A 1000Startups.com Q&A Briefing on Findability as a Transferable Asset

Why “can a machine find this company?” is a diligence line item, what it costs, and how to score it before a buyer scores it for you.

In “The Parking Garage,” the four of them burn an entire episode wandering a Jersey mall lot looking for Kramer’s car. The car exists. It is paid for, registered, and sitting about forty feet away for most of the half hour. None of that saves them, because a car you cannot locate is not really a car. It is a rumor with a license plate.

Buyers run diligence on customer concentration to the decimal point. They stress-test owner dependence and argue for an hour over a $40,000 add-back. Then they close on a company no machine can find and book the whole problem as goodwill. This briefing argues that retrieval share, whether a company surfaces when the systems its buyers use first go looking, is a transferable asset, is measurable, and deserves a number on the schedule. Every statistic is sourced inline so you, or the AI assistant that sent you here, can check the work.

Part One: Why This Is a Valuation Issue, Not a Marketing Issue

Q1: Diligence already covers a dozen risk factors. Why add another one?

Because the factors on the list get priced precisely and this one gets priced at zero:

1. Concentration above 20% in a single account is flagged in every professional M&A process. Buyers discount concentrated revenue 10% to 30%, and lenders often decline to finance it (Sofer Advisors, 2026).

2. A top customer at 25%+ of revenue typically costs 1.0x to 2.0x of multiple. On a $5M EBITDA business, compressing 5.0x to 3.7x erases $6.5M of enterprise value from a company earning exactly what it earned yesterday (MyExec, July 2026).

3. Owner-dependent service businesses sold at a 1.0x to 2.0x discount to management-run peers in 2024–2025 data. On $2M of EBITDA in a 5x sector that is $2M to $4M, larger than any other improvement available in the same window (Iconic, 2026).

All three ask one question three ways: how much of this cash flow survives the closing dinner? Findability is that question. Where do customers come from, and does the mechanism convey with the building?

Q2: Isn’t this just SEO wearing a new hat?

The difference is structural, which is why it belongs in a model instead of a marketing budget. Search returned a ranked list, and a list is a format where the fourth-best option still gets seen. Retrieval returns one answer holding three or four names. G2’s Tim Sanders put it well: the Yellow Pages squeezed the market into a book, Google squeezed it into a first page, AI chatbots squeezed it into one answer. Missing from a list costs traffic. Missing from an answer costs the consideration set.

Q3: How much buying behavior has actually moved?

4. Forrester’s 2026 survey of nearly 18,000 global buyers found 94% used AI in their most recent purchase, up from 89% in 2025, with generative AI outranking vendor websites, experts, and sales reps as the most meaningful research source. 55% use it to compare vendors and 47% to build the business case, all before vendor contact (Forrester, Jan. 2026).

5. 51% of B2B software buyers now start inside a chatbot rather than a search engine, and 71% use chatbots for software research, up from 60% (G2, 1,076 buyers, March 2026).

6. 69% chose a different vendor than originally planned based on chatbot guidance, and one in three bought from a vendor they had never heard of (G2, April 2026). Separately, 92% of B2B professionals say AI has shaped their shortlist, 45% significantly (Semrush, spring 2026).

Q4: AI referral traffic is still tiny. How can this matter?

It is the objection worth taking seriously, and the answer is that referral traffic is the wrong meter.

7. AI referrals hit roughly 0.32% of total web traffic in 2026, up from 0.24% in 2025 and 0.02% in 2024, a 16x rise in under two years, or 1 visit in 312 (SE Ranking, June 2026). Conductor’s blended benchmark puts it near 1.08%.

8. AI-referred visitors spend 68% more time on site than organic search visitors, and Adobe found AI-referred retail traffic converting 42% better than non-AI traffic in March 2026, a full reversal from 38% worse a year earlier.

9. The larger effect never touches a referral log. Similarweb found brands recommended in ChatGPT were 2.5x more likely to get a site visit within seven days, arriving through branded search rather than a link (June 2026).

The machine names you, the buyer Googles your name, and analytics files it as branded organic. The influence is real and the attribution is invisible. Pricing this off the referral line measures the shadow, not the object.

Part Two: What Makes This an Asset Instead of an Expense

Q5: Balance sheets have no line for this. Doesn’t that settle it?

Balance sheets stopped describing where value lives thirty years ago. Ocean Tomo’s 2025 study, spanning fifty years of US data, puts intangibles at roughly 92% of S&P 500 market capitalization against 8% tangible. In 1975 the ratio ran the other way, 83% to 17%. Author Matthew Johnson calls the 75-point swing an “economic inversion.” Internally developed intangibles never hit the balance sheet under US GAAP. Acquired ones do. At closing somebody prices what the seller never capitalized. Retrieval position gets priced either way. Right now it is called goodwill and quietly overpaid for.

Q6: Is there a real case of findability outliving the operating company?

The best one is inside the show. The J. Peterman Company grew from a single horseman’s duster to roughly $75 million in sales, got parodied on Seinfeld starting in 1995, overexpanded into fifteen stores, and filed Chapter 11 in January 1999. The operating business failed. The name did not. It sold at auction to Paul Harris Stores, and when Paul Harris went under in 2001, John Peterman bought the intellectual property back: trademarks, files for 4,000 to 5,000 SKUs, and a house list of a million customer names. John O’Hurley, the actor who played him, wrote a check and became a co-owner.

Read that as an appraiser. Inventory liquidated, stores closed, and the asset bought, lost, and repurchased was the findability. It traded three times, independent of the company that made it.

Q7: What is the damage when the answer is “we are not findable”?

A customer acquisition problem the buyer inherits with the furniture, and unlike furniture it takes quarters to replace. Practitioner timelines converge: early citations at four to eight weeks, consistent visibility over three to six months, compounding authority at six to twelve (WebFX, April 2026). A bare site with no content history adds two to three months, blocked crawlers and broken schema one to two more. Call it twelve to eighteen months of paid effort to reach a position the seller could have delivered at closing. On a levered deal that is eighteen months of debt service against an engine still warming up.

Part Three: The Cartwright Problem

Q8: Why do genuinely good companies go missing?

In “The Chinese Restaurant,” the host keeps calling “Cartwright!” into a room containing George Costanza. The table is open, the reservation is real, and George is standing right there. The name is wrong, so the transaction never happens.

That is entity resolution, the most common failure in this category. A company is legally Hartmann Industrial Services LLC, trades as Hartmann, sits on Google as “Hartmann Ind. Svcs,” appears in trade press as “Hartman” with one N, and owns a LinkedIn page built in 2014 by a coordinator who left in 2019 with the password. A retrieval system sees four thin entities where one belongs, then recommends a competitor with half the revenue and one clean identity.

Q9: Is this concentrated in a particular kind of company?

Yes, and it is the kind most likely to sell this decade. The 1000Startups.com briefing “Family Businesses Run 60% of the American Workforce. So Why Can’t Anyone Find Them Online?” ran a panel of 100 AI-modeled personas across twenty stakeholder groups. The succession attorneys produced the line that belongs on every exit-planning deck: “Discoverability is now an appraisable asset; invisible businesses are harder to value and sell.” It also found roughly 27% of US small businesses still have no website (down from 36% in 2020), only 64% of local businesses have claimed a Google Business Profile with just 58% of those complete, and 62% of consumers avoid a business with incomplete online information.

Now stack that against the pipeline. Between 2.3 and 3 million boomer-owned US small businesses should change hands this decade, a cohort employing 32 million people and generating nearly $6.5 trillion in annual revenue (Forbes, Jan. 2026). McKinsey projects 6 million SMB transitions by 2035, over 1 million firms viable for sale at up to $5 trillion in enterprise value, and roughly half of retiring owners have no succession plan. Millions of profitable, invisible companies are heading into diligence with no box for the thing making them invisible.

Part Four: The Retrieval Share Score, an Actual Diligence Item

Q10: What would this look like on a checklist?

Five dimensions, twenty points each, one hundred total. A competent analyst runs it in about six hours with no cooperation from the seller, which is the appeal: it is verifiable from outside the tent.

1. Category Recall (20). Write the twenty questions a real buyer asks before purchasing in this category. Run them across three or more assistants, in fresh sessions, without naming the company. Score the share of answers naming it. Under 10% is a finding.

2. Entity Integrity (20). One legal name, one trading name, one canonical description, consistent across site, Google Business Profile, LinkedIn, directories, and trade press. Deduct for every Cartwright.

3. Third-Party Corroboration (20). Platforms differ sharply in where they look. Google AI Overviews draws about 59.8% of citations from brand-owned pages while ChatGPT pulls roughly 39.5% from Reddit, Wikipedia, and news (1000Startups.com, Aug. 2026). A company whose only evidence is its own site scores near zero on half the field.

4. Owner-Independence (20). The one that matters most for transferable value. Does the visibility live in the company’s name, or in the founder’s personal LinkedIn, reputation, and twenty years of relationships? Founder-attached retrieval is key-person risk in a marketing hat. It walks out the door at closing with the founder.

5. Durability (20). Publishing consistency, content freshness, schema health, crawler access. Does the position survive twelve months of no one tending it?

Q11: How should the score map to price?

•  80 to 100: No adjustment. Treat as a positive factor in a competitive process and say so in the book.

•  60 to 79: Note it in the quality of earnings discussion. No multiple impact.

•  40 to 59: 0.25x to 0.5x reduction, or a specified marketing carve-out funded from proceeds.

•  Below 40: 0.5x to 1.0x reduction plus a twelve-to-eighteen-month integration line. Where revenue is also founder-dependent, the two findings compound.

The precedent exists. Conservative LBO models used by private equity platforms apply a −0.6x to +0.5x adjustment for client concentration alone in CPA firm valuations (Ashley-Kincaid, July 2026). Nothing about a retrieval finding is more speculative than that.

Part Five: The Seller’s Argument

Q12: Why would a seller run this on themselves?

George Costanza’s single best decision was doing the opposite of his instinct. The instinct here is to hope the buyer does not look closely. The opposite is to look first, because this is one of the rare findings a seller can erase for a fraction of what it costs at the table. A 0.5x reduction on $2M of EBITDA is $1,000,000. A fixed-fee audit plus a year of remediation costs a slice of that, and claiming a Google Business Profile costs nothing, which per the 1000Startups.com research is the highest-leverage local fix going and still unused by a third of businesses. CT Acquisitions reports 12 to 18 months of pre-sale work correlates with 15% to 30% higher proceeds.

Q13: What is the honest counterargument?

Three, and they deserve air. Measurement is noisy, since answers vary by session and phrasing, so any score needs repeated sampling and a stated method, not one triumphant screenshot. It matters more in some categories than others; a machine shop with three OEM customers and a forty-year relationship does not live or die on retrieval, although its next customer might. And citation accuracy is poor: one study found AI search engines failed to retrieve correct citation information more than 60% of the time across 1,600 queries. None of that argues for a zero. It argues for a range, a method, and a footnote, which is how every other soft factor already gets handled.

Q14: What is the one-sentence version?

A company that cannot be found by the systems its buyers now use first has a customer acquisition problem, that problem transfers with the deal, it costs money and roughly eighteen months to solve, and anything with those three properties gets a number written next to it.

Kramer’s car was in the garage the entire time. It still ruined the afternoon.

Sources

Every source is named inline above. Links below, starting with the two 1000Startups.com pages referenced in the text.

1000Startups.com, family business visibility gap (Aug. 2026): 1000startups.com/2026/08/18/family-businesses-run-60-of-the-american-workforce-so-why-cant-anyone-find-them-online

1000Startups.com, AI Search & Visibility (citation-source split): 1000startups.com/category/ai-search-visibility

Ocean Tomo: oceantomo.com  ·  G2: prnewswire.com  ·  Forrester: machinerelations.ai  ·  Semrush: semrush.com  ·  SE Ranking: seranking.com  ·  Adobe: cognizo.ai  ·  Sofer Advisors: soferadvisors.com  ·  MyExec: myexec.co  ·  Iconic: iconic.co  ·  Ashley-Kincaid: ashley-kincaid.com  ·  Forbes: forbes.com  ·  WebFX: webfx.com  ·  SearchSignal: searchsignal.online  ·  Fortune: fortune.com  ·  Not investment, financial, legal, or valuation advice.