Genuine thanks to those who recognise that I am, at least, trying, to provide realistic and useful advice, without getting drawn into the rights or wrongs of the situation.
It is inherently unclear what evidence is being relied on. Clearly, what GA can
potentially access is:
- List of all claims made
- Details of train services customers claim to have used (or did use!)
- Personal information e.g. name, address, bank account number etc (needed to claim usually)
- Upload of a more traditional type of ticket, so an image. Ticket barrier data for this type of ticket is available, but seemingly not useful enough to break it down into something especially useful, except unless using a more unusual ticket which can stand out in the data
- Smartcard tickets - presumably they DO have access to the tap in and tap out data. Certainly I would work on that assumption.
- On the occasions I had to buy a season ticket, I would have to give my details etc in case I lost it (I think!) Not sure what else that (database?) potentially stores.
- Staff checks. Have customers who flag been monitored by plain clothes teams / CCTV etc? I wouldn't expect this to be common, but I'd imagine for the hardcore fraudsters (which I'm sure exist) they'd be doing a lot more.
- At the start of this, I found some references to cross checking of claims against other operators, presumably to stop duplicate claims, or to identify people in two different places at once!
There's probably more industry systems available, but I'll leave that to others to determine their usefulness. If I were approaching the investigation, I would be approaching it this way:
I would run some sort of comparison, e.g. Between X time frame, on average, how often does a person holding a season ticket claim from Ipswich to London.
Get a list of people who are above that average. Some cursory filtering. Possibly apply a tolerance to the average to eliminate borderline (e.g. slightly above average) cases.
Possibly an additional step in further filtering to produce a list of people who trigger the "above average" flag on more than 1 occasion over a longer timeframe.
----My gut feeling is that GA is doing this or something fairly similar, and at this point, letters go out.
If settled, no further investigation.
If disputed/ignored - at this point I think more in depth analysis is made, utilising whatever data they have available (including any responses received back from the customer). If at this point the investigation turns up suspicious claims, it goes to the police, or, potentially, the matter is discontinued. The reason I think more investigation happens before going to the police, is that GA will know that they need the BTP to have some level of confidence in their accusations, or it will affect their own relationship, and contrary to some belief on this forum, both BTP and GA will not want significant levels of adverse publicity. In any event, an interview under caution based just on number of claims is likely to be a very short one!