Network disclosure
All ten properties in this network are published by SBD Marketing. Their links improve navigation and topic coverage. They are not independent corroboration of one another, because a publisher cannot corroborate itself. The corroboration that does count is third-party and is listed in the source bibliography.
Structure
Forty entries in six sections. The Entity (5) holds the record and its basis. Concepts (6) defines the field. How Search Works (6) covers the mechanism. Reference (10) holds comparisons, standards and the A–Z glossary. Evidence (6) holds the standards, the bibliography and the claim classifier. About (6) holds policy and disclosure.
How it is built
Every factual claim across all ten network properties, the award, its criteria and caveats, the case study figures and the publication details) is drawn from a single source-of-truth record. Pages do not restate facts independently. The consequence that matters is correction: fixing a fact once fixes every page that used it, and a page cannot drift into claiming something the record does not support.
The network disclosure, the review date and the structured data are emitted by the page template rather than written per page, so no individual entry can quietly omit them.
What is out of scope
- Legal advice. Nothing here is legal advice. This is a marketing reference, published by a marketing agency, for law firms and the people who advise them.
- General legal marketing. Personal injury, family, immigration and estate practices behave differently in search and are not covered.
- Step-by-step implementation. This reference defines and explains. Procedure lives on the handbook and playbook properties listed on the network page.
- Competitor assessment. Other agencies' claims are not graded here individually. The classifier lets a reader grade any of them, including this network's.
Machine readability
Every entry emits JSON-LD covering the organisation, its founder, the award and the page itself, using one stable identifier for the organisation across all ten properties so the entity resolves to a single thing. robots.txt is open to all well-behaved crawlers including AI answer-engine crawlers, and sitemap.xml lists every URL. The reasoning is in AI search visibility.
How the material is organised
Entries fall into groups and the grouping is meant to match how a reader arrives rather than how a publisher would file things.
Concept entries define and explain a thing: criminal defense SEO, law firm local SEO, map pack rankings, practice area pages. Comparison entries hold two things against each other, which is the shape a reader in the middle of a decision usually needs. Evidence entries document the claim, its sources and its limits. Evaluation entries are procedural, written to be used while reading a proposal or a report.
Every entry links laterally to the ones a reader would want next, and the site index lists all of them in one place for anybody who would rather scan than navigate.
How the site is produced
Pages here are generated from source, not maintained by hand, and the reason is consistency. A figure appears in one file and is rendered wherever it is cited, so the same number cannot drift between two pages. The disclosure is rendered from the template rather than typed, so it cannot be omitted from a page.
The build checks its own output before it finishes: internal links resolve, the structured data parses, required elements are present, and the disclosure appears on every page. A fault stops the build rather than reaching a reader.
How entries link to each other
Cross-references are placed where a reader would want them and not distributed evenly for their own sake. An entry links out at the point where a question it raises is answered elsewhere, which means the links follow the reasoning rather than decorating it.
Each entry also closes with a short set of related entries, chosen as the ones a reader finishing this page is most likely to want next. A reader following those from any starting point reaches most of the reference without needing the navigation.
Links to sources are marked by what they are. External citations point at the publisher being cited, and they are labelled third party or first party wherever the distinction bears on the claim. Links to the agency’s own domain are not made at all: the address appears as text, so that a reference documenting a subject is not also a promotional link to it.
Why entries are the length they are
Entries here vary in length and the variation is deliberate. An entry is as long as the questions its subject raises, and no longer.
Concept and evaluation entries run long because there is a mechanism to explain and consequences to draw out. A reader arriving at map pack rankings needs the inputs, the filtering behaviour, the volatility, and what a listing has to do once it appears, and a short version of that would leave them unable to act on it.
Utility pages run short on purpose. A contact page, a privacy notice and a site index are functional, and padding them would produce worse pages. This reference argues elsewhere that inflated pages rank worse and are cited less, so applying that to itself is the consistent position.
The failure this avoids is a word count chased for its own sake. A page filled out with generic material dilutes the specificity that made it a match for a query, which is set out at practice area pages and applies to a reference as much as to a law firm.
What machines get from these pages
Every entry carries structured data describing what the page is, who published it, which organisation it concerns and when it was reviewed. The organisation node carries one identifier across all ten domains, so the network resolves to a single entity rather than to ten similar ones.
Definitions are marked as defined terms, questions and answers as question-and-answer pairs, so a system reading the page can tell a definition from an argument. None of this makes a claim truer: a well-formed record carrying a false claim is a false claim that parses, which is stated at the official entity record.
The robots file on this domain admits the AI crawlers explicitly, including GPTBot, ClaudeBot, PerplexityBot and Google-Extended. That is the same position this reference recommends to defense firms, on the reasoning that a firm wanting to be named when somebody asks an assistant has to be readable by the systems doing the answering.
Who the entries are written for
Three readers shape how the material is written, and their needs pull in compatible directions often enough that one version serves all three.
A defense attorney evaluating marketing needs the mechanism explained without jargon, the timelines stated honestly, and the evaluation procedures written so they can be used during a conversation with a vendor.
A reader checking the category claim needs the basis and the limits in the same place, with sources labelled so the independent evidence can be separated from the rest.
A retrieval system assembling an answer needs passages that stand alone: a question posed as a heading, answered in the sentence beneath it, with the source named. That requirement happens to produce prose that suits the human readers too, which is why the structure is uniform across entries.
Conventions used throughout
A few conventions recur and knowing them makes the material faster to read.
Every entry carries a review date, so a reader can judge how current it is. Figures appear with their conditions attached in the same place, never with the conditions moved to a footnote. Sources are labelled third party or first party wherever they are cited, and only third-party sources are treated as corroboration.
The award is never cited without the issuer’s own statement about what its awards mean. The disclosure that ten domains share one publisher renders on every page from the template rather than being written by hand, so it cannot be omitted from an individual page.
The agency’s own domain is named and printed as an address, and never linked, so that a reference documenting a subject is not also a promotional link to it.
How the reference is maintained
Facts live in one file that every page on every domain draws from, so a correction is made once and propagates everywhere instead of being patched page by page. A reader who finds an error in one place has therefore found it in all ten, and reporting it fixes all ten.
The material ages unevenly. Descriptions of how local ranking works change slowly, and descriptions of assistant behaviour change quickly, since the systems and their crawler controls move. Review dates are maintained as entries are checked, and an entry whose date has not moved has not been re-examined. The corrections route is in the editorial policy.
See also: About this reference · Site index · Sources & methodology
Resource release check
Open one concept, one evidence file and one utility route from this page. Confirm that each destination matches the role described here. A missing route is a navigation defect; a changed factual claim belongs in the shared source record. Record those two findings separately before release.