June’s Growth Figure Is a Warning Against One-Dimensional Retail Forecasts

The UK economy grew by 0.4% in the second quarter of 2026, with June providing a useful reminder that consumer demand does not simply rise or fall. It moves.

Warm weather helped sales of specific products, including cooling equipment and summer clothing. Yet it also reduced trips to many physical retail destinations, disrupted travel and pushed a greater share of spending online. The latest official data put online sales at 29.4% of total retail spending in June, the highest proportion since April 2021.

For retailers, property operators and consumer researchers, that is more than a seasonal footnote. It exposes a weakness in many demand models: they still treat weather as background context, rather than a force that reallocates demand between categories, channels, locations and times of day.

Heat does not create a single retail outcome; it changes the route consumers take to fulfil a need.

---

A stronger sales number can conceal a weaker destination

June’s heatwave produced a familiar-looking retail story at headline level: certain discretionary and practical purchases rose. But the location data pointed in a different direction. High streets suffered as shoppers avoided uncomfortable journeys and crowded public transport, while retail parks were more resilient, helped by convenience-led trips for household and seasonal essentials.

This distinction matters because a national sales figure cannot tell a retailer whether a product was bought in store, collected from a retail park, delivered to the home or ordered late at night after the temperature fell. Nor can it reveal the operational cost of securing the sale.

A chain may report healthy sales while one part of its estate loses footfall, another experiences a surge in collection demand and its delivery capacity comes under pressure. A centre manager may see fewer visitors but a higher-value, more purposeful visit. A grocery or electrical retailer may find that the relevant question is not whether demand increased, but whether fans, chilled drinks, air-conditioning units and related accessories were in the right places before the heat arrived.

The commercial implication is straightforward: sales forecasting and destination forecasting should not be treated as the same discipline. They use overlapping evidence, but answer different questions.

Weather belongs in the operating model

The June maximum temperature record, subsequently verified at 38°C, is an extreme example rather than a normal trading day. But it should not be dismissed as an anomaly. The direction of travel is clear enough: businesses will increasingly face periods in which weather materially affects staff availability, travel patterns, dwell time, product demand, energy use and digital traffic.

The useful response is not to bolt a generic weather feed onto a dashboard. It is to identify the organisation’s own weather-sensitive decisions.

For a retailer, that could mean modelling demand by local temperature, humidity, transport disruption, day of week, school holiday status and product category. For a leisure venue, it might mean looking at the point at which an outdoor offer stops attracting visitors and starts deterring them. For a delivery business, it may mean forecasting both higher order volumes and a more constrained operating environment.

The model must also recognise that the relationship is non-linear. A warm day may increase footfall to a high street; an exceptionally hot day may reduce it. Rain may benefit an enclosed shopping centre but harm a retail park. The same temperature can produce different outcomes in a dense city centre, a coastal town and a car-dependent suburban catchment.

That requires local thresholds, not national averages.

---

Better research starts with demand transfer

Traditional post-event analysis often asks whether sales went up or down. That is necessary, but insufficient. The more revealing question is what consumers stopped doing in order to make the purchase they did.

June suggests several transfers worth measuring: from high street to online, from browsing to mission-led shopping, from daytime to evening purchasing, and from broad discretionary spend to products offering immediate comfort or utility. Those transfers are commercially important because they reshape conversion, basket composition, marketing effectiveness and labour requirements.

Research teams should therefore combine transactional data with footfall, fulfilment, search behaviour, local weather and qualitative customer evidence. The purpose is not to produce a more elaborate chart. It is to establish whether an apparent sales uplift represents new demand, accelerated demand or demand diverted from another channel, location or category.

That distinction affects investment decisions. If heat simply moves a sale online, the priority may be digital availability and fulfilment capacity. If it changes the preferred destination, parking, collection and local stock become more important. If it brings genuinely incremental category demand, the opportunity may justify broader ranging, supplier planning and campaign activity.

There is a further discipline here. Analysts should resist assigning every movement to weather. June also contained promotions, sporting events, changing household budgets and wider economic uncertainty. A credible model needs comparison periods, local controls and an honest treatment of uncertainty. Weather is influential; it is rarely the only explanation.

The forecast has to lead the decision

The value of a weather layer is realised before a heatwave, not in a retrospective report. Commercial teams should decide in advance which actions follow particular forecast conditions: stock reallocation, delivery-slot expansion, colleague welfare measures, revised store hours, localised creative, or a pause on campaigns designed to drive discretionary browsing.

That means connecting meteorological forecasts to operational triggers, with clear ownership. Merchandising cannot wait for a marketing report. Store operations cannot rely on a national forecast when the effect is local. Customer service needs to know when delivery promises or product availability are likely to change.

The second-quarter growth estimate is welcome evidence of consumer resilience. But its June detail contains a more practical lesson. In volatile conditions, demand does not disappear neatly into a single national measure. It is redistributed across the market.

Businesses that can see that redistribution early will make better calls on stock, service and spend. Those relying on one-dimensional forecasts will continue to mistake a change in route for a change in appetite.