Chapter 02

Market and competitors

The wedge is vacant at the intersection and crowded on both flanks, and the flanks are closing. That sentence is the whole chapter, and it is a downgrade from what we believed a week ago.

What matters

1,144,530 UK homes changed hands in the 2025-26 financial year. That is the trigger volume, and it is published.S11
The buyer is a small firm in a shrinking, fragmenting trade. Over 80% of installer businesses are sole traders and 60% have considered leaving the sector.S14 That is who we are asking for £29 a month.
Nobody joins the render to the letterbox. Four axes were tested against 2025-26 UK evidence. No player holds more than two, and none holds the two that matter together.S21
Both flanks shipped product in November 2025. Zoopla and Rightmove both launched AI restyling of live listing photographs, and Rightmove additionally shipped a floorplan-derived, kitchen-and-bathroom-specific renovation cost estimator.S22
There is a clock. In July 2026 Hg put over £200m into Street Group, whose Spectre product is AI-targeted direct mail to UK homeowners. It mails sellers, not buyers, so it is not a competitor today.S25

2.1 The trigger, and why it is the whole business

Everything Housewarmly does hangs off one event: somebody buys a house. HMRC publishes the count monthly, and it is the most solid number in this guide.S11

Financial year	England	Scotland	Wales	NI	UK total
2021-22	1,166,510	109,080	62,310	32,470	1,370,370
2022-23	1,034,020	102,140	53,520	27,260	1,216,940
2023-24	840,030	93,130	44,060	23,470	1,000,690
2024-25	1,035,310	101,750	50,270	27,380	1,214,710
2025-26	959,900	105,220	53,540	25,870	1,144,530

Read the volatility, not just the level. The market fell 27% from 2021-22 to 2023-24 and has come back most of the way. A business whose entire input is house sales inherits that swing. Nothing in this guide should assume a stable 1.14 million.

The upstream funnel is larger and earlier. TwentyCi recorded over 1.7 million new instructions and over 1.25 million sale-agreed in 2025.S29 Sale-agreed is the interesting trigger because it is weeks before completion, before the buyer has chosen anybody, and about 9% larger than the completion count. It is also the point at which the listing is still live on the portal, which matters for the product.

Two figures that circulate and must not go in a deck. TwentyCi's "homemovers spend £29 billion annually" and "3 times more predictable than a non-mover", and Royal Mail's "£7bn on home improvements in the first year", are unsourced marketing claims made by companies selling mover data.S29 They establish that a priced market exists. They are not statistics, and no model here uses them.

2.2 The prize, at the other end

£17,500

Median UK kitchen renovation spend, Houzz, up 34% year on yearS12

£7,000

Median UK primary bathroom project. See the warning belowS12

£5.4bn

End-client spending on fitted kitchen products, 2025S13

1.0m

Kitchens installed in 2025, the lowest annual total since 2013S13

The shape of that last pair is the commercial story. Fewer kitchens, each worth more, so the fight per job is harder and the value of winning one is higher. The luxury segment now takes £2.15bn, which is 35.6% of the market's value from 7.35% of its projects.S13 And the consumer-facing half of the trade invoiced almost £3bn against just over £1.1bn trade-facing, while supplying kitchens in nearly equal volumes.S13 Our buyer sits on the consumer-facing side of that gap, which is where the money is.

Two numbers in this section are weaker than they look, and both are used in sales conversations. The £7,000 bathroom median comes from Houzz fieldwork carried out in November and December 2023 with a sample of 228 self-selected platform users, and Houzz has since stopped printing bathroom spend figures.S12 Given kitchen medians moved 34% in a single year, a 2023 bathroom median is probably materially low. And all Houzz respondents are users of a renovation platform, so the medians sit above the true population medians. Quote them with the date attached.

Do not use the "£10bn UK KBB market" figure. It circulates everywhere. It traces to an undated, unattributed standing panel on a trade blog with no base year and no source.S13 The kitchen figure of about £5.4bn is genuinely good, recent and clearly defined as end-client price. There is no verifiable public figure for a whole-KBB total: the bathroom segment's best public number is manufacturer sales from 2021, and the real numbers sit behind paid reports. If total-KBB sizing ever becomes load-bearing, buy the report rather than repeating folklore.

The trigger works, and there is finally a UK source for it

The claim that movers renovate has been asserted in this business for weeks on vendor marketing. There is a properly documented UK survey behind it.

44%

Of UK homebuyers renovate within the first year of moving inS46

64%

Renovate within the first two years. 71% where the home is olderS46

31%

Of all UK renovations are driven by a recent home purchase, up from 27%S46

The first two come from a nationally sampled Opinium survey of 1,000 UK adults, fieldwork 23 to 28 January 2025, published by Zoopla with the methodology printed.S46 The third is from the 2026 Houzz and Home study and is the better number for sizing the mover-driven share of the renovation market, with the 27% to 31% move supporting a rising-share argument.S46 Historically, Houzz also found recent buyers spent close to double the national median at £30,000.S46

One trap. A separate study reports that "44% of homebuyers discovered they needed costly repairs within the first year". The match with Zoopla's 44% is a coincidence and it measures remedial repair rather than discretionary renovation.S46 Do not present them as corroborating each other. And the "84% of movers renovate within the first year" figure that circulates on marketing blogs is American.

2.3 The buyer population, honestly counted

Here is the first place where the guide has to be careful. There is no authoritative census of UK kitchen and bathroom installers. Nothing published enumerates them.S18 What exists is a good proxy and two partial views.

The proxy: ONS enterprise counts by SIC code and size band

From "UK business: activity, size and location: 2025", published 24 September 2025, table 4, cross-checked against Nomis.S19 Counts are control-rounded to base 5.

SIC class	0 to 4	5 to 9	10 to 19	20 to 49	Total
43.22 plumbing, heat and air-conditioning installation	40,140	3,860	1,460	720	46,450
43.32 joinery installation	24,730	2,440	830	340	28,435
43.39 other building completion and finishing	20,865	1,715	630	170	23,430
47.59 retail of furniture, lighting and household articles	6,555	1,600	590	215	9,060
All four	92,290	9,615	3,510	1,445	107,375

Read this as a ceiling, not a market. These SIC classes are much broader than kitchen and bathroom fitting: 43.22 includes every plumber and heating engineer in the country. The register also covers only VAT- or PAYE-registered enterprises, so sole traders below the threshold are excluded entirely.S19 What the table does establish soundly is the shape: 86% of these businesses have four employees or fewer, and only about 13,100 firms across all four classes sit in the 5-to-19 band that this product is built for. The addressable population is a fraction of that, and nobody has published which fraction.

The two partial views, which agree with the shape

Over 3,000 installer businesses responded to a single BiKBBI survey in 2021, of whom over 80% were sole traders, only 9% thought there was a good supply of qualified installers, and 60% would consider changing sector.S14 A survey response count is a floor on the population, not a measure of it.
KBSA's retail member directory enumerates to roughly 205 studios, counted by hand. That is a quality subset, not a census.S18

State the consequence, do not paper over it. We can say the shape of the population with an ONS source. We cannot say "there are N thousand target customers" with one, because no published figure separates kitchen and bathroom fitting from plumbing and joinery generally. The addressable slice is the minority who employ fitters and run a showroom or a design service, and the 52 firms in chapter 05 were found one at a time because there is no list to buy. If a partner or investor asks for the count, the honest answer is that the closest published figure is 13,100 firms in the 5-to-19 band across four over-broad SIC classes, and that the real number is smaller and unpublished.

Why the sole-trader dominance is a pricing fact, not a trivia fact. An 80%-sole-trader population, shrinking in volume, with 60% considering leaving, does not buy £1,000 to £2,000 a year CAD seats on twelve-month non-cancellable contracts.S20 It does plausibly buy £29 a month it can stop. The price in chapter 03 follows from this paragraph.

2.4 The four axes, and who holds them

A competitor only matters if it holds several of these at once. This is the test the competitor sweep was built around, and it was written as an attempt to prove the wedge was already taken.S21

The buyer's actual room

AI-renders the specific room, not a template and not generic staging.

From a listing

Sourced from Rightmove or Zoopla, so the customer uploads nothing.

Installer-branded

The output carries a named local trade, who gets the enquiry.

Printed and posted

It lands on a doormat, not in an inbox.

Player	A	B	C	D	What it actually is
Housewarmly	yes	yes	yes	yes	The only one holding all four. That is the entire position.
Roomvo (Leap Tools), via ZooplaS21	yes	yes	no	no	Buyers restyle live listings with real products, paint and flooring. Branded party is the manufacturer, not the fitter. 7,000+ retailers, product-placement revenue. No cabinetry, no bathroom suites, no print.
Rightmove renovation suiteS22	part	yes	no	no	"Style with AI" restyles listing photos. The Renovation Cost Estimator reads the floorplan for real room dimensions and costs kitchens and bathrooms at three finish levels. Beneficiary is the agent. No trade connected to any of it.
RenoLeadsS23	no	no	yes	yes	New-homeowner data by postcode at £25 a month, direct mail add-on at £1.50 a letter with trade-specific templates. Our go-to-market, minus the product.
NEST DataS24	no	holds	yes	yes	29.3m properties, 200+ attributes, a "Property Images" database, an automated "Just Moved In" mailing campaign and home improvement named as a target sector. One render away.
Street Group / SpectreS25	no	agent feed	no	yes	AI propensity-to-sell model driving prospecting mail on behalf of estate agents, to prospective sellers. Same machinery, opposite direction.
MatchenS26	upload	no	match	no	The closest UK startup by shape. Consumer uploads a photo of their own kitchen, gets a render, is matched with three local specialists. Renders are Matchen-branded, not installer-branded.
TailoredQuote, Driveway VisualiserS27	upload	no	yes	no	Trade-branded AI before-and-after inside a quote, from £29 to £35 a month. Proof the render-for-trades model sells in the UK, and that it stops at the quote.
Convertly AI, Keystone AIS28	no	no	yes	no	Sold as "AI for kitchen fitters", from £249 a month, one trade per postcode. Turns out to be lead reply and follow-up. No visualisation at all.

The finding, stated exactly. The market has split cleanly in two and the halves have not met. The render half (A and B) is held by the portals. The delivery half (C and D) is held by unglamorous data and mail bureaux selling generic creative at £1.00 to £1.50 a letter. No UK company found renders a specific property's kitchen or bathroom and posts it to that property's new owner under a local installer's name.S21

The caveats on that finding, which belong beside it every time it is used. The sweep was done on 9 September 2026 by one researcher in one day. Several primary domains bot-block automated fetching, including Rightmove listing pages, Zoopla, Howdens, Checkatrade and Roomvo's main site, so parts of it rest on trade press rather than the source. Spectre's mail creative could not be seen, so whether its letters carry a picture of the property is undetermined. RenoLeads publishes no company number or founding date and could be a one-person data reseller. Houzz Pro UK is bot-blocked and whether it now does photo-to-render is unconfirmed.S21

2.5 Threat ranking, and what would actually hurt

Rightmove

It owns the photographs, the floorplans, the audience and the agent relationship. It has already shipped floorplan-derived room dimensions with a kitchen and bathroom, three-tier renovation spec, which is the hardest part of our stack to copy, plus AI restyling of listing photos.S22 It faces none of the copyright and terms-of-use exposure in chapter 10, because it is the counterparty. It needs only to decide that trades are an advertiser class.

What holds it back: it monetises agents, not trades; it is defending a £1.5bn Competition Appeal Tribunal class action over agent fees, which makes any new agent-facing charge awkward; and it has shown no interest in print or in the buyer after completion.S22

Our answer: be too small to be worth the product decision, and be somewhere they are not, which is the doormat

Roomvo / Leap Tools

Already in the UK on Zoopla, already has the product-placement revenue model that answers "who pays for the render", already sells to 7,000 retailers.S21 Extending from paint and flooring to cabinetry, and from portal placement to a named local fitter, is a product decision rather than a data one.

Our answer: cabinetry and sanitaryware are a harder render than paint, and the fitter relationship is a sales motion they do not run

NEST Data or RenoLeads adding a render

Cheap, obvious, and they already have the trigger, the trade customer and the printing press. NEST Data already holds a property images database and runs an automated "Just Moved In" campaign with home improvement named as a sector.S24 Neither shows any sign of it.

Our answer: get to their customers first, and be better at the render than a bureau will bother to be

Street Group / Spectre

Best capitalised, best machined, and pointed the other way. It would need a new customer segment, not a new feature.S25

Our answer: none needed yet, but know it cold, because every serious partner will raise it

2.6 The uniform negatives, which are as useful as the threats

Four whole categories were checked and came back empty. Each absence is a small piece of evidence that the wedge is real.

Agent-side martech

ValPal, Yomdel, Sprift, Nurtur, Iceberg Digital, Homeflow, the Reapit AppMarket and the Alto marketplace: no AI room-restyling or renovation-render app anywhere, and nothing that markets to the buyer after completion.S21 A genuine hole in the UK agent-tech stack.

KBB retailers and suppliers

Every "visualiser" is a template configurator, not the customer's room. Wren's Unity planner, Howdens' 2020 Ideal Spaces, Wickes, B&Q TradePoint Spaces, Magnet, IKEA, Crown Imperial's Kitchen Explorer. KBBG's mybathroomspecialist.com has no visualiser at all.S21

Virtual staging

Nine vendors checked. Only BoxBrownie and Styldod ship a dated-room-to-modernised product, both sold to agents and photographers. None of the nine offers print or mail fulfilment and none targets trades as a buyer.S21

Trade lead marketplaces

Checkatrade's one AI product of 2025-26 is TradeMore, a free job-management app. MyBuilder, Bark and Rated People ship nothing beyond lead matching. Houzz Pro is a presentation tool used after the lead exists.S21

2.7 What the trade currently pays for a lead, from audited accounts

The per-lead prices that circulate for Checkatrade and MyBuilder are SEO affiliate content, not published rate cards. MyBuilder publishes no rate card at all. Do not use them. The audited numbers are more useful anyway.S17

Company	Turnover	Operating profit	Margin	Note
MyBuilder Ltd, FY2025	£25.0m	£5.67m	22.7%	"Turnover represents shortlist fees", so 100% lead revenue
Vetted Ltd (Checkatrade), FY2024	£105.9m	£2.67m	2.5%	50,824 members. Published rate card: Free, Approved £30/month, Growth from £59/month, twelve-month fixed terms

Two readings, both useful. UK tradespeople pay roughly £131m a year for shared leads, so the willingness to pay is large and established. And Checkatrade turns £105.9m of it into £2.7m of operating profit, so selling shared leads to sole traders is a hard business. We are not trying to be Checkatrade. We are trying to sell one installer an unshared, house-specific, pre-drawn approach for less than the monthly fee they already pay for shared ones.S17

2.8 What the wedge is worth, and what it is not worth

Two sizings, both labelled.

Line	Value	Basis
UK completions per year	1,144,530	published HMRC, FY2025-26S11
Share with a usable kitchen or bathroom photo and a floorplan	65%	assumption untested at volume; new-builds and probate sales are photo-poor
Packs the market could absorb	~744,000	derived from the two rows above
At £5 a pack, gross market for the pack itself	~£3.7m/yr	derived, and this is the honest ceiling of the SaaS as currently priced

Read that last row properly. A pack business priced at £5 that captured every single UK house sale would turn over under £4m. That is not a venture-scale market and pretending otherwise would be the exact kind of bad assumption this guide exists to avoid. The scale case is not more packs at £5. It is (a) a supplier or manufacturer paying for the packs on the installer's behalf, which chapter 07 covers, or (b) the same machinery at US volumes and US price points, which is documented but is option value, not a plan.S45

2.9 Where we lose

Against a fitter who is already busy. The trade has a skills shortage, not a work shortage. An installer booked six months out does not want more enquiries, and "I'm turning work away" is the most common honest objection. Qualify these out early rather than argue.
Against the free design at the depot. Howdens gives every trade customer a free surveyed 3D design in every depot.S30 If the conversation drifts to design, we have lost it. Keep it on marketing.
Against a portal that decides to do it. If Rightmove connects a trade advertiser to its renovation estimator, our A and B become their by-products. We would still hold C and D. Plan for the flank, not the frontal assault.
Against "the photo is not mine". Chapter 11. This is the objection that ends deals with anyone who has a legal function.

← The product, in the buyer's words Pricing and packaging →