Analysis of Octopus Agile electricity prices

Guy Lipman
6 min readSep 1, 2019


In Great Britain, while most domestic electricity customers are on a standard variable tariff or a fixed tariff (usually for 12 months), for the past 18 months Octopus have offered a floating tariff. Customers on Octopus’s floating tariff pay for whatever electricity they use at the prevailing half-hourly price (they therefore have to use a smart meter to take this option). This leaves the customer exposed to the risk that prices will be higher than expected, but gives them flexibility to pay less by shifting consumption to lower-price times.

There are a number of things to think that might influence whether you are better to go with a floating or fixed tariff (Please note, this blog post aims to give some things to think about. It won’t cover everything, so please don’t take it as advice for your particular situation). If your budget can’t handle any price uncertainty, you should probably go with fixed. There are a lot more deals in the fixed market, and they can get extremely competitive (often unsustainably so), so at times the best fixed deals will be less than the expected floating prices.

However, I believe the two biggest factors will be when you tend to use electricity, and how easily you can shift consumption. Fixed price tariffs tend to assume you’ll use a standard profile, which looks like:

(Note that this standard profile is calculated by Elexon, and is used to estimate the profile for domestic customers that don’t have smart meters.)

As a result, if you know that you tend to use more electricity at times when prices are lowest (eg overnight, summer) and not when they are highest (4–7pm, winter), then you should be aware that a fixed price is probably going to overcharge you. I’m very much in this category, with an electric hot water tank and gas heating, and those with electric vehicles will be too.

To help these customers who use more electricity at night, retailers offer deals like economy7 or Octopus Go, in which customers are given a cheaper rate for an overnight period (7 hours for economy7, and 4 hours for Octopus Go) and a higher rate for the rest of the day. These require a smart meter or a special economy7 meter. However, these deals won’t give you the same benefit as a floating tariff if you use less electricity than they expect in winter, or you have flexibility, for example being happy to charge your car more on windier nights (when prices are cheapest).

So, the first question I was keen to answer was how the average floating price varies, and how much it depends on the profile. I used Octopus Agile prices for London, including VAT, which are reported in pence per kWh. I calculated the average price for each month and year, both weighted by number of hours and by standard domestic profile weight. Each month is grouped by calendar days in local time (ie March is 743 hours and October 745 hours). This shows that the profile-shaped prices are about 10% higher than baseload (ie unweighted average) prices (which was less than I expected):

I next calculated what the equivalent prices would be for Octopus Go (which runs from 00:30–04:30 each night) and Economy7 (I used 23:30–06:30, but it can vary between suppliers). I also included calculations for the SuperPeak (16:00–19:00) and NonSuperPeak periods.

Again, I was surprised that the offpeak prices were as close to the peak prices, given how large the variance between offpeak and peak is with some of the tariffs on the market (for example, Octopus Go is 5p for offpeak, 14.7p for peak). It may well be the case that suppliers are letting peak use subsidise offpeak use, and if you’re actually going to use most of your use during offpeak, you’d be better off with one of these deals than floating. It is also clear how much more expensive the SuperPeak period is.

Considering Flexibility

As mentioned above, another benefit of floating tariffs is that customers can take advantage of flexibility, that is, the ability to reschedule consumption to cheaper times of the day. Obviously most customers will have at least some consumption that they won’t want to shift, for example, you’re not going to cook your dinner at 3am just because the prices are lower. But for some things like water heating, electric vehicle charging and running the washing machine, many customers will have at least some willingness to shift.

I decided to run experiments across the historical Octopus Agile prices:

  • The cheapest half hour of each day tends to be 4:30am. What was the average price at 4:30, and how much less would you have paid if you were willing to shift to the actual cheapest price each day?
  • There is some consumption that you con only do from 9am–4pm (I live in a flat, so couldn’t run my washing machine at 2am!). The cheapest half hour in this window is 3pm. What was the average price then, and how much less would you have paid if you were willing to shift to the actual cheapest price between 9am and 4pm?
  • There is some consumption that you have to do between 5:30 and 7pm (perhaps cooking dinner). The cheapest half hour in this window is 5:30pm. What was the average price then, and how much less would you have paid if you were willing to shift to the actual cheapest price between 5:30 and 7pm?
  • The cheapest half hour of the week tends to be 4:30am on a Sunday. What was the average price then, and how much less would you have paid if you were willing to shift to the actual cheapest half hour of each week?

I did the calculations, and got the following results:

It seems to me from these results that there isn’t currently an enormous amount of value from flexibility within a day. Being willing to shift from Sunday 4:30 to other times of the week does offer more value, but I should note that this analysis assumes you could predict in advance which day would be cheapest — it isn’t clear how true this would be.

Just because there isn’t currently a huge amount of value from flexibility, doesn’t mean there won’t be in future. For example, if we increase the amount of wind power, and reduce the amount of on-demand generation, we would be likely to get more variation between half hours. Also, the prices that I have used for this analysis are day ahead prices — there would be likely to be more value if customers could offer flexibility in real time. As a result, I believe it is definitely worth keeping an eye on these numbers going forward.



Guy Lipman

Fascinated by what makes societies and markets work, especially in sustainable energy. Views not necessarily reflect those of my employer.