National press. Forty years of wells. And an entity the machines can't see.

The rarest authority in the field, sitting behind an invisible entity

Rob Thompson is a Santa Rosa, California third-generation water dowser — dowsing since 1978, more than four decades, former co-owner of one of the largest drilling companies in Northern California, a licensed engineering contractor, and the owner of Thompson Builders. He has earned the single hardest asset in his entire niche: genuine national press — The New York Times, CBS News, ABC7, The Press Democrat, NorCal Public Media, and Vice have all featured him.

Almost none of that authority is visible to Google, AI search engines, or a first-time visitor in a structured form. The site carries no business-identity schema — no Person, no LocalBusiness, no Service, no Review markup — so answer engines see no entity to cite. There is no Google Business Profile with reviews, zero reviews on Yelp and Birdeye, an inconsistent NAP (two phone numbers, two addresses across listings), and the two strongest proof points — named winery case studies with hard numbers — appear nowhere in machine-readable form.

[ Data verified ] Live fetch of robthompsondowsing.com (all six pages, robots, sitemaps, JSON-LD), Search Atlas keyword API, NYT/CBS/ABC7/Press Democrat/Vice, Birdeye — pulled June 16, 2026.

The seven questions, answered

A third-generation dowser with a driller's credibility

Verified: Thompson Well Location Inc, 2777 Yulupa Avenue, Santa Rosa, CA 95405. Rob is a third-generation dowser, "dowsing since 1978," with "over four decades" of experience and a claim of "over 4,000 individuals and companies" helped to find water. His real differentiator is not mystique — it's that he co-owned one of the largest drilling companies in Northern California for ten years, holds an engineering contractor license, and owns Thompson Builders. Method: two L-shaped stainless rods and a pendulum, GPS and Google Earth mapping, with remote/map dowsing offered. Service area: Northern California (Sonoma, Napa, Saint Helena, Knights Valley), travels anywhere.

The credibility wedge is the drilling background: it lets Rob speak the buyer's language — cost-per-foot, depth, dry-hole pain — instead of the mystic's. That is the single most under-leveraged asset on the site. What's missing is the structure around the story: the entity name resolves three different ways across the site (schema says "Thompson Well Location," titles say "Thompson Well Location Inc," the brand is "Rob Thompson Dowsing"), the membership claims are unverified, and the proof lives in prose rather than in anything a machine can read.

• Entity name should canonicalize to one form, with the others as alternateName, so AI engines stop disambiguating across three labels.Needs confirmation
• American Society of Dowsers membership is claimed on-site but could not be independently verified — confirm before featuring it as a credential.Needs confirmation
• The "4,000" claim vs an older "900 wells" figure should resolve to one defensible number used consistently everywhere.Needs confirmation

The money isn't in "dowsing" terms — it's right next door

Search volume in this category splits three ways: large informational demand (people researching the concept or buying rods), tiny local-service demand (a handful of "near me" searches a month), and a large adjacent pool — the well-drilling buyer journey Rob sits beside. The strategy follows the demand: roughly 20% direct dowsing terms (won through entity authority, not volume), 50% adjacent well-drilling content + partnerships, and 30% authority/AI-visibility so Rob is the named answer.

[ Data verified ] Search Atlas keyword API, June 16, 2026. Volumes are US monthly averages; difficulty 0–100.

What that demand is worth

Assumptions stated: dowsing pricing is hidden across the entire field, so average engagement value ≈ $1,000 is an assumption for modeling only — single-site visits run lower, multi-site vineyard/ranch jobs materially higher. Confirm Rob's actual fee Current inbound from search is assumed minimal given the entity and local gaps.

All figures directional, built only on the stated assumptions. The structural point holds regardless of the fee: the largest pool of buyers near Rob is already searching 18,000 times a month for where to drill — and he is the person who knows where.

A small field with no gatekeeper

The most important structural finding: the dowsing SERP is page-winnable, not directory-mediated. Searches for "water dowser / water witcher near me" do not surface Yelp or Thumbtack dowser listings — those return water-treatment contractors instead. Individual practitioner sites rank directly. There is no aggregator to fight, so entity authority plus schema plus content can take the page. The flip side: no aggregator funnels demand either — the practitioner is the funnel.

How the credible operators overcome the skepticism (worth copying): third-party validation (a named driller, not self-claims), quantified predictions (gallons-per-minute and depth versus actual), longevity and lineage, and borrowed scientific authority. One lane is wide open: not one reviewed competitor features a "no water, no fee" guarantee — and pricing is hidden across the entire field, so a risk-reversal offer and transparent terms would both differentiate.

The honest balance sheet

A clean foundation, missing its top layer

The site is a well-built WordPress + Divi foundation on managed hosting, with clean robots and sitemap, real keyworded titles, and genuine content across six pages. The gaps are the next layer up — the structured authority and conversion infrastructure — not the basics.

One item to confirm: the site runs GA4 G-3KT4C16M4V site-wide — the same legacy property ID seen elsewhere in the Bonsai portfolio. Verify it is correctly scoped to this client and not cross-counting traffic.Needs confirmation Net: the highest-impact technical task is the entity schema graph; the second is reconciling the NAP and conversion basics.

Indexed, but under-told

The link layer is healthy — clean robots, a valid sitemap, correct redirects, and keyworded titles. What holds it back is telling: per-page metadata is missing on /find-oil/ and /news/, the brand name fragments across title, schema, and OG, the heading hierarchy is broken (two H1s per page, eight on the News page), the money page (Dowsing Services) is the thinnest at ~358 words, and the genuine press the News page links to is never used as a citation or authority signal.

• Schema: add Person (Rob), LocalBusiness (NAP/geo/areaServed/priceRange), Service, FAQPage, and Review — the dominant gap.
• Metadata: add titles/descriptions to /find-oil/ and /news/; unify the entity name across title, schema, and OG.
• Content: deepen the Services page; turn the Info page's Q&A into FAQ-schema content; expand the case studies into rankable, citable pages.
• Headings: a single H1 per page; demote News article headlines to H2/H3.

Be the answer — today AI engines see no entity at all

Ask an AI engine "who is the best water dowser in Northern California?" and Rob should win — he has the press and the track record. Today he can't, because none of it is machine-readable. There is no Person entity for Rob, no LocalBusiness with an address or service area, no Service, no Review, and no FAQ on the page built exactly like an FAQ. The press (NYT, CBS, ABC7, Press Democrat) is not wired in as sameAs, and the entity name resolves three different ways.

Fix path: Person + LocalBusiness + Service + Review + FAQ schema, a sameAs graph to the press, YouTube, and Thompson Builders, an llms.txt, a consolidated authority page, and a reconciled NAP. AI Search Readiness moves from 2/10 toward 8/10 in a single phase.

The cheapest win on the board: a Google Business Profile

No Google Business Profile with reviews was found — for a location-based service in a defined geography with a real street address. That is the single cheapest, fastest win available. Alongside it, the NAP needs standardizing: two phone numbers and two addresses appear across citations (Chamber of Commerce, Birdeye, Yelp, hub.biz, licensed.contractors).

• Standardize to one canonical NAP: Thompson Well Location Inc · 2777 Yulupa Avenue, Santa Rosa, CA 95405 · one confirmed phone.Confirm primary number
• Reviews are an open field: 0 on Yelp and Birdeye, none on Google. Even the field leaders hold few reviews — parity is reachable inside a year.
• Capture is easy: winery and ranch references already exist in the case studies; a simple post-job review request turns them into structured proof.

Real-world authority that's never been wired in

The paradox of this account: the raw authority is exceptional, the structured authority is near zero. The fix is mostly aggregation and markup — making what already exists verifiable in one click.

A good form, standing alone

The contact form on every page is solid. But it's nearly the only path, and the trust signals that justify hiring a dowser live off the page. The gaps are quick to close.

Offer ladder: free phone consult → site visit (the core service) → multi-site/vineyard engagement → remote/map dowsing for distant land → the oil/exploration advisory lane for qualified landowners.

A real lane, an honest verdict

Rob already markets oil location on his /find-oil page, so the question isn't whether to enter — it's how hard to lean in. The honest answer: the lane is nearly empty of competitors and nearly empty of search demand, with severe trust barriers. Treat it as a deliberate, low-cost authority bet — not a near-term revenue engine.

The buyer is a small owner-operator or geologist at one of ~9,000 US independents, concentrated in Texas, Oklahoma, North Dakota, Wyoming, and New Mexico. The siting decision runs geologist → landman → owner — often the same few people — which makes this a personal-authority sale, exactly where Rob's drilling lineage helps. But the barriers are real: controlled tests, a documented oil-dowsing fraud legacy (the "doodlebug" era), and professional embarrassment for any geologist associated with it. The only credible pitch is "a cheap supplemental tiebreaker alongside seismic, with risk reversal" — never "trust me over the geophysics." Historical precedent gives one citation worth using: Paul Clement Brown, an MIT-trained electrical engineer who reportedly dowsed oil for Standard, Signal, Getty, and Mobil.

Wildcat Market Opportunity Score

Scored 1–10, higher = more favorable.

Three lanes, in sequence

Strong authority on a thin foundation — very fixable

The first five dimensions score current state (higher = stronger today). The last three score headroom (higher = more opportunity to capture). Scores are directional, built from the verified findings in §6–§12.

From invisible to cited and ranked

What to do first — ranked by impact

The first five moves

• Approve the entity schema build and grant access — WordPress admin, Google Business Profile, GA4.
• Reconcile the two phone numbers, pick the primary, and move email off sbcglobal.net.
• Create and verify the Google Business Profile at the Santa Rosa address; standardize the NAP.
• Publish the Jack Neal & Son and Delectus case studies with names (permission) and Review schema.
• Add the booking calendar and the "no water, no fee" guarantee to the services page.