Energy Usage Graphs useless and misleading

I have been communicating (8 emails now) with Bulb about the completely useless Energy Usage graphs that bear no resemblance to reality, especially the “Actuals” to date!
I haven’t smart meters but provide readings at the end of every month, I would expect the actuals to date to reflect at least closely the actual bill paid but this is not the case.
Graph Dec £451 actual bill £279
Graph Jan £480 actual bill £274
Graph Feb £496 actual bill £333
Graph Mar £449 actual bill £277
I have been given various explanations for this the latest being:

“I’ve received feedback from my team and they’ve advised me that the graphs for legacy meters are always estimated.
This is due to Bulb not receiving regular meter reads, this results in Bulb using your estimated annual consumptions to update these graphs.”

To which I replied:

If what you are saying about “legacy meters are always estimated” is correct then that makes the graphs absolutely useless.

Even if the bill for a month is based on estimated readings the “Actuals” part of the graph must reflect the actuals paid not over £150 more as is happening in my case.

As said previously “The “Predicted Usage” should be a calculation of previous years actual usage for that month together with the variation of previous actuals for the current year (i.e. In case you have started using more/less than previous years due to a change in circumstances as in my case).”

In my case, except for 2 months when I was grieving for my late wife, I have always submitted regular meter reading every month when requested thus this should not be any different from the smart meters.

I have just submitted my readings for May, which includes the 2 days that were not included at the end of April as the reading for April was done on 28/4/22.

This has caused the graph to now put actuals for April of £223 which is very close to the actual bill for April (based on the readings on 28/4/22) of £207.68.

I therefore cannot understand why the “Actuals” for Dec, Jan, Feb and Mar 22 are so wildly different from the actual bills paid by hundreds of pounds?

I think your team need to take a closer look at how the graphs are calculated as they are wildly inaccurate making them useless.

I would be very grateful if they could explain to me how they arrived at a figure of £480 for Jan 22 when the actual bill was £274 (i.e. over £200 more).

There is no way I could have used £206 of energy in a couple of days between the meter reading and the end of the month of Jan!!!

Sorry to keep on about this but I think Bulb are completely misleading their customers with these graphs as I suspect the “Predicted Usage” part is also incorrect.

If these figures are being used to show the yearly predictions of energy usage which may well be reflected into the expected direct debit then I think OFGEM would also be interested in how you are coming up with the figures."

As yet no reply


I can guarantee exactly what Bulb will say “IF” they bother to reply…

“ Our usage graphs are in beta mode, so offer an indicator to your usage but are not 100% accurate.”

They’ve said the same thing over and over to everyone who comments on this subject. What bugs me is that the usage chart has been in “beta” mode ever since it was launched god knows how long ago. It’s just an excuse to hide behind their lack of funding for the app and their piss poor level of customer service.

The only advice I can offer is to send daily readings to bulb so there’s no errors in your usage. Doing this has meant I haven’t had extortionate rises in my direct debit and or ridicules estimates of my usage.


This is pure laziness on Bulb’s part, I’m in the same situation with the energy usage graphs implying the actual energy use being real data based on actual bills, when in fact it’s just a random number generated by an algorithm.

It’s one thing estimating the future figures, but estimating the actuals very unprofessional & misleading the customer.

While Bulb pay there employees retention bonuses the customers can’t get even the most basic correct information.


I can concur with all comments above.

I just checked my energy usuage graph & compared against actual payments, the current graph for Dec’21 to Mar’22 is way off, but April 22 is OK:

Dec’21 on graph = £200, actual = £120
Jan’22 on graph = £380, actual = £271
Feb’22 on graph = £310, actual = £208
Mar’22 on graph = £140, actual = £100
Apr’22 on graph = £240, actual = £240

Of course the graph for actuals should match payments exactly. I had screen grabbed earlier graphs & I noticed the vertical scaling had changed from a few months ago, so I wonder if bulb’s algorithm has somehow screwed up on that & incorrectly mixed bars of different scales. They should just keep the graph scale constant at £100 vertical intervals & not mess with it. I’m currently better off just loading my own data in a spreadsheet.

This is school-boy programming errors which should be easy to fix.

Not impressed at all.


The more worrying aspect is that they seem to be using these figures to calculate the predicted annual usage which then reflects into the suggested Direct Debit. If you total up the “actuals” and “predicted” values on the graph they are very similar to the “12 month personal projection” on the most recent bill. Divide this by 12 and it gives you the “recommended monthly payment” which in my case would increase my DD by £200 a month! What a mess!
Recently the energy companies have been accused of over inflating their DD to bring in extra cash, all have denied doing this, but if the data they are basing the DDs on is incorrect (as this appears to be with Bulb) then it’s a way of over charging.
I am keeping my DD as it is and if the actuals for a month are greater then I just top up to give a zero balance.
As said earlier this is schoolboy stuff and should be sorted out immediately.


Bulb has set me £186 per month, this can be adjusted in the direct debit payments section, it’s suggested at £168 minimum, so I have done just that. We are already in credit so it’s not much point in giving money away to a failed company currently in the hands of the administrators while paying the CEO £250,000 per year.

I have just received a response from Bulb regarding my last email:

"This is just a follow up email to the last email that I’ve sent to you.

I’ve received an update from my team and they’ve advised me that the only way you’re able to correct this is have your meter reads submitted each month.

In order for Bulb to bill you accurately and for your usage to be updated, you would need to submit a meter read on the 1st of every month.

I hope this helps."

I am just going to respond to this load of rubbish, I’ll post my response here later.

So much for being within three days of the last day of the month.

I sent mine yesterday which means another incorrect month.

I’m going to play their game and send in readings today and tomorrow just to see what utter rubbish this might be.

When I took the meter reading yesterday I used the ‘scan your meter’ function on my Bulb smart phone app, it rejected the reading because it needed a 6 digit number, it was a six digit number and it accepted it when input manually, another example of technology failures.

1 Like

My response to the most recent email listed above:

Sorry but the response from your team is rubbish and I notice that they have not explained how they arrived at a figure for Jan 22 which is £200 more than the actuals.

Do they even know how it is calculated???

I worked in the IT industry for 45+ years some of which was as an independent consultant in an Energy company.

As stated previously I have submitted meter readings each month (except for two due to circumstances), I have done this when requested by Bulb that is within 3 days of the end of the month NOT on the 1st of the month.

You cannot submit on the 1st of the month as it is too late for the billing period and estimates will already have been made to make the bill for the previous month. Even if you submit the last day of the month that may well be too late for the billing period. By submitting on the 1st of each month would cause rolling estimated bills month after month which on present form could be wildly out looking at present calculations!

I think your “team” need to look more closely at the calculations for the graph especially since these figures seem to be used to calculate the predicted annual usage which then reflects into the suggested Direct Debit.

If you total up the “actuals” and “predicted” values on the graph they are very similar to the “12 month personal projection” on the most recent bill. Divide this by 12 and it gives you the “recommended monthly payment” which in my case would increase my DD by £200 a month!

This is a complete mess and needs looking at urgently as you may be over charging your customers.

Recently the energy companies have been accused of over inflating their DD to bring in extra cash, all have denied doing this, but if the data they are basing the DDs on is incorrect (as this appears to be with Bulb) then it’s a way of over charging.


Yep totally agree with your findings is that the graphs appear pretty useless and as i don’t have a functioning Smart meter. I.e. not connected to anywhere and have to send monthly readings for over a year now. I cannot tell if changes to our heating and energy consumption are having any effect. It would be better to graph the actual energy (KWHr’s ?) consumed as value is meaningless as price changes so bleeding rapidly.

I tend to agree with you about the graphs being based on KWHrs as the data must be calculated to give the predicted usage. The ideal would be to have two graphs (or switchable) one with £ usage and the other with KWHrs. Fat chance of that happening and god knows what useless information it would show based on the present graph!
So the ideal would be:

  1. Graph 1 in KWHrs with “actuals” showing ACTUAL usage to date and “Predicted” showing a prediction based on previous years usage modified to account for changes to usage in the actuals to date (i.e. due to reduced/increased usage/month to date compared to previous years)
  2. Graph 2 in £ usage with “actuals” showing ACTUAL bills and “Predicted” based on the predicted KWHrs modified to incorporate any known increase in price (e.g. increase in the price cap).

This would then give a much more accurate estimate as to what the monthly DD should be.
As I said FAT CHANCE of this as they cannot even get a basic graph correct.

As an aside I have just got my bill for May, the predicted was £326 but the actual was only £118 which when my DD goes out today will make me £50 in credit!
As this has now given “actuals” for April which are just about correct for a change (the prediction was about £200 more than the actual) it has reduced my predicted yearly usage by £200 (widely inaccurate). My suggested DD is still about £200/month extra.

This is a complete shambles!!!

Yep. Great idea if only??
Did a spreadsheet and graph in excel showing KWh usage over last year. Will keep adding usage numbers to see if our attempts at saving energy are working​:woozy_face::pray:


First of all I’m in total agreement with the comments in the thread. I too find the graph infuriatingly unhelpful. I think that the main problem is that someone decided that it would be “helpful” to express the graph in pounds-cost rather than gas/electricity units used as this would, perhaps, be clearer to less mathematically confident customers. (I’m guessing.) But even the most challenged mathematically can compare a graph total with a bill total and see that it is wildly different. And as Bulb don’t explain how they reached their totals that leaves all of us perplexed as to what exactly is going on. I see various potential issues with the graph design. As follows:

First they really ought to say that the monthly bill refers to the previous column on the graph. As in the May 1st-3rd bill relates the April graph column. Obvious I know, but it’s really easy for a person to go “May bill, therefore May column” and the design should take account of this tendency to make a simple error.

Second, although Bulb claim that the graph shows “actual” usage it does not. It shows billed units. If an estimated is used then that is what is used to calculate the graph. The column will be modified in the following month to reflect an actual reading, but the original estimate will be left in place. Perhaps understandable as this will match the bills, but in fact neither the estimated column nor the modified column will be “actual usage” which is what Bulb claims them to be.

Thirdly I don’t know if Bulb factor in either or both the daily standing charge and/or the VAT (currently 5%) into the columns as shown on the graph. If they don’t then bills (which do include both) ought to be consistently higher than the graph. But if they do include these factors then the graph isn’t actually showing energy usage, which is what they claim it is showing.

Fourthly Bulb are going to have a problem every time the unit cost changes as they cannot have different rates of scale on a single, simple, graph. I noticed, because I was looking out for it, that when the unit price rose at the beginning of April the “cost” of each column (for months prior to April) had been recalculated at the new, higher, rate (although of course I had paid the bills at the old, then applicable, lower, rate). As the rise in unit cost was in the region of 50% (differing for individual rates and the daily standing charge) the graph column totals therefore suddenly significantly rose above the actual bill cost. However if the Bulb predictive algorithm (I’m guessing again) just uses the last 12 months totals - as recalculated and shown on the graph - then this would explain the enormous and unjustified jump in predicted bills. Under normal conditions this would work itself out over 12 months, but this year we will be getting a further rise in October - and a potential readjustment every three months after that - so this may be an issue that stays with us fora year or more.

I’m at an additional disadvantage as I have Economy7 and the graph doesn’t even try to manage that. My night usage is about 80% during about three months in the winter when I have the heaters on, but the graph just lumps both rates together and averages them with a weighting for the last few months. Which results in massive overestimates in the winter when my usage peaks, followed by dramatic falls in annual prediction during the summer when my usage is significantly less. And it just doesn’t even try to factor in the effect of the different rates for day and night. For example one recent month my night units used were a third for the total units used, however, because the night units cost less they cost only an eighth of the total bill price. Again it means that my predicted usage tends to be wildly off.

Honestly Bulb, just give me a graph of units billed, gas, electricity (day and night rates if applicable) and leave it at that.

1 Like

Your first point about the graphs referring to the previous month is incorrect, this only appears to be the case because if you give a meter reading 3 days before the end of the month then (as requested by Bulb to meet the billing run) then the actual month is 3 days short of a reading. This will not be completed until you give a reading which includes those 3 days (i.e. 3 days before the end of the next month). I’ll give an actual example:
i) I gave a reading on the 28/4/22 (i.e. 2 days before the end of April), the graph gave an “actual” for March as all March was now covered but April was still estimated as it was 2 days short.
ii) I gave a reading on the 29/5/22 which as it included the last 2 days of April gave an “actual” reading for April but still an estimated on for May (as May was 2 days short).
iii) As suggested by Paul-S I gave another reading on 1/6/22 which because it included the 2 days in May gave an “actual” for May, so the graph is up to date although the “actuals” are slightly different from the bills as the bills are based on the readings given just before the end of the month.
So the answer is to give a reading 2 days before the end of the month so that a bill can be generated without using widely out estimates, then submit another reading on the 1st of the month. If you don’t submit regular readings then the readings are estimated by Bulb and the graph will show the estimate.
None of this explains why the “actuals” for Jan, Feb and March are so widely out from the actual bills or why the “predictions” for June-Nov have not been recalculated based on increased/reduced usage for April and May compared to previous years.

The above I think explains your second point about “estimate will be left in place”.

Regarding your third point I think the graphs do include both VAT and the standing charge as my graph “actuals” for March and April are very close to the bills (which are 2 days short of readings).

Your 4th point is a good one, there is no way history should be changed on the graph, so a change of rate must not affect previous readings. I didn’t sport this but if true then it shows Bulb do not have a clue what is going on, I wish I could rewrite history!!!

Regarding your Economy 7, I assume that it is a separate meter reading at a different rate? If that is the case then this should be reflected on the graph as a separate entity. That would then show your increase in usage during the winter months. Fat chance of that happening. The other way is to forget about the 2-3 days difference in readings at the end of the month and just show the “actuals” on the graph to match the actual bill. That would be OK for the “actuals” but problems for the “predictions”. You would have to have a separate prediction for Economy 7 usage as it’s a different rate.

I can now take this uselessness to a completely new level.

Today, the 7th of June, I looked at my usage figures for the 6th. I used £1.36 of electricity and somehow used £3,605,638.37 of gas in one day. If I could post the screenshot I would.

I must have left a pan boiling on the hob.



Bulb are looking to you to help them out of administration :grinning: :grinning:

1 Like

For the record I’m NOT Elon Musk.