Great Western Coffee Shop

Newly published data - all Wiltshire stations plus some close neighbours

April 22 to March 23	April 23 to March 24	Station
5468478	6118136	Bath Spa
2588034	2831130	Swindon
1621562	1812122	Salisbury
1444888	1578070	Chippenham
734772	840926	Trowbridge
518998	535644	Westbury
426702	486806	Bradford-on-Avon
295484	331932	Warminster
207090	254698	Frome
151630	170286	Tisbury
81534	93570	Bedwyn
64424	63350	Melksham
28230	30004	Dean
23402	24136	Avoncliff
12076	19300	Dilton Marsh

Analysis to follow

Dilton Marsh stands out in particular, to me.  :o

The effect of providing a better service.

Melksham decreasing?

...and it's the only one to do so.

Sorry - just seen "analysis to follow".

Not unexpected - in fact better than it might have been.   

The service is cancelled far too often if they're short of crew, or short of trains, or doing engineering work somewhere and sending expresses through Melksham, or (as has happened today) sending trains our way because of an incident at Limpley Stoke.   And people can't travel on cancelled trains. Most avoid the rail replacement buses if they can, and that's even if they run them.  And this got worse in the 23/24 year over the 22/23 year. Beware, it's an even worse prognosis for the 24/25 year.   Do not expect to see better ORR stats for Melksham until they publish in November 2026. Once someone is let down badly, they take a lot of convincing to get back onto the "same" service.  And cancellation of even a single train when the next one isn't for 2 hours is a bad let down whereas a single cancellation on a 2-per-hour or 3-per-our service is a let down, but not a bad one for most people.

The good news is that people travel on trains and not on statistics so if we can persuade whoever's running the trains and making timetable and on the day decisions to be aware of what best serves the community, and actually run it, improvements could come once Westbury re-opens. 

Ah right - I just thought they based it on the town's predicted population growth; didn't realise it took factors such as this into account.

yep :)

Having used the bus instead of the train a few times recently I must admit I'm more inclined to use it regularly than I was. Especially since they brought the first 271 service forward about 7 mins so there's a much better chance of getting to Bath in time to catch the service I'd have changed onto at Chippenham if I'd taken the train. And even if I don't, there are more services running to Bristol through Bath than there are at Chippenham.

If only the 271 wasn't so crowded and/or had seats wide enough for humans. And it will be £2 more expensive soon; at the moment bus+train is only slightly more expensive than the train only option.

I'm answering in two halves and may split this part into a new thread.    The term "Estimates of Station Usage" can be misleading and is worthy of general coverage.   And, yeah, Melksham (MKM) is a good example.


It's nothing directly to do with the population.   It's based on ticket sales.    If I buy a return from Melksham to Sugar Loaf (yes, I have done) it counts a2 2 journeys for Melksham - 1 arrival and 1 departure, and 2 journeys for Sugar Loaf - again 1 arrival and one departure.

Simples - except

1. If I buy a season ticket, how is it counted?    For a week's season, I think it's five return trips each way, end to end. So a week's season ticket Melksham to Swindon adds 10 to the headline figure for each of Melksham and Swindon.  The fact that on Thurdsday, my partner dropped me off at Chippenham in the morning so it should have only been 9 and 1 for Chippenham is not recorded.  And the fact that I took a train on the season on Saturday (one that was running) up to Chippenham to meet the  station friends isn't counted.    For longer seasons, there's algorithms to work out the numbers based on expected holidays etc

2. If I buy a split ticket - e.g.Melksham to Didcot and Didcot to London and back it should be 2 for Melksham and 2 for London, but it ends up as that plus 4 for Didcot - so much of an issue these days that adjustments are made.

3. At one time, Melksham to Bristol was cheaper than Bradford-on-Avon to Bristol and an awful lot of "Melksham Specials" were sold to people going nowhere near Melksham.  So much of an issue that at one time I'm pretty sure that it accounted for two thirds of the Meksham ticket number.  To this day, I will buy an off peak Melksham to a Severn Beach line station return to go to Temple Meads, and for Melksham to Trowbridge it's cheaper to buy a Westbury ticket

4. Go to London and the ticket is "London Stations".  So from Melksham, it's Paddington.  Probably, but it could be Waterloo or Victoria

5. My trip to Sugar Loaf?   No - it wasn't. It was to Llanshamlet, out the long way round via Craven Arms, week's work for DVLA with a break of journey at the nearest station to them

6. What about rangers, rovers and freedom tickets?

7. It's ticketed journeys counted, so if there are any fare evaders they ain't counted.  Which is probably fair enough because really the future train service levels shoudn't be based on the defrauders.

So, you see, estimates.   Internally, the data is there between every pair of stations is available - not yet for the year to '24 thought - see http://www.firstgreatwestern.info//flows.html?stn=MKM and the grand total number is the same.

I am pretty certain that a member on this forum answered this one a little while back, Grahame - perhaps that helpful member might confirm?