5 Customer Map Signals for Pricing, Flexibility and Retention Teams

1. Meter, meter on the wall, what's the fittest tariff of them all?

Somewhere in a basement there’s an off-white plastic box with numbers blinking.

It clocked the exact Tuesday the heat pump went in, the week the teenager discovered gaming, Midsummer when the whole family vanished and left for the countryside, leaving the rest of us without a cottage connection to swelter in the city. It knows which house on the street whose owner swears it’s just a server is really mining crypto in the attic. It has seen more of your customers than their GP.

Unlike a magic mirror, though, it won’t talk back to tell you which tariff is best and for who.

Spain put 11 million households on time-of-use pricing during the energy crisis and found people used less at peak but didn’t move much overall. Italy found that people shifted, used more overall, and sometimes made the peak worse. Norway found that a tariff will cut the peak or move it but rarely both, depending on how you design it. And none of that is your portfolio. 

Customer Map runs the tariff against your portfolio, so pricing teams can know how a new one will affect it before launch, including how long until people start shifting their load, how much to expect, which assets do the moving, and what it does to your churn.

It also separates people changing their habits from machines being scheduled only once. 

2. The Great English Boil-Off 

It's the 4th of July, 1990, and somewhere in a control room a man is watching a dial.

Well, a wall of dials, and a chart recorder scratching a line onto a roll of paper, because it's 1990 and that's the technology. His job is keeping energy demand and supply in balance. Tonight the whole country is holding one breath watching the same thing, and he can see the line on the paper has dropped. This is England versus West Germany, the semi-final of the World Cup, and it’s gone to penalties, the part of football that exists to hurt people. 

It’s down to Chris Waddle to score and keep England in. He takes a long run-up and hits it, and the ball climbs over the crossbar up into the Turin floodlights.

The man in the dial room opens up the generators that have idled all match, then waits the three seconds it takes for grief to travel from a stadium in Italy to a kitchen in Sunderland. 

Across the country, a nation that has just been let down stands up and goes to the kitchen, because the only correct English response to national grief is to put the kettle on. 1.2 to 1.5 million of them. 

36 years later, the kettle has company: EVs, heat pumps, PVs, and batteries. All sitting on a smart meter that’s logging what happens. 

This makes national events the closest things you get to real demand response calls without the expense of running a programme. You also get a natural control group as some markets will watch their team crash out early.

3. No country for old EV data

Buying an EV is the most open a customer will ever be to changing how they use energy and also the best moment to move them onto a smart tariff, before charging habits are set.

Trouble is, you find out about the EV the way you find out about anything your customer does: a little too late. Energy customers aren’t known for being eager to share good news. So the number sits in the CRM, too low and behind as the market moves faster than any signup form.

We read that continuously off the meter to see the flow: new EVs arriving, EVs going quiet, cars sold on. You then get your adoption rate tracked as it happens.

4. Curb your capacity

A lot of solar is invisible in the official numbers because smart meters capture what crosses into the grid but not what a house makes and uses on the spot. 

SolarPower Europe estimates about a third of the EU's solar output isn't captured in the statistics, mostly rooftop systems that were not fully registered or properly measured. That's an issue for the grid operator, who can't see it coming, and a headache for anyone who has to reconcile the books.

We reconstruct the full picture with machine learning by using the meter readings together with local weather and sunshine data. The model works backwards to fill in how much the panels produced and how much the house consumed itself. Once you add those back in, you can also work out the size of the installation.

The same method can flag unregistered batteries adding to more flexible storage.

5. The 3 stages of flexibility 

When we think of flexible assets we tend to lump everything together into one big beautiful number that’s supposed to tell you how much demand you can shift off peak when the grid needs it. But owning an EV doesn’t mean you can move when it charges, the same way counting all the cars in a car park doesn’t tell you how many could theoretically drive to Portugal tomorrow. 

The number you want has to go through a few stages first.

First: availability. Owning the asset isn’t the same as being allowed to touch it, and being allowed to touch it isn't the same as there being anything to touch. 

Owning the asset means nothing until the customer has enrolled and agreed to let you steer it, and enrollment dies by a thousand clicks: review terms, authorise data, confirm email, drop off at every step. Then the car has to be plugged in during the window you care about. Everyone running a program already does it, more or less, so I'll just say the number that remains after this is your starting point.

Second: Rebound. You have to model what the portfolio looks like after you've moved it. Steer every enrolled EV out of the 6pm peak and into the cheap window at 2am and you've built a shiny brand new one at 2am. 

Third: Benchmark. Say 30% of your EVs charge at peak, is that terrible or totally normal? Looking at yourself only won’t help you, because it needs market-wide knowledge.