The weekâs rate decision, Juneâs retail sales figures, and a new pay-per-sale route to customers for UK merchants, with our read on what each means for a founder-led brand.
The hold itself was expected. The movement inside the committee was not the comfortable kind: the argument gaining members is for a rise, not a cut.
The Monetary Policy Committee held Bank Rate at 3.75% at its meeting ending on 29 July, in a decision published the following day - the fifth hold this year. The vote was 6-3, with Megan Greene, Catherine L Mann and Huw Pill preferring an increase of 0.25 percentage points, to 4%. The Bank noted that CPI inflation has fallen to 2.6% but expects it to rise later in the year as the effects of higher energy prices pass through, and judges the risks to that path as tilted to the upside.
When we covered Juneâs inflation figures last week, the reading was that services and core inflation gave a cautious committee no clean case to cut, and that two members had voted for a rise as recently as June. That two is now three. Huw Pill joined Megan Greene and Catherine Mann in preferring 4%, against a 7-2 vote to hold in June, and the committee closed with the line that it âstands ready to act as necessary to ensure that CPI inflation remains on track to meet the 2% target in the medium termâ. Nothing in that combination makes a cut this autumn the base case.
For a founder the practical reading is last weekâs, made firmer. A plan that quietly assumes cheaper borrowing in the second half is leaning against the visible direction of the committee. The places to look are the ones where a rate assumption hides: stock built for peak season on borrowed money, a working capital facility coming up for renewal, capex financed on floating terms. Each is worth testing with todayâs rate held through year-end, and once more at a quarter point higher. If the plan only works in the version where rates fall, it is a position, not a plan.
If a rate assumption is baked into your forecast somewhere hard to see, a Discovery Call is a good place to surface it.
Source: Bank of England, Monetary Policy Summary and minutes, July 2026, 30 July 2026
June was a good month for retail as a whole. The number that should reach second-half planning is the channel split underneath it.
Retail sales volumes rose 1.0% in June, the ONS reported on 24 July, following growth of 1.2% in May, with warm weather and promotions credited for the lift; across the second quarter, volumes were up 0.6% on the first. The sharper detail is in the channel mix. Non-store retailers saw volumes rise 4.4% on the month, and onlineâs share of total retail spending rose from 28.9% in May to 29.4% in June, its highest since April 2021.
April 2021 is a telling comparison. The last time online held this large a share of the till, non-essential shops were only just reopening and the channel had something close to a captive audience. This time there is no lockdown doing the work. Shoppers with the whole high street available to them still put more than 29 pence in every retail pound through online channels. For DTC brands, most Baseline readers among them, that is demand moving toward your channel at exactly the point in the year when peak-season plans get locked.
The planning question is capacity and mix rather than celebration. A growing channel attracts spend as well as shoppers, so attention will cost more this peak, not less. And if Juneâs share is a new base rather than a heatwave artefact, stock cover, fulfilment capacity and acquisition budgets set on last yearâs channel split are calibrated to a smaller online market than the one arriving. The time to test that is now, while there is still a quarter of runway before the season starts.
Phoenix can help you turn a channel-mix assumption into a stock and working capital plan for peak.
Source: ONS, Retail sales, Great Britain: June 2026, 24 July 2026
When a channel charges only on the sale, acquisition spend starts to behave like a cost of sale rather than a bet. That changes the question a founder needs to be able to answer.
Shopify has expanded Shop Campaigns beyond the US and Canada to new markets including the UK, in a rollout reported on 29 July. The tool lets eligible merchants run a single campaign across the Shop app and third-party channels including Meta, Google, X, Snap and Pinterest, with budgets and billing in local currency. Merchants set the most they are willing to pay for a new customer, and are charged only when a qualifying sale completes.
The reach is not the interesting part; the shape of the cost is. Conventional paid acquisition is spend-first: the money goes out, and the cost of a customer is discovered afterwards. A pay-per-sale model inverts that. The cost attaches to the order rather than to the attempt, which caps the downside of a campaign that does not convert and makes acquisition spend look, unusually for most channel mixes, like a variable cost of sale. For a founder-led brand where cash discipline matters more than reach, that is a different financial instrument, not just another place to advertise.
Known is not the same as cheap. A channel that fixes your cost per sale prices that certainty in, and the honest comparison is contribution margin per order, by channel, after the platformâs share. What counts as a sale matters too: a model that pays to win a customer is worth different money depending on whether that customer comes back. The discipline before switching anything on is knowing what a new customer is worth to you, first order and repeat, so a guaranteed price can be judged rather than accepted.
Worth a conversation if your channel decisions currently run on revenue rather than contribution.
Sources: Practical Ecommerce, New Ecommerce Tools: July 29, 2026; Retail Times, Shopify expands Shop Campaigns to new markets including the UK; Shopify Help Center, Understanding Shop Campaigns
After three weeks in the high mountains of Europe, a UK classic with a different discipline at its centre. On an unmarked course, the result belongs to the runners who never lose track of their position.
The Montane Lakeland 100 ran from Coniston over the weekend of 24 to 26 July: a course branded 100 miles that measures closer to 105, with around 6,300 metres of climbing, no course markings, and navigation by map and route description, much of it in the dark. Mark Darbyshire won the menâs race in 18:51:31 and Anna Troup the womenâs in 25:39:33, while the accompanying Lakeland 50 went to Ry Webb in 7:44:29 and Katie Kaars Sijpesteijn in 8:23:21.
The event is built around its fourteen staffed checkpoints, and the organisers are blunt about the attrition: 40 to 50 per cent of the 100-mile field does not finish in a typical year, and most of those who drop are gone by halfway. On ground like that, the racing skill is unglamorous. It is knowing where you are, all the way round: position against the cut-offs, food against the miles remaining, minutes spent standing still in checkpoints. The runners who finish are rarely the ones who moved fastest between any two points; they are the ones who never stopped knowing the state of the whole.
A trading year makes the same demand. Plenty of founder-led businesses run it as an unmarked course, with the true cash position and the real margin by product discovered at year-end, the way a lost runner discovers a wrong valley: too late to do anything about the route. The alternative is checkpoint discipline. A small set of numbers, arriving reliably and often enough that correcting course is still cheap when the need shows up. Phoenix builds that visibility with founders: the handful of measures that tell you where you are, while it still matters.
Sources: Open Tracking, Lakeland 100 2026 results (official event results); Ultramarathon Database, Lakeland 100 2026 results
A free 30-minute Discovery Call is the place to start. No obligation, no sales pressure.