How the California Cash Home Buyer Index Is Scored

What this Index is

The California Cash Home Buyer Index is an annual, per-city study of the companies that advertise cash home purchases. For each company we capture primary public records on fixed dates and score them under published math. Inclusion criterion: companies actively marketing to sellers in the chapter's market at capture time, regardless of physical presence, as evidenced by their own published claims. Sacramento was the first chapter and remains the 2026 example throughout this page's worked math; the same criterion and the same math apply to every chapter that follows it. We do not survey companies, accept submissions, or take payment for placement. The author operates one of the scored companies; that conflict and how it is handled are documented at the bottom of this page.

Sacramento 2026 snapshot dates: DRE and Secretary of State records 2026-08-08, supplemented 2026-08-09; BBB 2026-08-08; Google 2026-08-08; company site claims 2026-08-08. Scores computed 2026-08-09. Each cell in the dataset names its capture date. Later chapters carry their own snapshot dates on their own pages.

The five axes and their exact math

Each axis is scored on a native scale, then weighted. Native-scale math first; weighting follows in the next section.

LICENSE VERIFICATION (native scale 30). Link points: 18 if the California DRE licenses a corporation in the brand's own name. 12 if an individual principal is licensed AND the person-to-brand link is confirmed (see identity standards below). 0 if no license is found or the link is unconfirmed. Display points, only where link points exist: 6 if the site displays the license number accurately with a DRE label; 3 if displayed with a defect (a bare number without a label, or a number formatted differently from the state record); 0 if not displayed. Clean-record points, only where link points exist: 6 if the record shows no disciplinary action, else 0. Absence of a license is a verifiability finding only. Buying property on one's own account does not require a DRE license under Business and Professions Code 10130-10131. AB 1850, which would require a license for wholesaling, is pending, not law, as of publication. What a license changes is recourse: licensed operators answer to the DRE.

ENTITY STATUS AND LONGEVITY (native scale 20). Scored only where a state registry record is confirmed to be the brand's operating entity. Status points: 12 for Active or Good Standing; 0 for a suspension, scored only when the suspension is verified on two independent registry endpoints AND the entity's link to the brand is confirmed. Longevity points by years from formation date to the scoring date: 10 or more years 8; 5 to 9.99 years 6; 2 to 4.99 years 4; under 2 years 2. Where a brand has more than one confirmed entity, the oldest sets longevity. A brand whose legal identity cannot be found after exhausting the Secretary of State index and the county fictitious business name index with a control-verified search scores 0, reported as verifiability, never as wrongdoing.

BBB RECORD (native scale 20). No BBB profile: 0 on the whole axis. Grade points: A+ 10, A 9, A- 8, B+ 7, B 6, B- 5, C+ 4, C 3, C- 2, D range 1, F 0, NR (no rating issued) 2. Complaint points from the 3-year complaint count, summed across identity-linked profiles: 0 complaints 6; 1 to 5 complaints 4; 6 to 20 complaints 2; 21 or more 0. Accreditation: 4 if accredited, else 0. Multi-profile rule (standing): where a brand has more than one identity-linked BBB profile, complaint counts sum, and the letter grade of the licenseNumber-linked primary profile governs, with secondary grades disclosed in the evidence notes.

REVIEW RECORD (native scale 15). Formula: 15 x (rating / 5) x min(1, log10(review_count + 1) / 2), rounded to one decimal. Only Google profiles identity-linked to the scored brand count. Where a brand has several identity-linked profiles, counts sum and the rating is the review-count-weighted average; this consolidation applies uniformly to every multi-profile brand. The logarithmic term reaches full weight at 99 or more reviews, so a perfect rating on a handful of reviews cannot outscore a strong rating on a large base. A brand with no identity-linked Google profile in the chapter's market scores 0: the axis deliberately measures presence and standing in the market the chapter covers.

CLAIMS CONSISTENCY (native scale 15). Start at 15. Subtract 3 per documented inconsistency cluster, floor 0. A cluster is one claimed quantity in conflict: either two statements by the same brand that cannot both be true as written, or a claim contradicted by a verified public record. Conflicts count per quantity, not per reference: three different figures for years in business is one cluster. A statement that is true as written but could be misread (a genuine contractor license displayed in a home-buying context) is reported in the narrative and not deducted. A defect already scored in another axis is not deducted twice. Every deduction is published with its evidence.

Weights: what was pre-registered, what was ruled, and why

We pre-registered one weighting, scored everything under it, then made one adjustment before publication and froze the result for the annual cycle. Both tables publish because the adjustment changed the author's own rank and readers should be able to check it.

Pre-registered weights: License 30, Entity 20, BBB 20, Reviews 15, Claims 15. Final ruled weights (FROZEN for the 2026 cycle): License 20, Entity 25, BBB 20, Reviews 15, Claims 20.

The rationale for the change, recorded at ruling time: wholesaling without a license is lawful in California, so licensure cannot dominate the rubric; the license axis remains as a consumer-recourse measure. Entity status (does the company legally exist, and in what standing) and claims accountability were weighted up in its place.

Weighted score = sum over axes of (native score x ruled weight / native scale). Concretely: License x 20/30, Entity x 25/20, BBB x 20/20, Reviews x 15/15, Claims x 20/15. On the 2026 Sacramento data every rescaled component lands exactly on one decimal, so no rounding enters the totals.

Both weight tables, final data, side by side

Both columns include the two pre-publication identity attributions (the Property Sales Group license link and the Cash for Keys operating entity; see identity standards below), so the only variable between the columns is the weights.

The weight change moved the author's company down one rank (fourth to fifth on final data; it had stood third before the two competitor attributions). It was ruled anyway. The weights are now frozen: no further tuning is available until the 2027 cycle, and any 2027 change will publish with the same side-by-side disclosure.

TIES: HomeVestors and Westbrook land at exactly 34.0. No tie-break rule was pre-registered, so none was invented after seeing the data; both publish as rank 11 and the next rank is 13. A tie-break rule will be pre-registered for 2027.

Evidence tiers: which axes are artifact-backed, and which are not

Not every axis in this Index rests on the same quality of evidence, and a reader is entitled to know which is which before deciding how much weight to give a score. For the 2026 cycle the five axes sit on two tiers, and the tier is the same in all fifteen chapters.

ARTIFACT-BACKED, three axes: the DRE licence record, the Secretary of State entity record, and each company's own published claims. For every scored cell on these axes we hold the retained raw page or the raw search result that produced it. That means the figure is reproducible: you can run the same lookup and compare, and if you get a different answer, ours is checkable against a file rather than against our memory of it.

SUMMARY-BACKED, two axes: the BBB file and the Google review record. These rest on the dated written record our capture session made at the time, not on a retained raw page. They are real observations, made and dated, and they are a weaker form of evidence than the three axes above, so we label them rather than let the difference pass unmentioned.

THE REASON IS A LIMIT IN OUR OWN TOOLING. It is not a statement about the BBB, not a statement about Google, and not a statement about any company in this Index. The BBB declines our automated requests while serving the identical pages normally to an ordinary browser, which is an entirely reasonable thing for a website to do. Google review counts are drawn onto the page by JavaScript after it loads, so the page we would save is the page before the numbers arrive. Between those two, the missing piece on our side is the ability to write a fully rendered browser page straight to disk. We have not built it. It is scheduled for the next annual cycle.

We considered publishing nothing on those two axes until the tooling exists, and rejected it. A BBB grade and a review record are two of the five things a seller actually looks at, and dropping them would make the Index less useful while making it look more rigorous. Reporting them at a named, lower tier is the more honest of the two options.

One consequence worth stating: where a chapter's licence, entity or claims data was re-captured during the 2026-08-27 evidence pass, its BBB and review figures were NOT re-captured and still carry their original capture date. Those dates are printed under each chapter's table. A number under this Index is only ever as fresh as the date beside it.

Identity-linking standards

Scores attach to identity, so identity has rules.

RECORD-CONFIRMED LINK: the default standard. An entity or license attributes to a brand when at least two independent public records link them, such as a displayed license verifying to a licensee whose registered office matches the brand's published address, or a site's displayed license number resolving to a person plus a BBB license field carrying the same number, or a fictitious business name filing tying the brand name to the entity. One record is a lead, not a link. This is the same standard applied to Joe Homebuyer's principal and to Cash for Keys CA's operating entity: both attributions rest on public records alone.

DISCLOSED-SOURCE LINK: author knowledge was used for exactly one attribution in this Index, the Property Sales Group license link, and it is labeled as such wherever it appears. Property Sales Group's licensed principal was linked to the brand by the author's direct industry knowledge, confirmed on the record, rather than by two public documents. We publish the basis so readers can discount it if they choose. Confirming this link moved a competitor above the author's own company.

CANDIDATE MEANS ZERO: a plausible but unconfirmed link scores zero on the affected axis, every time, for every company. Scoring a candidate as confirmed would be fake data. Before publication, two candidates moved to attributed on stated bases: Property Sales Group's license by the disclosed-source link above, and Cash for Keys' operating entity by the record-linkage standard (three independent public sources: the footer's displayed license numbers verifying to licensees, those licensees' DRE broker office address matching the brand's published address, and the matching Google profile address). Each basis is recorded.

NAME-TWINS ARE NOT LINKS: a registry entity sharing a brand's name is not that brand without linking evidence. Two FTB-suspended entities named "Cash for Keys" are unrelated name-twins of the Sacramento brand and are not attributed to it; an FTB-suspended California LLC named like Express Homebuyers is not attributed to the Virginia brand for the same reason. The one suspension this Index does attribute (Westbrook REI, Inc.) is attributed because the entity name matches the brand exactly AND the principal matches AND the suspension is confirmed on two independent registry endpoints.

FORMER NAMES: one entity with successive names is one entity. Realty Helpers LLC previously operated as Fortune Homes LLC; the chain is a name amendment, not two companies, and an unrelated terminated corporation with a similar name is not attributed to the brand.

ABSENT IS ABSENT: where a company's legal identity cannot be found after exhausting the Secretary of State and county fictitious-name indexes with control-verified searches, the Index reports exactly that. Absence from a registry is never published as inactivity, wrongdoing, or an accusation.

How to run every check yourself

1. DRE license: the DRE's public license lookup is free. Search the number a site displays and read whose name is on it; then search the company name and the owner's name. Check status, expiration, employing broker, and the disciplinary section.
2. Secretary of State: bizfileonline.sos.ca.gov, search the company name and any entity name the site or BBB discloses. Read status and standing, formation date, and the agent's name.
3. County fictitious business name: if no state entity turns up, the county FBN index maps trade names to owners. Sacramento County's index is searchable online; note that a company officed in another county files there instead.
4. BBB: read the profile's business details, not just the letter grade. The legal name, license fields, start date, and alternate names are self-reported but checkable against the registries above.
5. Claims: save a dated copy of any page whose promises you are relying on. Our dataset does exactly that, and our capture dates are published.

Direct links: California DRE public license lookup · California Secretary of State business search · Better Business Bureau · AB 1850 bill status (leginfo.legislature.ca.gov)

Limitations, stated plainly

1. BBB COMPLAINT COUNTS HAVE NO VOLUME DENOMINATOR. 71 complaints against a national operator doing enormous transaction volume is not comparable to 1 complaint against a small local buyer, and no public per-company transaction denominator exists to normalize them. The complaint brackets are therefore coarse, and the axis caps how much they can move a score.
2. THE CLAIMS AXIS MEASURES CONSISTENCY, NOT CANDOR. A site that publishes almost nothing checkable can score perfectly on claims, and one company in the 2026 set does. The license and entity axes are where opacity costs points; the claims axis only catches what a company actually says. Read a perfect claims score next to the identity columns, not alone.
3. NO TITLE-VERIFICATION AXIS IN v1. The strongest possible check, whether a company actually takes title to houses and how long it holds them, is not in this edition: county recorder verification was designed, partially piloted, and deferred. It is planned as a v1.1 update to the Sacramento chapter and as a standard axis in future cycles. Until then, the Index verifies who companies are and what they claim, not their deed history.
4. REVIEWS CAN BE GAMED. Google reviews can be bought. The axis carries the lowest weight, the log curve makes small fake batches nearly worthless, and counts publish so year-over-year spikes are visible. But a determined operator could reach full marks on this axis; that is why it is worth 15 points and not 30.
5. THE LOCAL-PROFILE RULE ZEROES NATIONAL OPERATORS on reviews. Opendoor, Offerpad, Express, and HomeVestors have review activity elsewhere; none has an identity-linked Sacramento Google profile, so all score 0 on the review axis. This is a deliberate choice: a Sacramento chapter measures Sacramento presence. It is disclosed here so nobody mistakes it for those brands having no reviews anywhere.
6. LONGEVITY BRACKETS ARE CLIFFS. An entity at 9.9 years scores 6 where 10.0 scores 8. On the 2026 data no company sits close enough to a bracket edge for this to move a rank, which we checked; brackets publish anyway so the cliff is visible.
7. SNAPSHOT SCORING. Scores reflect the capture dates. Companies change their sites; two statements deducted in this edition were corrected after capture, including by the author's own company. Corrections after the snapshot date do not change the edition's scores for anyone.

Conflict of interest

The author of this Index operates Fast Home Buyer California, one of the 14 scored companies. Handling: identical criteria and identical strictness; the company's row cites the same classes of evidence as every other row; the full dataset is downloadable so anyone can recompute its score. On the 2026 data the author's company took the most claims deductions of any licensed company. The author confirmed one competitor's license link from direct knowledge, a second competitor's operating entity was attributed on public records alone, and the author ruled a weight change against his company's strongest axis; together those moved his company from third to fifth. The methodology does not ask readers to trust that; it publishes the math so they can check it.

Corrections

Correction, August 9, 2026: an earlier version of the Sacramento chapter's disclosure stated the author had confirmed Cash for Keys CA's operating entity. That attribution rests on public records alone, as the company's profile describes; the author's direct-knowledge confirmation applies only to the Property Sales Group license link. Scores were not affected.