Stats SA published June's CPI on 22 July. Headline inflation came in at 5.0%, up from 4.5% in May and the highest reading since June 2024. That is the number that ran in every summary.
Underneath it:
Food at 1.6%. Transport at 12.7%. Within transport, diesel up 50.8% and petrol up 31.7% over twelve months.
The headline is an average of a household's whole basket. You do not buy a household's basket. Reading 5.0% and assuming your costs rose 5% is the most common mistake operators make with this release, and this month it is wrong in both directions at once.
Why food at 1.6% is not the relief it looks like
Two reasons to be careful before banking it.
CPI food is retail, not wholesale. It measures what a shopper pays in a supermarket for a basket weighted to household consumption. Your input mix is different, your volumes are different, and your supplier terms move on their own schedule. A low retail food number tells you the direction of travel, not your invoice.
Averages hide the lines you actually buy. June's largest annual increases were pork, tomatoes, sausages, corned meat and hake. The largest decreases were seasonal fruit, beetroot, white rice, potatoes and dried beans. A menu built on the first list and a menu built on the second had very different months inside the same 1.6%.
Where the squeeze actually is
Transport at 12.7%, with diesel up more than half in a year, reaches you three ways, and only one of them is obvious.
- Delivery. Every own-fleet drop and every aggregator fee sits downstream of diesel. Where you are absorbing delivery cost to hold a price point, that absorption got materially more expensive.
- Your suppliers' costs. Everything arrives on a truck. Fuel at these levels flows into supplier pricing on a lag, which is a reason to be sceptical that 1.6% food holds.
- Your customer's wallet, before they reach you. This is the one that gets missed. Petrol up 31.7% is money spent before anybody decides whether to eat out. Fuel is non-negotiable and takeaways are not, so a transport-led inflation print is a discretionary-spend problem wearing a cost-of-living label.
That third channel is why a low food number and a high headline is a worse combination than it looks. Your inputs are stable and your customer is poorer.
How to read the release in five minutes
It lands mid-month, monthly, free. A useful pass is short.
1. Headline, for context only. It tells you the story everyone else is reading. 2. Food and non-alcoholic beverages, as a directional check against your own invoices, not a substitute for them. 3. Transport, because it is the best available proxy for pressure on both your delivery cost and your customer's spare cash. 4. The largest annual increases and decreases, which name specific products. This is the most useful part of the release for a kitchen and it is routinely skipped.
The next release covers July. On the current pattern, the question worth holding is whether food stays near 1.6% while fuel runs at these levels, because those two do not usually stay far apart for long.
