators manage it at about 97%, so the network you happen to be on matters less here than it does in most rural areas.
One oddity worth knowing: nearly all postcodes in the area sit near a road modelled at or above 65 decibels, the level at which official noise mapping treats road noise as significant, and the loudest band present reaches 75 decibels. That says something about how the settlements here are strung along their roads rather than anything about how any individual house sounds.
For electric vehicles, the area has 5 public charge points offering 5 connectors between them, and none of them are rapid. Charging here is something you plan around overnight, not something you do while picking up milk.
Connectivity is the weakest part of the case for Watchet. The car works, the broadband is adequate, and everything else asks you to either own a vehicle or genuinely not need to go anywhere.