Attorney Local SEO as a Selection Problem: Export-Style Thresholds in a Local-Only Legal Market

Visibility is a set you enter, not a rank you climb

For most of the search era, a solo or small firm competed for a position in an ordered list. The list was long, the ranking was continuous, and the difference between the fifth result and the eighth was a matter of degree. A generative answer changes the shape of the contest. When an engine reads the sources and returns two or three named firms, it has replaced a continuous ranking with a binary membership decision: a firm is in the answer or it is not. Presence below the answer, like presence on a second page, increasingly means absence from the moment of choice.

That binary is the reason the ranking metaphor undersells what is happening. The relevant object is no longer a firm's position in a market but its membership in a set the engine curates on the buyer's behalf. Membership problems have thresholds. And thresholds, in economics, are where selection effects live.

What Melitz actually proved about who goes global

In 2003, Marc Melitz published a model that reorganized how economists think about trade. Earlier theory treated firms in an industry as interchangeable. Melitz allowed them to differ in productivity, then asked what happens when they face a fixed cost of entering a foreign market, the sunk expense of building the distribution, compliance, and capability that selling abroad requires before a single unit is sold.

The result is a sorting mechanism. Firms above a productivity threshold find the fixed cost worth paying and self-select into exporting. Firms below it rationally decline and serve only their domestic market, and the least productive exit altogether. Trade does not lift every firm equally; it reallocates market share toward the more productive, and aggregate industry productivity rises because of that reallocation, not in spite of it. The load-bearing idea for our purposes is narrow and precise: a fixed cost of market entry produces a cutoff, and the cutoff sorts otherwise similar firms into the seen and the unseen.

The export data that makes the model concrete

The pattern Melitz formalized is visible in plain US statistics. Small firms are overwhelmingly the number of exporters and a small fraction of the value. In 2023, small firms were 97.2 percent of all identified US exporting firms, some 270,014 businesses, yet they accounted for only 33.0 percent of identified-firm export value. Vast participation, systematically under-scaled per firm. That is exactly the signature of a productivity threshold: many firms clear the bar to participate at all, but the firms above the bar carry most of the volume.

This is the anchor for the analogy, and also its first caveat. The export data is real and established. The claim that a similar threshold governs local visibility is an analytical framework applied by extension, not a measured finding about law firms. We keep those two things separate throughout.

The local market has borders after all

A solo practitioner in personal injury or family law never exports anything. There is no customs form, no foreign distributor, no border. So why would an export model describe the market at all? Because the mechanism Melitz identified is not about geography. It is about a fixed cost of entering a market that screens firms by their capacity to pay it. The generative-answer layer installs precisely such a cost inside a local market that used to have none.

Being named in the answer to "best injury lawyer near me" is not free and not automatic. It requires a firm to resolve to a single verifiable entity across its site, its Google Business Profile, the legal directories, and the state bar record; to publish practice-area content an engine can extract and cite; to carry current, correct structured data; and to sustain a legitimate flow of real-client reviews. Most of that is a fixed, up-front cost, incurred before a single new client arrives, and it must be maintained as engines and policies change. A local market with a fixed cost of entry and a capability-based cutoff is, structurally, an export market wearing local clothes.

Why the threshold sorts firms: capacity, not willingness

In the trade model, the sorting variable is productivity. In the visibility analogy, the closest real-world counterpart is capacity, and the evidence on small-firm technology adoption is unusually consistent about what that capacity gap is made of. The OECD's cross-country survey work finds that the barriers to small-firm digital adoption are structural rather than attitudinal: low awareness, insufficient internal resources, skill deficiencies, and financial limits. The gap is not that owners refuse; it is that they lack the money, staff, skills, or time to clear the bar.

The same body of work shows the gap widening precisely where the tools get harder. Small-firm cloud adoption was 41 percent in 2021, and over the same period the gap with large firms grew from 31 to 33 points, with the divide largest for the most sophisticated tools. Machine-legibility, the structured, crawlable, entity-consistent presence a generative engine can trust, is a sophisticated tool. If prior technology layers sorted small firms by capacity, there is little reason to expect the newest and most demanding layer to sort them more gently.

What crossing the threshold actually costs a firm

If entry into the visible set has a fixed cost, it is worth naming the cost in the specific terms of a legal practice, because the components are unusually concrete here. Solo law firm SEO and small law firm AI search readiness share the same load-bearing pieces: a single, verifiable firm identity across the site, the Google Business Profile, the legal directories, and the state bar record; practice-area pages built around the exact matter and market rather than a generalist menu; current structured data an engine can parse; and a compliant, real-client review system. These are the fixed investments that put a firm above the cutoff, and, like the export model's entry cost, they are largely sunk before any new matter is booked.

There is a young but real optimization literature underneath this. The 2024 peer-reviewed paper that introduced Generative Engine Optimization measured which content levers change whether a source is cited inside a generated answer, finding that adding cited statistics, quotations, and authoritative sources raised a source's visibility in the engines it tested. That the object of optimization has moved from a rank to a citation is itself the point: the fixed cost of entry is now partly a content-and-entity cost, and it is measurable, if not yet guaranteed.

Reputation belongs in the same ledger. Reviews are, for a law firm, both a trust signal a human reads and a corroboration signal an engine reads, and the asset is getting more expensive to hold honestly. BrightLocal's long-running consumer survey finds that 93 percent of consumers read reviews before visiting a business and 95 percent trust businesses with many reviews more, even as trust erodes at the margin: 75 percent are now concerned about fake reviews and 82 percent say they have encountered one. A steady stream from real clients, earned inside the advertising and platform rules, is part of the fixed cost of staying above the line, not an optional extra.

Reading the evidence: what we do and do not know

The strongest temptation in this domain is to reach for a single dramatic number about how much narrower generative engines are than classic search. One widely repeated industry figure holds that ChatGPT recommends roughly 1.2 percent of local businesses in category queries, against Google Local surfacing about 35.9 percent, an order-of-magnitude difference. That direction is consistent with the independent Pew click-through findings, but the specific percentages come from marketing-industry monitoring, not an audited academic or standards-body study. We treat them as contested and decline to publish them as fact.

The larger and more important gap is vertical. No verified academic or government dataset was found that quantifies the digital-adoption or AI-visibility position of solo and small law firms as a distinct sub-population. The general MSME framework is well evidenced; the legal-specific numbers are not yet measured. That absence is itself a citable observation about the state of the literature, and it is why the useful move for any individual firm is to measure its own position rather than to import a sector-wide statistic that does not exist.

A sorting mechanism, not a verdict on any firm

It matters what the selection model does and does not claim. In Melitz's world, a firm below the export threshold is not a worse firm; it is a firm for which the fixed cost is not yet worth paying, or not yet payable. The cutoff sorts by capacity to clear a specific bar, not by merit, competence, or the quality of the underlying legal work. The analogy carries that humility with it. A solo practice absent from the AI answer is not a lesser practice. It is a practice that has not yet paid, or has not been shown how to pay, the fixed cost of entering the set the engine now curates.

The practical consequence is that the threshold is crossable, and crossing it is an infrastructure question rather than a talent one. But the first step is not to spend. It is to read where a specific firm actually stands against the specific firms it loses clients to, on the questions real clients ask, so that any fixed cost is aimed at the gap that decides the matter rather than at guesswork.