Prefer to listen?
This article is also available as a podcast episode
The construction market did not slow down in 2026. It split in two, and the average figure everyone quotes conceals which half you are standing in.
If your workload is privately financed and housing-led, you have had a genuinely bad year and the forecasts do not promise you a good 2027. If your workload touches infrastructure, energy, water, education or public-sector refurbishment, you have had a decent year and the pipeline is strengthening. Both statements are true simultaneously, and the headline output figure — down 3.3% this year, up 1.2% next — describes neither of them.
For a small practice or contractor, the question worth asking before the year end is not “is the market recovering.” It is “which market am I in, and can I get into the other one.”
What actually happened
The proximate causes were external. A Gulf conflict spiked oil and industrial energy prices in the spring, which fed straight into double-digit construction product inflation, mortgage rates and client confidence. On top of that the UK imposed 50% import tariffs on steel from 1 July 2026, with the quota cut by around sixty per cent. Almost every downgrade published this year traces back to one or both.
The output data:
- CPA summer forecast (27 July 2026): total construction output -3.3% in 2026, +1.2% in 2027 with risks to the downside. Downgraded from -2.5% in the spring.
- Private housing new build: -10.0% in 2026, flat in 2027.
- Private housing repair and maintenance: -8.0%, flat in 2027.
- Infrastructure: +3.2% in both years — the only reliably positive sector in the forecast.
- ONS: new work has now contracted for nine consecutive months. New orders fell 11.8% quarter on quarter.
- PMI: 38.2 in May, a six-year low. 44.7 in July — the best in four months, and still firmly in contraction.
The sector split, which is the part that matters
Two datasets tell the story better than the headline.
Glenigan’s index of construction starts to end June 2026, year on year:
| Sector | Year on year |
|---|---|
| Offices | +8% (and +51% quarter on quarter) |
| Education | +7% (and +17% quarter on quarter) |
| Retail | -9% |
| Social housing | -18% |
| Health | -28% |
| Industrial | -29% |
| Utilities | -36% |
| Hotel & leisure | -42% |
| Infrastructure | -51% |
| Private housing | -63% |
| Overall | -38% |
RICS Q2 2026 workloads, net balance:
| Sub-sector | Net balance |
|---|---|
| Energy | +39% |
| Water & sewage | +23% |
| Communications | +22% |
| Rail | +10% |
| Roads | +7% |
| Other public works | +9% |
| Public housing | +1% |
| Private commercial | -7% |
| Private industrial | -9% |
| Private housing | -12% |
The two datasets disagree about infrastructure because they measure different things — Glenigan counts project starts, which are lumpy and were depressed by a few very large schemes not starting, while RICS measures surveyor-reported workload, which captures the design and enabling work that runs ahead of a start. Read together, the honest reading is: infrastructure and utilities work exists and is being designed, but it is converting to site slowly.
Which is itself the defining problem of 2026. As BCIS put it, “there is reported difficulty converting the pipeline into activity.” Sixty-seven per cent of RICS respondents cited financial constraints as the main obstacle. Barbour ABI recorded planning applications down 56% in a single month, and warned that a sharp deterioration in applications raises real concerns about the pipeline entering next year.
Planning consent is no longer the bottleneck. Finance and viability are.
What this means for a housing-dependent practice
If most of your fee income comes from private residential, the strategic question is uncomfortable but simple: the sector you serve is forecast flat at best through 2027, after a year down ten per cent.
The reflexive answer is to chase infrastructure work. That is right in direction and usually wrong in execution, because it is attempted as business development — new logos, new frameworks, new relationships — which has a two to three year lag that a firm under cash pressure does not have.
The faster route is capability and evidence. Public-sector, regulated-asset and infrastructure clients buy differently from private developers. They buy against a pre-qualification questionnaire, a framework, and a demonstrable process. What gets a small firm through that door is not a relationship, it is:
- Demonstrable information management to ISO 19650, because the public sector procures to the UK BIM Framework under the Information Management Mandate and asks you to evidence it.
- A competence framework you can show — named individuals mapped to the published standards, not a capability statement.
- Framework readiness — the administrative groundwork done before the opportunity appears. Central Digital Platform registration under the Procurement Act takes minutes and is required at award, not at bid; doing it cold when you have won something is friction you do not need.
- Payment data literacy. The Procurement Act now publishes authorities’ payment performance. Local government spends 34% with SMEs and rising; central government manages 10%, worse than 2020. Chase the pipeline where the money actually goes.
None of that is a three-year programme. It is a quarter’s work, and most of it is documentation you should have anyway.
Where the specific opportunities are
Fit-out and refurbishment is the most SME-accessible growth market in the whole picture: a £9.85bn UK interior refurbishment and fit-out market, up 4%, split roughly commercial offices 25%, education 23%, leisure 20%, healthcare 12%, retail 11%. Short-cycle, design-intensive, and increasingly requiring proper information management because of energy performance evidence and, on larger buildings, building safety compliance.
Education is one of only two sectors positive on every measure, with starts forecast up 8% this year and 20% next, and further education described as particularly strong.
Retrofit carries the most reliable policy-backed pipeline available to a small firm — £15bn, five million homes by 2030 — but note the contradiction with the CPA’s -8% on private housing repair and maintenance. Both are true: discretionary owner-occupier work is collapsing while grant-funded and social-landlord retrofit grows. Different clients, different procurement, and PAS 2035 roles are the gate.
Consented but unbuilt schemes are a market in themselves. Build-to-rent starts fell 79% year on year, and 84% outside London, while more schemes than before secured planning permission. That leaves a large stock of consented schemes with a sunk consent and no route to site — which is a value engineering and redesign brief, not a construction one.
Data centres are real — a pipeline approaching £100bn — but concentrated with tier one contractors and specialist MEP. The accessible layer is design coordination, not main contracting.
The thing to be careful about
A discipline worth applying to every pipeline announcement you read this year: a spending target is weaker than a funded programme; a funded programme is weaker than a procurement notice; and a framework ceiling is permission to spend, not committed spend.
The New Hospital Programme is the cautionary example. £60bn total estimated cost, and final hospitals now opening in 2045-46 rather than 2030, with some schemes delayed nine to ten years. Anyone who built a business plan on new hospital construction in 2023 has had a difficult three years. NHS backlog maintenance, decant and enabling works are the realistic near-term market there — not new build.
AMP8 is the same shape with a different timeline: £104bn committed, but effectively started a year late, with projects due by 2030 slipping to 2032 and beyond. The risk that creates is back-loading — five years of work compressed into three — which turns delivery and design throughput into the binding constraint rather than funding. That is a 2027 and 2028 story worth positioning for now.
The market has not stopped. It has moved, and it has moved toward clients who buy on evidence rather than on relationships. That is genuinely good news for a small firm with its documentation in order, and genuinely bad news for one without.
If you are looking at a housing-weighted workload and want to know what it would actually take to be credible on public-sector and framework work — we do that assessment, and we do the capability evidence that follows it.