SEO and Digital Marketing Expert Witness Services

What an SEO expert witness actually does

An SEO expert witness examines search, advertising, and analytics records and explains what they show — and what they do not show — to attorneys, judges, and juries. The work is evidentiary, not promotional: reviewing productions, reconstructing what a result page or an ad account looked like at a given time, testing whether a technical claim is supported by the data, writing a report, and defending it under cross-examination.

Two roles get conflated. A consulting expert works behind counsel — testing whether a technical theory holds, identifying what to demand in discovery, drafting questions for an opposing expert's deposition — and is generally not disclosed and files no report. A testifying expert is disclosed, serves a report satisfying the disclosure rules, sits for deposition, and can be examined on everything considered. Moving from the first role to the second carries discovery consequences and is counsel's decision.

The disputes where this evidence turns up

Search engine expert witness work clusters into a few recurring fact patterns:

• Trademark and keyword advertising. Lanham Act claims — section 32, 15 U.S.C. 1114 for registered marks, section 43(a), 15 U.S.C. 1125(a) for unregistered marks — where a competitor bought a mark as a keyword or used it in ad copy or landing pages. The technical work is reconstructing the result page as it appeared, separating paid from organic placement, and establishing keyword targeting, match types, negative keywords, and ad rotation from account data.
• Click fraud and invalid traffic. Competitor clicks, bot traffic, publisher fraud in affiliate networks, domain spoofing, ad stacking, misrepresented inventory — read through server logs, IP and user-agent patterns, timing distributions, conversion anomalies, and reconciliation of ad-server, analytics, and billing data.
• Advertising performance and agency disputes. Work not performed, results misreported, undisclosed markup between platform spend and invoiced spend, unauthorised account changes, refusal to transfer account ownership, lead quality.
• Lost profits computed from traffic data. A de-indexing event, an algorithm update, a defamation campaign, a hosting outage, a hijacked domain. The reliability of the analytics is usually the battleground.
• Website and platform migration failures, where technical causation is contested between agency, developer, platform vendor, and market.

Domain disputes run down two different tracks

The UDRP (Uniform Domain Name Dispute Resolution Policy) is a contractual administrative proceeding, not litigation. Under Policy paragraph 4(a) the complainant must prove all three elements: that the domain is identical or confusingly similar to a mark in which it has rights, that the respondent has no rights or legitimate interests in it, and that the domain was registered and is being used in bad faith. Paragraph 4(i) limits the remedy to cancellation or transfer — no damages.

The ACPA (Anticybersquatting Consumer Protection Act), 15 U.S.C. 1125(d), is federal litigation. It requires bad faith intent to profit plus registration, trafficking in, or use of a qualifying domain — disjunctively, which is broader than the UDRP's conjunctive test — and section 1125(d)(1)(B)(i) lists nine non-exclusive bad-faith factors. Section 1125(d)(2) provides an in rem action against the domain itself, and 15 U.S.C. 1117(d) allows elected statutory damages of not less than 1,000 and not more than 100,000 dollars per domain name.

What the evidence consists of, and how it gets authenticated

The record in these matters is almost entirely machine-generated:

• Analytics exports, with date range, filters, and segment definitions stated
• Raw server and CDN logs, often the only unmodelled record of what hit the site
• Ad platform data — search terms and keyword reports, auction insights, billing records, change history
• Archived pages from the Wayback Machine and other archives, with capture dates
• WHOIS history, DNS history, registrar and hosting records
• Crawl and index-status data showing what was reachable and indexable at a point in time

Because all of it is output rather than testimony, authentication runs through FRE 901(b)(9): evidence describing a process or system and showing that it produces an accurate result. That clause is why the configuration of the measurement system is foundational rather than a matter of weight. If nobody can describe how the data was generated, the exhibit has a problem before anyone reaches the opinion.

The standard the testimony has to meet

Federal Rule of Evidence 702, as amended effective 1 December 2023, permits expert opinion only where the proponent demonstrates to the court that it is more likely than not that each of four requirements is met: that the specialised knowledge will help the trier of fact; that the testimony is based on sufficient facts or data; that it is the product of reliable principles and methods; and, under subdivision (d), that the expert's opinion reflects a reliable application of the principles and methods to the facts of the case.

Allocating reliability to the court, on a more-likely-than-not standard, means an analytics-based opinion cannot be waved through as a matter for the jury to weigh. The rewritten subdivision (d) is aimed at overstatement: the opinion must stay inside what the method and the data support. A report served after 1 December 2023 that recites the superseded subdivision (d) language hands the other side a free cross-examination point.

What Rule 26 requires in the written report

For a retained testifying expert in federal court, FRCP 26(a)(2)(B) requires a report prepared and signed by the witness containing six things: a complete statement of all opinions with the basis and reasons for them; the facts or data considered in forming them; any exhibits that will be used to summarise or support them; the witness's qualifications, including all publications authored in the previous ten years; a list of all other cases in which the witness testified at trial or by deposition in the previous four years; and a statement of compensation.

Two items cause most of the trouble in data-heavy matters. Considered, not relied upon — material reviewed and then set aside is still disclosable, which matters when large exports were pulled and only some used. And exhibits: SERP captures, log extracts, and analytics screenshots need capture dates and a stated method, not a first appearance at deposition.

Why technical evidence in this field falls apart

This is the part worth reading twice, whichever side you are on. Most weak search and analytics evidence fails in one of four ways.

• Third-party traffic estimates presented as measurements. Figures from tools that estimate another site's traffic, keyword volumes, or market share are modelled from panels and clickstream samples. They are useful for orientation and they are not measurements of the site in question. Offered as the quantum of loss, they invite a challenge.
• Correlation offered as causation for a ranking change. Traffic fell in the week of an algorithm update, therefore the update caused it. A defensible causation opinion controls for what else changed — the release calendar, redirect and canonical changes, robots and noindex directives, paid spend, seasonality, competitor activity.
• Analytics whose configuration was never examined. Filters, internal-traffic exclusions, cross-domain setup, bot filtering, consent-mode behaviour, sampling, retention windows, a mid-period property migration, a changed attribution model — any of these can move a number more than the conduct in dispute.
• Screenshots with no provenance. A capture with no date, no URL, no location or device stated, and no description of method proves very little, because results vary by location, device, signed-in state, and time.

How an engagement runs, and how to make contact

The sequence is consistent: a conflicts check against every party and counsel of record; a scoping call to fix the technical question, the role, and the deadlines; an engagement letter and retainer; intake, meaning an inventory of what has been produced, what is referenced but missing, and what must be requested or preserved before it ages out of a platform's retention window; then analysis, report, deposition, and trial.

This site is published by Bill Hartzer of Hartzer Consulting, who has more than twenty-five years of experience in technical SEO, search engine marketing, website and domain migrations, domain names, and internet investigations, and who serves as an expert witness in matters involving search engine optimization and search engine marketing, Google Ads and online advertising, web analytics and website traffic analysis, internet marketing, domain names, and domain ownership disputes. Engagements have come from both plaintiffs and defendants, and each is approached as independent analysis rather than advocacy.

A useful first message states the parties and case caption, the court and jurisdiction, the operative deadlines — expert disclosure, discovery cutoff, trial date — the technical question in dispute, and whether the need is consulting or testifying.

This page is general information about how search, advertising, and analytics evidence is handled in litigation. It is not legal advice.

Frequently Asked Questions