Take a look at two customers on your book today. They sit in the same profile class, pay the same flat tariff, and on every report you run they look identical. Now fast-forward to mid-2027, when Market-Wide Half Hourly Settlement (MHHS) is fully in place and your portfolio is being settled against actual half-hourly consumption rather than industry profiles. One of those customers, it turns out, has a deeply evening-peaking shape that costs you significantly more to serve than their tariff reflects, and they have been quietly eroding your margin every weekday for years. The other is the opposite: their consumption sits comfortably outside the expensive part of the day, they have effectively been overpaying for the shape they have, and as soon as a competitor prices them more accurately, they are going to leave. You will be able to see both of them clearly, for the first time. The question I want to spend some time on is what you decide to do about that.

This is the conversation Tom Goswell from Cornwall Insight and I had on a recent webinar, MHHS Unpacked, looking at MHHS and what it really means for retailers once you get past the migration timelines. Tom set out the regulatory picture very well, and I am not going to repeat that here. What I want to talk about is what happens commercially on the other side of all of this.

If MHHS is still new to you: MHHS is the shift from settling consumption against industry-defined profiles to settling against actual half-hourly data, for every meter. Migration started in October 2025, all suppliers need to be qualified to take on new customers on a half-hourly basis by October 2026, and the full cutover lands in July 2027. Ofgem and DESNZ have made it clear there will be no further delays. I think most retailers are now operationally focused on hitting those dates, and rightly so. But the regulatory programme is only the gateway. 

Here is the way I would put it. Profile classes have always been a useful fiction, and the fiction worked because settlement was profile-based. The model and the underlying reality were the same thing. If you priced a Profile Class 1 customer against the PC1 profile, and you were also settled against the PC1 profile, then in a sense you could not be wrong about that customer, even when you were quite wrong about that customer. The actual differences between the real human beings inside the profile got socialised across your book, smoothed out and absorbed. They existed, but they did not show up in your numbers. This is the visibility problem we talk about a lot at Gorilla. Most of it has been hidden, for years, by the settlement model itself.

Once you settle against actual half-hourly data, that protection goes away. The two customers I described at the start are not a hypothetical. They are a feature of every retailer's book, multiplied by however many meters you have. Some of them are profitable on the tariff they are on, some of them are not, and post-MHHS you will know which is which. That is a very good thing, ultimately. But it does mean two specific failure modes start to appear that were not really visible before. The first is margin erosion: customers whose actual consumption shape costs more to serve than their price reflects, where the gap shows up directly on your P&L. The second is churn risk in the other direction: customers paying more than they should for the shape they have, who will leave the moment a competitor prices them more accurately. Both of these things are happening in your book today. MHHS doesn’t create them, it just makes them visible.

What that means for how a retailer actually operates is, I think, the part of this that needs the most thought. Point-in-time pricing followed by an annual renewal cycle is not really enough anymore. The economic profile of a customer can drift through the year based on how their consumption actually behaves, and if you are not watching for that drift, you are paying for it. What you end up moving towards is something that looks more like continuous portfolio management: tracking margin in-life rather than only at the point of sale, pricing acquisition and renewals based on real behaviour rather than against a profile class, and paying close attention to which of your high-margin customers might be at risk. There is a first-mover argument too. The retailers who get to the high-margin customers first will pull them onto their books while everyone else is still finishing migration plans.

So what does the toolkit for any of this look like in practice? Some of it is already in place today. Pricebooks become considerably more useful in a post-MHHS world, because the constraints that limited them, essentially profile class and region combinations, fall away. You can start using variables that map more directly to behaviour: LLFID, market segment, connection type, business type. Time-of-use design becomes a free parameter rather than something constrained by Standard Settlement Configurations, which means you can build genuinely tailored products for EV drivers, heat pump customers, demand response participants, or whoever else you want to design for. Pass-through tariffs become a more viable product for harder-to-forecast customers, because you are transferring shape risk to the consumer in exchange for a price that better reflects their actual situation. And alongside all of that, we have been building out billing determinants, because, frankly, there is no point being able to design clever new products if your billing engine cannot actually invoice them at the end of the month. That has been a real constraint historically, and it is one we are actively working on.

Other parts of the toolkit are further out, and I want to be honest about that. Behaviour-based clustering as a real replacement for profile classes, behaviour-driven cross-sell into heat pumps and EVs, in-life margin detection that flags drift early, properly targeted retention for high-margin customers at risk: this is where I think the operating model for retail is heading, and it is where we are investing as a product team, but it is a direction rather than a finished story. The retailers I talk to are at very different stages of being ready for any of it, and that is fine. The point I would encourage anyone to take away from the webinar, and from this piece, is simply that the commercial reality of MHHS is worth taking seriously alongside the compliance one, and that the gap between those two conversations is where the real money will be made and lost over the next few years.

I genuinely believe the future of energy retail is margin-led, and that is a big part of why we moved to describing what we do as energy margin intelligence rather than something narrower. MHHS, more than anything else happening in the UK market right now, is the moment when margin becomes visible at the meter, and that visibility is the thing that lets you do everything else: better products, better retention, better forecasting, a healthier portfolio overall. If you want to dig into any of this, the full webinar with Tom is here, or you can come and talk to us directly.

‍