This statement by consumer champion Martin Lewis has been widely reported today: "The rollout of the smart meters has been a disgrace in this country. Twenty percent of them don't work ...”. I saw it at second or third hand from this article, but I haven’t been able to find the source if it wasn’t in an attributable interview the Express had with him.
Does anyone know what the basis for this assertion is and where it was first made? Is it true?
EDIT: The government has revised up these official figures as it turns out some Suppliers gave them inaccurate figures. The revised current figures are in a later post in this thread
----------
The government Comptroller and Auditor General (C&AG) has a lower figure in its last report. See below.
This is an often quoted figure, although recently the government has been using a 2.7m figure sometimes from DNEZ.
It is difficult to know how accurate these figures are, for example, we do see examples on the forum where it appears ovo think a smart meter is working but it isn't in some way.
I do see short-term issues with my meter connection, whether they are picked up in the stats i don't know. It doesn't mean my bills are wrong on a regular basis, just I have to be careful what I am looking at sometimes.
It may be Martin Lewis has been mis quoted without knowing exactly what he was referring to with the figures. I haven’t heard or seen where Martin said this.
So the majority of smart meters/connections/accounts are working, perhaps around 90% plus or minus a bit at any one time, but clearly we see some painful examples with issues on the forum. It won't be the same people with issues all the time. Some are short term, some are more serious than others.
Unfortunately several organisations use surveys and present their findings as valid data whereas they are a group of opinions from a selection of chosen contacts - that’s one reason why I don’t find surveys generally useful. I’ve found some respond with what they think was required, rather than offering any factual information. That’s not to say that some smart meter connections (internal and external) are in need of strengthening although in many cases, better guidance has a greater effect
Bring me Sunshine ~ E&E {Solar PV, Battery Storage, EV, ASHP and Home Assistant automation}
I had it installed a few years ago and it worked fine after a few weeks or so. The Octopus dashboards are fantastic...
Still every time I would get someone knocking and taking a reading it would mess up the bills, really shocking values. This never hit our bank account but still caused some concern.
Eventually they fixed it. Something to do with the old reading point not having been completed disabled. We now have a policy that we only let anyone in that we are pre-warned about. Stopped it for sure. Apparently even if the gas meter readings are also sent, manual readings are needed...
In other words, smart meter setup is not as squeaky clean as it should be.
But we love the pricing that changes every 30 minutes and getting paid to charge our EV, so only look forward. 😁
I bet it’s the old thing where SMETS1 meters sometimes became “dumb” after folk changed suppliers. But that didn’t prevent them being used and read in the same way as traditional analogue or digital jobbies.
A 20% “failure” rate sounds ridiculous, that would surely be a ludicrously unworkable level!
I agree with waltyboy here, I wonder if it’s to do with the SMETS1 issue where they weren’t compatible with all suppliers? I’ve worked for OVO for a long time and would be very surprised if this were the case. I wonder what the source is? If it is the Which survey the problems are ‘self reported’. I know that sometimes there can be confusion when it comes to smart meters- with estimated readings for example, when a customer is on monthly meter reading schedule and this causes them to believe that the smart meter isn’t in communications.
The latest government "official" figures are here as of end 2023. Quarterly updates.
This is collated from data supplied by energy suppliers so will include data that ovo have supplied rather than subjective customer data in the which survey and other surveys. Historical figures have recently be revised up as some suppliers had given the government inaccurate figures.
So 11.4% of smart meters in traditional mode at the end of 2023 according to supplier data for a number of reasons that are described in the report that are important for context. It is a snapshot in time.
This includes some non domestic and advanced meters.
There is a spreadsheet for those really interested in the detailed breakdown over time and splitting out domestic and non domestic etc.
I wonder if OVO would be willing to share their slice of the data in the official figures out of curiosity. It would be interesting how the figures vary across suppliers. Some Suppliers may well have more smart meters "working in traditional mode". We know ovo have been fined for the speed of their rollout of smart meters so there are likely to be differences in the splits for this reason.
This official data will never line up with the which data
1. The which data asks a sample of customers if they have had problems over the last 12 months.
2. The official data is a snapshot of meters that the supplier cannot contact at a specific time.
The percentages in the which data will always be higher as the time period is 12 months.
And today’s article summarising the figures in the online Guardian supports the kind of figures suggested by @Jeffus and others above. Oddly, the article refers only to “homeowners” benefitting from smart meters.
Final two paragraphs imply some 11% below par…but even so, that seems only to refer to the smart meters being no worse than traditional meters requiring quarterly estimates or submission of readings by consumers.
”Smart Energy GB, a government-backed non-profit set up to help with the smart meter transition, said there are now almost 35m smart meters in Great Britain “and the vast majority are operating as intended”.
It added that there had been an overall improvement in the proportion of smart meters operating in smart mode since 2022, increasing from 87.3% at the end of 2022 to 88.6% at the end of 2023.”
Walt. 2013: 3.2kW PV East/West + Fronius IG, as at 01 Sep 2026 30 MWh, of which 22 MWh exported. 2023: 2.4kWh LiFePo4 battery. 2024: Tadö wireless smart stat v3; Electriq split unit 2.6kW AAHP (“air conditioner”). 2026: Hyundai Inster 02 EV.
And today’s article summarising the figures in the online Guardian supports the kind of figures suggested by @Jeffus and others above. Oddly, the article refers only to “homeowners” benefitting from smart meters.
Final two paragraphs imply some 11% below par…but even so, that seems only to refer to the smart meters being no worse than traditional meters requiring quarterly estimates or submission of readings by consumers.
”Smart Energy GB, a government-backed non-profit set up to help with the smart meter transition, said there are now almost 35m smart meters in Great Britain “and the vast majority are operating as intended”.
It added that there had been an overall improvement in the proportion of smart meters operating in smart mode since 2022, increasing from 87.3% at the end of 2022 to 88.6% at the end of 2023.”
An additional challenge is whether some of the functionality in smart meters can realistically be utilised any time in the near future.
There are load control features that currently are not being used by the likes of OVO and Octopus. So OVO Anytime relies on the Internet for switching as does Intelligent Octopus Go.
It will be interesting to see how the market evolves and whether the stats can get to a point that suppliers look to utilise more of the smart meter features. Else we as customers have spend an awful lot of money on the smart meter rollout when a cheaper cut down solution could have been deployed. The latest estimate is £13.5 billion at 2011 prices... so an eye watering amount in real terms.
From the impenetrable figures in the tables Jeff referred to, we can see that dumb meters account for 11% of all supposedly smart ones, We don’t know whether they’re ‘working’ or not, but assuming they are, that leaves 9 of Lewis’s 20% unaccounted for. Some of those 11% will be SMETS1 devices that couldn’t be connected to DCC, which would become dumb on a change of supplier as Emmanuelle points out. Any customer in this situation will have been made aware of the possibility.
I’m still wondering about the large number of non-working IHDs - is it likely that 10% (40% of the 23% of Which? respondents who said so) of them aren’t working properly? I’m sure that there is a significant number of customers who simply give up on their IHD when it won’t tell them what they want to know and stuff it in a drawer somewhere. That isn’t evidence of a non-working smart meter.
As Balfour stated, “There are three kinds of falsehoods: lies, damned lies and statistics.”
Noel | I have no official status; I'm just a volunteer who comes here to help other customers. My gear: Aclara SGM 1416-B Electricity-only E7 meter; Chameleon IHD3-PPMID-AAA | It may look as if I know what I’m talking about, but don’t let that fool you. |
From the impenetrable figures in the tables Jeff referred to, we can see that dumb meters account for 11% of all supposedly smart ones, We don’t know whether they’re ‘working’ or not, but assuming they are, that leaves 9 of Lewis’s 20% unaccounted for. Some of those 11% will be SMETS1 devices that couldn’t be connected to DCC, which would become dumb on a change of supplier as Emmanuelle points out. Any customer in this situation will have been made aware of the possibility.
I’m still wondering about the large number of non-working IHDs - is it likely that 10% (40% of the 23% of Which? respondents who said so) of them aren’t working properly? I’m sure that there is a significant number of customers who simply give up on their IHD when it won’t tell them what they want to know and stuff it in a drawer somewhere. That isn’t evidence of a non-working smart meter.
As Balfour stated, “There are three kinds of falsehoods: lies, damned lies and statistics.”
It is not a like for like comparison
This official data will never line up with the which data
1. The which data asks a sample of customers if they have had problems over the last 12 months.
2. The official data is a snapshot of meters that the supplier cannot contact at a specific time.
The percentages in the which data will always be higher as the time period is 12 months.
My meters would definitely be in the 23% if I had filled in the which survey but very unlikely I would have been in the snapshot official data from the suppliers seeing if they could see my meter.
