The Retail Note - Retail – outperforming the economy (again)

Key Messages

• 2025 retail sales values grew +3.4% YoY
• Retail sales volumes were up +1.8% YoY
• Both significantly higher than likely GDP growth of +1.4%/+1.5%
• Strong growth in both Grocery and Non-Food
• Grocery values +3.8%, volumes +0.4%
• 1^st year of positive volume growth since 2021
• Non-food values +2.9%, volumes +2.6%
• Discretionary categories (e.g. books, furniture, jewellery) outperform non-discretionary
• Online spend grows +4.3% YoY
• Online penetration increases marginally to 27.4%
• Multi-channel (+5.2%) outperforms pure-play (+3.3%)
• UK consumer continues to defy economic logic.

A cool £537bn. That was how much the UK retail economy was worth in 2025. Actually, that figure refers purely to retail sales. Adding in any number of multiplier effects (e.g. job creation – direct/indirect, import/exports, manufacturing, farming, supply chain stimulation, taxation etc etc), retail’s contribution to the UK economy is worth that figure many times over.

In terms of perspective, £537bn is higher than the annual GDP of countries such as Norway, the Philippines or Vietnam. If, purely hypothetically, the UK retail market were a country unto itself, it would rank 30^th in the world.

OK, I’m being slightly disingenuous. That £537bn includes fuel. Excluding fuel, annual retail sales last year were a measly £416bn. That’s still higher than the annual GDP of Pakistan and the Czech Republic and not far shy of Romania, South Africa and Hong Kong. And it would rank a lowly 40^th in world economies on this basis.

Remarkable annual growth?

Big absolute numbers, but we all know that retail is highly challenged, the sick man of the UK economy. We know that because we are told it on an almost daily basis. And if you say something often enough, it becomes engrained as truth.

The actual numbers present a very different story, yet are seldom given any airtime. Last year, retail sales values (exc fuel) grew by +3.4% on 2024. This figure was in line with the long-term (30-year) average of +3.4% and only very marginally below the short-term (10-year average) of +3.6%, which included all the vagaries of COVID.

Ah yes, but those figures are flattered by inflation I hear every economist the length and breadth point out. Well, actually retail sales volumes also grew last year by +1.8%. This marked the first year of volume growth since the post-COVID bounce back of 2021. In terms of context (word of the day), this is above the short-term average (+1.0%) and not far off the long-term average (+2.0%).

Anyone else noticing that all the above figures are positive, not negative?

The golden rule of forecasting: never revisit an historic forecast. But I genuinely can’t recall any economist or finance house predicting that retail sales performance would be anything like this strong in 2025. I may be wrong, but I can’t recall a single one that even predicted positive volume growth. I could name and shame, but some had penciled in volume declines of -2.5%. A predicted slump of -2.5% vs actual growth of +1.8%. Hmmm. There’s getting it wrong, and there’s getting it hopelessly wrong…

Food vs Non-Food

Ah yes, but that growth must have been entirely driven by food as that is non-discretionary and demand is inelastic. As the cost-of-living crisis rages on, discretionary spend must come under pressure and consumers must cut back on non-food, non-essentail purchases.

Wrong. That may have rung true on occasions in the past (actually, not as many as you’d think), but it certainly wasn’t the case in 2025, with both food and non-food delivering the goods.

Grocery sales values grew by +3.8% in 2025, above both short-term and long-term averages (+3.4%, +3.6%). More positively still, volumes grew +0.4% in 2025, the first year of positive growth since 2021 (short-term average flat, long-term average +1.1%).

Non-food sales values grew by +2.9%, while volumes were ahead +2.6%. In both cases, this compares favourably with short-term (vals. +2.6%, vols. +0.7%) and long-term (vals. +2.7%, vols. 2.3%) averages. Obviously, there were considerable variances between non-food sub-categories and there is not a universal picture of health across the high street, but as headline figures go, these are more than decent.

The stand-out non-food category last year was books (+25.3%), while demand was also strong for jewellery (+5.4%), furniture (+4.9%) – hardly non-discretionary categories – clothing (+4.6%) and garden centres/petstores (+4.0%). At the other end of the performance spectrum, demand was weakest for chemists (-9.2%, a non-discretionary category), cosmetics (-9.2%) and carpets (-3.3%).

In essence, discretionary spend categories generally outperforming non-discretionary ones, the reverse of the economic rulebook.

For a more detailed breakdown, please refer to accompanying Retail Sales Dashboard.

Seasonality

Seasonality is a way of life in retailing, an occupational hazard for retailers themselves. But something economists struggle to comprehend and are at pains to ‘smooth out’.

In terms of broad trends across 2025, there were two months when retail sales performance wobbled horribly – January (vals. +0.4%, vols. -1.1%) and May (vals. +0.3%, vols. -1.1%). Thankfully, these setbacks proved very temporary and underlined one of my mantras of it being short-sighted to read too much into one month’s figures. Aside from these two blip months, the picture was of generally strong and accelerating growth, particularly in the second half of the year. Q3 vals. were up +4.2% YoY (vols. +2.4%), Q4 values were up +4.8% YoY (vols. +3.0%).

Compare this with the ONS narrative and that adopted by the economist community as a result. The narrative is seemingly constantly of month-on-month or quarter-on-quarter volume declines, punctuated by the odd fluke month of positive growth. But these numbers are mercilessly ‘seasonally-adjusted’ to present a view that the ONS thinks is palatable to the world at large – a picture of ‘underlying’ demand that expunges inconveniences such as Christmas, Easter, Bank Holidays, Back to School periods. Inconveniences that are actually very important to retailers themselves. And inconveniences that somehow have to find their way back into the annual YoY figures.

The raw Pounds Data from the ONS lays bare the limitations of seasonal adjustment. It also perfectly reflects the seasonality of retailing – the raw figures do not lie. In 2025 as in every year, December saw the highest retail sales values by a country mile (11.5% of the 2025 total), as everyone would (or rather should) expect. Perhaps more surprisingly, June and September are the next highest spend months (both 9.2% of the annual total), followed by March (9.0%). Perhaps astonishingly given all the hype around Black Friday, November is only the 5^th highest spend month of the year (8.7%). February is always the lowest spend month (6.9%).

Spend values may fluctuate, but this seasonal pattern never changes. Canutian efforts to adjust these patterns are futile, especially when year-on-year comparisons obviate the need to even do so.

Online – at an inflexion point?

All retail sales growth being fueled exclusively by online – another chestnut that is increasingly open to question.

All online sales grew by +4.3% in 2025. The interesting sub-text to this is that multi-channel (i.e. store-based retailers with online arms) grew significantly faster than online pure-plays (e.g. Amazon, Shein, ASOS etc), at +5.2% versus +3.3%. As a result, multi-channel’s share of online increased to 52.9% vs. 47.1% for online pure-plays. This is a gap that is only going to widen going forward.

Overall annual online penetration increased modestly to 27.4% in 2025, from 27.1% in 2024. The interesting thing about this is that it is about where online penetration would have been, if very hypothetically COVID hadn’t happened and online had simply continued growing at its pre-COVID trajectory. In other words, now that the dust has firmly settled, it is increasingly fair to conclude that COVID didn’t massively and permanently turn retail towards online.

Again, most forecasts on this, be they made pre- or post-COVID, now seem woefully wide of the mark. Most had online penetration at 30%+ by now, some even more ridiculously still at 50%+. I would boldly predict that annual online penetration is unlikely to hit 30% this decade. Or 50% ever, or certainly not in any of our lifetimes.

Applying the multi-channel metrics to the penetration figure, around 83% of all retail sales involve a store in some capacity. A figure that is likely to be a constant, rather than shrink.

Make no mistake, online is still a growth channel, but not exponentially so. In many ways, it is growing in line with wider retail sales and as with other channels, is subject to seasonal and other fluctuations. Rather than an omnipotent force, it is just part of the retail furniture. Part and parcel of retailing, rather than wanton destroyer of the high street.

Retail sales vs GDP

Retail the sick man of the UK economy? We haven’t got the full-year GDP figures as yet, but the consensus is that last year we are likely to have seen growth of +1.4% to +1.5%. Either way this is below the +1.8% volume growth achieved in retail. Retail again outperformed the wider economy. As it so often has in the past.

It is widely accepted that retail sales will slavishly follow trends in the wider economy. After all, retail is one of the most conspicuous parts of the economy, part of eveyone’s everyday life, something that everyone can understand and relate to. And something that it is therefore easy (but often dangerous) to apply basic economic logic to.

But retail sales have a perennial habit of de-coupling from the wider economy. One for the statisticians but hopefully makes the point. Correlating GDP with retail sales values over the full ONS timeseries (1989 – 2025) gives an R-squared of just 0.193 (when 1.0 = total correlation, -1.0 = total inverse correlation). The same exercise with retail sales volumes should give a better correlation in that the comparisons are purer (‘real’ growth, net of inflation). It doesn’t, the R-squared coming in at just 0.08. In essence, both measures are pretty close to zero correlation.

Why? The eternal, million-dollar, unanswerable question. The only feasible explanation is philosophical rather than empirical - consumers are not robots, they are emotional beings. They will do what they want to do and are not bound by any economic rule book. Consumers are even like willful children – they’ll do what they want, not what they are told. Or if they are expressly told not to do something, take gleeful delight in doing it anyway.

Final thoughts

Why my apparent obsession with retail sales? In very simple terms, they are the single most important macro-economic metric and KPI in retail. There are no end of other economic and consumer measures, but this is the most accurate one in measuring consumer spend on our high streets, shopping centres, retail parks and online. Consumer confidence may be a barometer of how people are feeling, but retail sales are the measure of how they are actually behaving. Footfall merely a loose barometer of how much people are out-and-about, not of how much they are spending.

Other economic measures and national / global newsflow may ultimately affect and feed into retail sales performance. But it is a dangerous to assume that everything always will, just because the economic textbook says it should. The 2025 retail sales figures are further proof of that.

Are retail sales the sole arbiter of high street health?  Absolutely not, they are just a broad indicator. Consumer demand is and always will be the lifeblood of retail, but obviously there are a plethora of other factors in play, not least operational and competitive issues and pressures. Even if the consumer is spending freely, this may not necessarily filter through to strong high street performance, the correlation is not watertight.

But healthy consumer demand is always a good starting point and those operational and competitive pressures bite even harder without it. Contrary to all expectation, 2025 certainly delivered on that front. To the tune of £537bn, even.