Last updated: 30 September 2026
England’s housebuilders face a growing number of questions about solar, batteries, grid capacity, capital and resident handover. Supplier pages can explain an offer, but they do not show which names answer systems actually surface when a buyer asks about those decisions. This report supplies a small, current view of that gap.
On 30 September 2026, we measured 396 answers across 45 English-language questions about housebuilder decisions in England, using ChatGPT, Google AI Overview and Gemini. Seven names in a ten-brand tracked sample appeared at least once. Every rate in this report uses the same 396-answer denominator. The finding is deliberately modest: even the two most frequently named providers appeared in only 24 answers, or 6.1%. The category has no automatic answer because the offers solve different parts of the development decision.
TL;DR
- 396 answers were measured across 45 questions and three answer systems.
- Octopus Energy and Gryd were each named in 24 answers (6.1%).
- UPOWA was named in 21 answers (5.3%).
- E.ON Next and Sunsave each appeared in 7 answers (1.8%); SNRG appeared in 6 (1.5%).
- The tracked sample includes seven names that appeared at least once; it is not a complete market list.
What was measured
We record what the systems return to practical buyer questions rather than treating a provider’s own claims as an answer. “Named answers” means an answer includes a provider name. “Mentions” counts every occurrence. “Question coverage” is the number of distinct questions in which the name appeared. The public measurement receipt contains the complete methodology, distribution and retrieval table.
This is a current, bounded answer study. It does not measure consumer demand, sales, market share, technical eligibility or customer satisfaction. A name appearing in an answer is a research observation, not an endorsement.
The providers named in the answers
| Group | Provider | Named answers | Rate | Mentions | Question coverage |
|---|---|---|---|---|---|
| Funded plot-level provision | Gryd | 24 of 396 | 6.1% | 75 | 8 of 45 |
| Funded whole-home service | Wondrwall | 4 of 396 | 1.0% | 17 | 2 of 45 |
| Estate infrastructure | SNRG | 6 of 396 | 1.5% | 23 | 3 of 45 |
| Tariff and specification | Octopus Energy | 24 of 396 | 6.1% | 32 | 7 of 45 |
| Tariff and specification | E.ON Next | 7 of 396 | 1.8% | 12 | 3 of 45 |
| Installation delivery | UPOWA | 21 of 396 | 5.3% | 39 | 4 of 45 |
| Consumer-adjacent service | Sunsave | 7 of 396 | 1.8% | 9 | 4 of 45 |
This is not a league table. The grouped rows are the result: a tariff proposition, an installation partner, a communal infrastructure provider and a funded per-home service can all be relevant to a development without being interchangeable purchases.
The joint headline is a useful example. Octopus Energy and Gryd each appeared in 24 answers, but Gryd was mentioned 75 times against Octopus’s 32 and appeared across eight questions rather than seven. That does not prove one is stronger. It shows that the systems use the names in different contexts.
The stronger conclusion is the ceiling: each joint leader appeared in only 24 of 396 answers. A housebuilder should not use a generic AI recommendation as a substitute for deciding whether the site needs a tariff offer, shared infrastructure, a delivery partner or a plot-level asset-and-service model.
Four commercial routes that should not be blended
Funded plot-level systems
A plot-level model is designed around each home. The provider funds or owns solar and battery hardware, the developer coordinates installation, and the resident has a continuing service, subscription or purchase relationship. Gryd’s developer page says it funds panels, battery, hybrid inverter and monitoring hardware while the developer installs through its normal build programme. Its modelled 55-home analysis describes the split between hardware and developer-paid labour. It is company-published evidence, so it should be read as an explanation of the offer rather than independent market validation.
Wondrwall is the closest whole-home comparison. Its housebuilder page describes a funded system with solar PV and battery storage; its homeowner page describes PowrPlan installation, maintenance and service support. Its public terms should not be assumed to match any other provider’s. Check charges, term, transferability, ownership and buyout terms for the actual site.
Estate infrastructure
SNRG answers a site-wide question. Its smart-grid explanation describes rooftop solar with a central communal battery and private-wire arrangement. Its Longbridge case study illustrates that estate-level route. It can be relevant where grid capacity, shared assets and site resilience dominate, but it creates a different asset and resident relationship from a self-contained battery at each plot.
Tariff and specification
Octopus Zero Bills is an accreditation and tariff proposition for qualifying homes. E.ON Next’s developer service covers an integrated utility route. These offers are relevant when a developer is evaluating a home specification with a buyer-facing energy promise. They should not automatically be read as provider-funded plot-level hardware. Ask who buys the assets, who owns them, who bills the resident, and what happens if the home is sold.
Installation delivery and consumer routes
UPOWA provides design, installation, compliance documentation, handover and aftercare for developer projects. That is a valuable delivery route where the builder retains equipment procurement, but it is a different commercial choice from a provider-funded resident service. Sunsave is consumer-adjacent context: familiar subscription language may appear in answers, yet it should be treated as adjacent unless a developer channel and site economics are verified.
Question coverage
The study gives equal question coverage to five buying decisions, with nine questions in each: England compliance and specification, procurement and project economics, plot design and build delivery, contracts and handover, and resident outcomes. The questions move from choosing an approach to checking whether a supplier can deliver it on a particular development.
For example, a commercial director asks what costs remain when hardware is funded. A technical team asks whether its appointed installer can work with a third-party funder. A sales or legal team asks how a roof-space lease will be assessed during a home sale. These are different buying jobs, even when each conversation begins with “solar for new builds”. The public measurement receipt publishes the complete question map and the observed answer count for each area.
Sources: broad retrieval, focused operating use
The receipt reports domain retrieval for the same 45 questions and measurement date. gov.uk was retrieved in 91 answers, gryd.energy in 51, LinkedIn in 31, YouTube in 25, Facebook in 18 and Reddit in 13. The underlying URL capture was paginated sequentially, so it is not an atomic end-to-end URL census. It also includes overseas, Scottish and consumer material; use it as a directional source footprint, not an English housebuilder reading list. The public measurement receipt holds the domain distribution.
| Type | Source | Use |
|---|---|---|
| Institutional | GOV.UK: rooftop solar for new builds | Regulation and policy context |
| Trade | Housebuilder: Future Homes Standard launched | Delivery and housebuilder framing |
| Professional | Future Homes Hub: delivery at scale | Grid, specification and implementation learning |
| Video learning | Future Homes Hub Part L resources | Technical guidance and webinars |
| Provider-owned | Official pages above | Terms to verify against the proposed contract |
Gryd illustrates the difference between being found and being named: its domain was retrieved in 51 answers, while the name appeared in 24. These are separate measures, and the source capture was sequential, so they are not a conversion funnel. They do suggest a useful research question for any provider: which parts of the website help an answer explain the subject, and which parts establish that the provider fits the buyer’s exact requirement?
Three operating takeaways follow. First, use policy and trade sources to understand the standard, but use the actual contract to understand asset ownership, maintenance and resident liability. Second, treat professional communities as a research queue: a credible story is a documented lesson from a completed site, not a generic sustainability claim. Third, check handover before selecting the system. Ask every provider who owns the hardware, bills the resident, maintains it, manages a sale, and carries the support obligation.
What this means for an England housebuilder
Start with the site boundary, not the most frequently named provider. A constrained grid may justify a shared estate model; model load, export and communal-asset governance first. A site seeking capital relief and a per-home proposition may justify comparing funded plot-level offers; separate hardware funding from installation labour and compare resident terms line by line. A site with an equipment budget but delivery risk should compare partners on design coordination, DNO process, commissioning and handover; require a plot-by-plot responsibility schedule before award.
For broader guidance on how AI systems use first-party and third-party evidence, see Schmitdy’s AI search resources. Product pages establish an offer. Clear resident documentation and independently checkable delivery evidence make the offer easier to assess.
A practical comparison checklist
The headline rates are a reason to ask better questions, rather than to shortlist a provider from a table. Before a development team compares prices, it should make the intended end state explicit. Does every home need its own battery, or does the site benefit from a shared asset? Is the primary objective Future Homes Standard compliance, lower resident bills, constrained grid capacity, EPC performance, a sales proposition, or several of these at once? A supplier can be right for one of those tasks and poorly suited to another.
For a funded per-home model, request a schedule that separates hardware, installation labour, roof works, electrical works, monitoring, insurance, maintenance, replacement and customer care. It should identify the asset owner at every stage and the conditions under which ownership transfers. A resident-facing agreement should be reviewed for charges, indexation, sale or remortgage treatment, buyout mechanics, complaint handling and the support path after a developer has left site.
For a communal model, request the private-wire boundary, metering arrangement, governance, battery dispatch rules, tariff calculation, grid-backup assumptions and responsibility for common assets. The technical design may be stronger at site level, but a buyer needs to understand what the resident receives and which decisions remain collective.
For a developer-paid installation route, ask how system design will be kept aligned with SAP, roof geometry, DNO requirements, commissioning and sales handover. An installation partner should be able to show who signs off each plot, which certificates reach the buyer and who responds when a system is not behaving as described. These are ordinary delivery controls, but they are especially important where solar, batteries, heat pumps and smart tariffs interact.
The answer study cannot choose between these routes. It can show that AI systems surface all of them under the same broad phrase. The practical response is to narrow the brief before comparing offers, then assess the evidence against that specific brief.
Limits
This is directional evidence, not a definitive market map. It can show which names appeared in these 396 answers and how the seven non-zero names compare against the same denominator. It cannot prove incremental revenue, technical performance, customer satisfaction, future recommendation rates or causation. It cannot tell one developer to chase every question or select the highest-named provider. Inspect the question and sources in the receipt, then repair the weakest evidence layer for the site.
Frequently asked questions
Which provider was named most often?
Octopus Energy and Gryd were each named in 24 of 396 answers, a 6.1% naming rate. This tie is not a market-leadership result.
Which other provider appeared most often?
UPOWA appeared in 21 of 396 answers, or 5.3%. It is an installation-delivery route, not the same commercial model as a funded plot-level system.
Does a retrieved source mean the provider was recommended?
No. A system can retrieve a page for background information without naming or recommending its provider in the answer.
Does one provider fit every development?
No. The joint leaders appeared in only 24 of 396 answers each, and the providers represent different ownership, delivery and resident-relationship models.
Sources
- Public measurement receipt: methodology, distribution and domain source footprint.
- Gryd: official product and delivery description.
- Wondrwall for housebuilders: official whole-home service proposition.
- SNRG smart grid: official communal infrastructure description.
- Octopus Zero Bills: official tariff/accreditation proposition.
- E.ON Next developer services: official developer utility offer.
- UPOWA developer services: official installation-delivery offer.
- Sunsave Plus: official consumer service.
- GOV.UK new-build solar announcement: policy context, not validation of the study.





