In the short term, do as Hymie suggests.
The other thing to do is to check roughly how many units per year you were using before the meter was changed versus now. Do not assess it on the cost of bills - the price per unit has shifted so much over the last 5 years - you need to relate it to units used.
I will now tell you a little story related to smart electricity meters which I resolved last summer. I was working for a client who had a commercial electricity contract and one of the properties had a smart meter fitted 3 yeasr ago. Since then, the electricity usage shot up by a factor of about 10 for no apparent reason. All attempts to discuss the situation with the supplier fell on deaf ears ('these meters don't go wrong') - made worse because the client changed commercial supplier in the midst of the problem. Finally, last summer, after we commissioned an independent electrical contractor to test the system with a data logger (an industrial version of what Hymie is suggesting) the meter company agreed to fit a check meter for a month and, guess what, the original meter was shown to be clocking at 10x the true rate.
The problem was in the original meter set-up - it had been set to clock at the wrong rate.
It is my understanding that these new smart meters are quite accurate at what they do, but they have to be set-up right. So don't try saying that a smart meter is 15% wrong - an error will be a multiple. But if it is say 200%, 400% or even as in my case 1000% wrong, you are probably onto something. BM