SEO Expert Witness

I am retained as an SEO expert witness — a search engine optimization expert witness — in civil matters where the disputed facts live in search results, web analytics, server logs, and the history of a website. My job is not to argue that a party did good or bad SEO. It is to establish what changed on a website, when it changed, what caused the change, and what can be proven from the surviving data. Then to defend that opinion in a report, in deposition, and at trial.

I have testified in United States Federal District Court and in the Grand Court of the Cayman Islands, and I have been deposed in state, federal, and international matters. I work for plaintiffs and for defendants in roughly equal measure, and I take rebuttal assignments against other experts in this field.

What follows is not a sales page. It is a description of the standard my work has to survive, the evidence it is built from, and the places where SEO opinions most often fall apart. If you are counsel evaluating whether you need an expert in this discipline at all, that is the most useful thing I can give you.

What I am actually retained to establish

Search disputes almost always reduce to one of four questions, and it is worth being precise about which one you are asking, because they require different evidence and different methods.

• Causation. Traffic or rankings fell. Did the defendant’s conduct cause the fall, or did something else? This is the hardest question in the discipline and the one most often answered badly.
• Standard of care. Did an agency, a developer, or an in-house team do what a competent practitioner would have done at the time — not by today’s standards, but by the standards that existed on the date of the conduct.
• Attribution and reconstruction. What did this website, this page, or this search result look like on a particular date, and who controlled it? This is forensic work, and it is frequently the entire case.
• Valuation. What was the search visibility of a domain, a page, or a brand worth, and what remains of it?

An expert who cannot say which of these four he is opining on, and which he is deliberately not opining on, is an expert who is about to be excluded.

The standard the testimony has to meet

Federal Rule of Evidence 702 was amended effective 1 December 2023, and the amendment matters more to this discipline than to almost any other. Two changes: the rule now states on its face that the proponent must demonstrate to the court that it is more likely than not that each requirement is met, and subsection (d) was rewritten so that the question is whether the expert’s opinion reflects a reliable application of the methodology — not merely whether the expert applied one. The Advisory Committee said plainly that courts treating the sufficiency of an expert’s basis and the application of his method as questions of weight rather than admissibility had been getting it wrong.

Search opinions are structurally exposed in exactly the two places that amendment tightened.

Rule 702(b), sufficient facts or data. A great deal of what circulates as SEO evidence is not measurement. Third-party rank trackers, keyword volume figures, and link and traffic estimates from commercial tools are modeled numbers derived from proprietary panels. They are useful for orientation. They are a weak foundation for an opinion when the party’s own Search Console exports, analytics, and raw server logs exist and were not used. Where I rely on a third-party estimate, I say so, I state what it is estimating and from what, and I say what it cannot support.

Rule 702(d), reliable application. The characteristic SEO overstatement is the causal leap — rankings fell after the defendant acted, therefore the defendant caused the fall. The correlation of two timelines is a real method. Causation is a conclusion that method does not reach on its own, and after the 2023 amendment the gap between the two is an admissibility problem rather than something to be argued about in front of the jury.

The governing authority for an expert like me is not Daubert alone. Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993), established the gatekeeping role and the familiar reliability factors. Kumho Tire Co. v. Carmichael, 526 U.S. 137 (1999), is the case that actually reaches me: it held that the gatekeeping obligation extends to technical and other specialized knowledge, not only to scientific testimony, and that the Daubert factors are neither mandatory nor exhaustive. The Court also observed that it is sometimes useful to ask the Daubert questions even of a witness whose expertise rests purely on experience.

The practical consequence is one that a lot of people in this field have not absorbed. I have done SEO for thirty years is a qualification. It is not a methodology. What survives scrutiny is a documented, reproducible protocol — stated data sources, stated date ranges, stated exclusions, stated controls for confounding causes, and queries and crawls that another competent analyst could re-run and get the same answer from. That is what turns experience into something with an assessable error rate.

What Rule 26 requires in the written report

Under Federal Rule of Civil Procedure 26(a)(2)(B), a retained expert’s signed written report must contain a complete statement of every opinion and the basis and reasons for it; the facts or data considered in forming them; any exhibits that will summarize 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 during the previous four years; and a statement of compensation.

Two of those obligations are unusually heavy for a search expert, and they are worth raising with your expert before you retain him rather than after.

The rule says considered, not relied upon. For an analyst who runs site crawls, pulls Search Console and analytics exports, parses server logs, and queries link indexes, the set of things considered is large and easy to mishandle. I keep an ordered work log for exactly this reason, and I produce it. To ensure accuracy, I utilize state-of-the-art forensic SEO tools such as Screaming Frog, SEMrush, Ahrefs, and custom log file analysis scripts as well.

The ten-year publication list is the other one. I have written publicly about search since long before this was a rule I had to satisfy, which means the list is long — and an expert who cannot produce his own publication history will be impeached with a forgotten post of his own. I maintain mine as a working document rather than reconstructing it under deadline.

Rule 26(b)(4) protects drafts of the report regardless of the form in which the draft is recorded, and protects attorney–expert communications, with three carve-outs: compensation, facts or data the attorney provided that the expert considered, and assumptions the attorney provided that the expert relied on. That last pair is the reason I prefer to pull the underlying data myself from first-party sources wherever it still exists. It is better methodology under 702(b), and it keeps a clean line between what I measured and what I was told.

Litigation Support Across Diverse Digital Case Types

Trademark Infringement & Intellectual Property Disputes

Legal teams navigating trademark or copyright infringement disputes benefit from in-depth digital analysis. Bill Hartzer examines online brand presence, evaluates domain and content overlap, and pinpoints unauthorized usage or imitation that affects business reputation. His experience writing numerous expert witness reports and expert witness rebuttal reports streamlines the validation and presentation of highly technical evidence.

• Domain, Content, and Branding Analysis: Assessing similarity of names, site structures, digital assets, and search engine appearance to draw direct lines between online brand confusion and consumer impact.
• Rebuttal Expertise: After reviewing and dissecting opposing reports, Hartzer constructs clear, defensible rebuttal documentation rooted in empirical analysis and industry standards.

Internet Defamation

Defamatory statements made online frequently require technically precise investigation. Bill Hartzer uncovers origins of false or misleading claims within complex web environments. He compiles actionable reports that detail both the reach and impact of digital defamation—powerful evidence for litigation and settlement negotiations.

• Use of search engine cache and archival tools to identify content history, authorship, and dissemination patterns.
• Creation of detailed documentation mapping digital footprints, citing exact URLs, and compiling timelines of defamatory activity.

Online Advertising & Click Fraud

Unethical practices in digital advertising carve out new challenges for both plaintiffs and defendants. Hartzer deciphers traffic logs, pay-per-click advertising anomalies, and automated bot activity, and then demystifies these findings for courts. His trial testimony translates layered click-fraud patterns into concrete, court-admissible evidence.

• Analysis of ad campaign data to flag invalid traffic and document suspected fraud.
• Clarification of pay-per-click (PPC) issues and interpretation of clickstream data using industry benchmarking.

Google Penalties & Algorithm Updates

Website losses following search engine penalties, algorithm shifts, or manual actions require forensic-level cause and effect analysis. Hartzer establishes the timeline and scope of impact, then correlates technical actions (e.g., link building, on-page changes) with ranking and revenue losses. His depositions and trial experience support attorneys as they establish or contest liability.

• Forensic review of traffic patterns, search rankings, and business performance before and after penalties.
• Identification of penalty sources—manual action versus algorithmic—and linkage to defendant or plaintiff actions.

E-Commerce & Online Business Practices

Competitive disputes in e-commerce—ranging from website misrepresentation to digital product confusion—often hinge on expert analysis of website functionality, sales funnels, and messaging. Hartzer evaluates business tactics, uncovers misleading claims, and documents practices that may violate consumer protection or fair competition laws.

• Assessment of website architecture and user experience for deceptive or anti-competitive intent.
• Review of competitor practices with reference to FTC regulations and industry best practices.

Reflect on your current digital case: which of these areas—and Hartzer’s unique combination of technical proficiency with real courtroom experience—will best support your litigation strategy?

Search Engine Ranking Factors & Algorithm Impacts: In-Depth Expert Analysis

How Algorithm Updates Influence Website Authority and Domain Value

Major search engines such as Google refine their ranking algorithms frequently—documented core updates have appeared several times each year. Forensic examination of a website’s organic search visibility, particularly around update rollout dates, exposes patterns linking actions to outcomes. Algorithmic changes directly impact factors like domain authority, on-page relevance, backlink quality, and user engagement signals.

For example, in May 2022, Google’s broad core algorithm update caused significant volatility across multiple industries, with data from Semrush revealing up to 65% fluctuation in search rankings for some niches (Semrush Sensor, May 2022). After such updates, shifts in traffic, impressions, and keyword positions pinpoint affected content or technical elements—this provides a traceable path for explaining or quantifying a website’s organic reach over time.

Historical Review of Performance Pre- and Post-Algorithm Changes

Systematic analysis of analytics and server logs—before and after significant search engine events—offers clear evidence of causal impacts. Utilizing Google Search Console data, an expert scrutinizes impression and click-through rates by date, keyword, and landing page. By overlaying industry-acknowledged update timelines, an expert witness constructs a narrative that ties measurable ranking or traffic drops to algorithmic causes rather than business missteps or intentional sabotage.

• May 2020 Core Update: Sites reporting poor content and link profiles experienced average ranking drops of 7 positions per affected keyword, based on data from Moz (MozCast, 2020).
• Panda Update: Penalized thin, duplicate, or low-quality content; this update affected up to 12% of search results according to Google’s own statements (Official Google Blog, 2011).
• Penguin Update: Targeted manipulative backlink schemes; up to 3% of English queries were impacted and domain link profiles could be irreparably harmed (Google Search Central, 2012).

What patterns would you expect if an algorithm penalized a domain tied to manipulative practices? Reviewing time-series rank charts typically exposes steep, abrupt losses, not gradual declines—an observation uniquely valuable in legal matters.

Identifying Penalties, Redirects, Backlink Schemes, and Technical Issues

Algorithmic or manual actions by search engines may penalize websites for reasons such as link manipulation, content duplication, or improper redirects. Domain histories reveal manual penalty notifications inside Google Search Console, and backlink audits surface toxic links traced to schemes or negative SEO attacks. Redirect chains and misconfigured migrations frequently result in ranking and indexation loss—a domain’s SEO visibility graph often pinpoints the exact moment technical errors were introduced.

• Backlink Analysis: Toxic link ratios above 35% correlate strongly with Google Penguin penalties, as demonstrated by Ahrefs studies (2019).
• Redirect Problems: Sites with improper 302 temporary redirects in place of permanent 301s lose inherited link authority; investigations have shown 15-30% reductions in search traffic post-migration if wrongly set.
• Penalty Assessment: Unnatural Links penalties documented in Google’s Manual Actions report precisely quantify the domains and URLs affected, aiding forensic attribution.

Some technical questions to consider: Has the domain inherited legacy penalties after acquisition? Were unauthorized redirects or backlink floods weaponized to devalue an asset during dispute? A thorough technical audit, led by an authority such as Bill Hartzer—who has written extensive expert witness and rebuttal reports, participated in depositions, and testified at trial—produces definitive, verifiable answers.

Start me as a consulting expert

Under Rule 26(b)(4)(D), a retained expert who is not expected to testify is ordinarily beyond discovery absent exceptional circumstances. A testifying designation, once made with a served report, cannot be unmade.

In search matters the early work — crawl the site, pull what analytics and log data survive, reconstruct the timeline, determine whether the data can support the theory of the case at all — is classic consulting work, and it is frequently the work that tells you whether there is a case. I would rather do that first and be designated afterward, when we both know what the opinion is going to be. It is cheaper for the client and it is safer for everyone.

The disputes where search evidence decides the outcome

These are the matters in which I am most often retained as an SEO expert witness. The pattern across all of them is the same: a party asserts that something happened to a website’s visibility, and the only reliable record of it is technical.

Matter type	What the expert is establishing
Traffic loss and business damages	Whether an identified decline is attributable to the conduct at issue rather than to an algorithm update, seasonality, competitor action, or the party’s own changes
Defamation and negative search results	Prominence and persistence of the material for identified queries over time, reconstructed from dated evidence rather than from a screenshot taken last week
SEO malpractice and negligence	What competent practice required on the dates of the conduct, and whether the work departed from it in a way that caused harm
Agency and vendor contract disputes	What was delivered against what was contracted, and whether the deliverables had the effect claimed for them
Domain name disputes	The search visibility, history, and acquired value attached to a name — see my domain name expert witness practice
Trademark and keyword advertising	Whether a mark was used to divert search demand, in organic results or in paid placement, and what the diversion consisted of
Site migrations and redesigns	Redirect mapping, indexation, and whether the loss was foreseeable and preventable at the time of the launch
Negative SEO and link attacks	Whether an external link campaign occurred, who benefited, and whether it plausibly caused the alleged effect
Local search and map visibility	Profile control, listing manipulation, and the effect on a geographically bounded market
E-commerce visibility	Category and product-level indexation, platform migration failures, and revenue exposure by page cohort

How I establish causation — and where I stop

A naive before-and-after comparison of organic sessions attributes to the defendant everything that Google, the calendar, and the market did in the same window. Organic search has no stable counterfactual: core updates ship several times a year, plus continuous unannounced change; seasonality is strong and category-specific; competitors act independently; and query demand itself moves.

What search data does offer, and what most damages analyses in this field fail to use, is unusually good internal controls:

• The party’s other channels — direct, email, paid — as a control for demand-side movement that has nothing to do with the defendant.
• Unaffected page cohorts on the same site as a within-firm control for a sitewide algorithmic event.
• Competitor visibility over the same window as a market-level control.
• Query-level demand data for the affected keyword set, so a collapse in demand is not mistaken for a collapse in ranking.

A difference-in-differences design across affected and unaffected cohorts, spanning the event date, is a real method with an assessable error rate. It is what a Kumho inquiry is asking for, and it survives cross-examination in a way that a single overlaid line chart does not.

Where I stop is equally important. I opine on causation and on the traffic or visibility counterfactual. I do not convert that into a dollar figure. Revenue per session, incremental margin, avoided costs, mitigation, and present value belong to a CPA or an economist, who takes my traffic delta as a stated assumption. Experts who compute the money themselves invite an attack on the financial half of the opinion that takes the technical half down with it. Courts have excluded damages models for ignoring known market conditions, for growth curves capable of producing any desired result, and for resting on say-so rather than analysis. Keeping the two opinions interlocking but non-overlapping is the defense against all three.

Authenticating web evidence

A screenshot is not evidence of anything until somebody can say what it is, where it came from, and when. Under Rule 901(a) the proponent must produce evidence sufficient to support a finding that the item is what it is claimed to be, and the routes that matter for web material are 901(b)(1) testimony from a witness with knowledge, 901(b)(4) distinctive characteristics taken with the circumstances, and 901(b)(9) evidence describing a process or system that produces an accurate result. That last one is the hook for a crawler, an archiver, or a logging system.

Rules 902(13) and 902(14), added in 2017, allow records generated by an electronic process and data copied from a device or file to be self-authenticating on a qualified person’s certification — with data copies ordinarily authenticated by hash value, matching hashes attesting that original and copy are exact duplicates. This turns what used to be a foundation fight into paperwork, and it is why I capture contested material forensically, with a SHA-256 computed at the moment of capture, rather than by pressing Print Screen.

Internet Archive material is routinely used and is not routinely self-authenticating. Courts admitted it as far back as Telewizja Polska USA v. Echostar Satellite (N.D. Ill. 2004) on an Archive representative’s affidavit, and Lorraine v. Markel American Insurance Co., 241 F.R.D. 534 (D. Md. 2007), remains the standard starting point for any electronic-evidence foundation argument. But the failures are instructive and they cluster: archived material offered for a date range rather than a date, captures whose scope the witness could not vouch for, and material offered for the truth of its contents without a hearsay answer. The Federal Circuit has held that archive material merely from the same year as the critical date could not support a verdict. Nail the date. Where I know a capture will be contested, I treat the archive as corroboration and a contemporaneous forensic capture as the primary exhibit.

Exhibits that reach the jury

Almost everything an expert in this field produces is a chart or a summary: rankings over time, sessions by channel, link acquisition timelines, crawl error rollups, keyword overlap tables. As of 1 December 2024 two different rules govern them, with opposite consequences.

Rule 1006, as amended on that date, admits a summary, chart, or calculation offered to prove the content of voluminous admissible material that cannot conveniently be examined in court — whether or not the underlying material is itself introduced. A Rule 1006 summary is substantive evidence. New Rule 107 governs the illustrative aid, and states that an illustrative aid is not evidence and must not go to the jury during deliberations absent consent or an order for good cause.

So the difference between an exhibit that goes into the jury room and one that does not is a set of drafting choices made months earlier: whether the underlying data is admissible, whether it was produced to the other side, and whether the exhibit is a faithful non-argumentative summary or an argument with axes. I build to Rule 1006 by default and produce the underlying data with the exhibit.

Preserve the data before it deletes itself

The single most avoidable failure in search litigation is that the best evidence expires on a timer while everyone is still exchanging letters. Under Rule 37(e), severe sanctions for lost electronically stored information require a finding that the party acted with the intent to deprive — and ordinary, automatic, good-faith data hygiene destroys the most probative material in these cases silently, with no intent at all.

There is a second reason to involve someone like me early. Rule 37(e) sanctions require that the lost information cannot be restored or replaced through additional discovery — and in this discipline a surprising amount of it partially can be. Archive snapshots partially replace lost page content. Third-party link and ranking indexes partially replace lost link data. Bing Webmaster Tools partially replaces Search Console. Determining whether the threshold is even met is frequently an expert question, and it decides whether any sanction is available at all.

What changed in search after 2023, and why every older damages model needs revisiting

Every damages model built before 2024 assumed that the relationship between rankings and traffic was roughly stationary. It is not, and the change is large enough to swallow an entire claim.

Roughly two-thirds of United States Google searches now end without a click to any website, up sharply in two years, and AI-generated summaries appear on a substantial share of queries with a heavy effect on click-through where they appear. A before-and-after model spanning 2023 to 2026 that does not control for this attributes an industry-wide structural change to the defendant. That cuts in both directions: it is a serious problem for a plaintiff’s expert who ignores it, and it is a powerful confound for a defendant to raise — which is why the ability to quantify the effect on a specific keyword cohort, rather than to wave at it, is now the most valuable single analytic capability in this field.

Two other developments belong in any current assessment. Internal Google ranking documentation became public in 2024 and Google confirmed its authenticity; it contains no weights and no indication of which attributes are live, which makes it genuinely useful for impeaching the proposition that Google’s public guidance describes the system, and genuinely dangerous for an expert who treats a listed attribute as a confirmed ranking factor. And the federal antitrust proceedings against Google have produced sworn testimony about ranking mechanics that corroborates parts of that record. An expert who can navigate what those documents do and do not establish is materially more useful than one who quotes them.

Qualifications and testimony record

I have worked in search since 1996, before Google existed, first in-house as an SEO and webmaster at an international telecommunications software company running multilingual sites, and since then as a consultant. I still run search work for clients today, which is the reason I can speak to current practice rather than to practice as it stood when I stopped doing it. My work covers technical audits, site architecture, international search, link analysis, penalty and recovery work, paid search, web analytics, and the domain name system — and it is the combination, rather than any one of them, that most litigation actually needs.

• Recognized speaker at major SEO conferences (Pubcon, SMX, State of Search)
• Featured in Search Engine Land, Moz, and Search Engine Journal, setting benchmarks for industry best practices
• Developer and consultant for websites with millions of monthly visitors
• Client portfolio spanning regulated sectors such as healthcare, finance, and legal services
• Advisor to law firms on litigation-triggered SEO strategy and website forensics

I have been retained as a search engine optimization expert witness for more than eight years. In that time I have written 36 expert reports and rebuttal reports, been deposed eight times, testified at trial in five matters, and testified in one arbitration — in state, federal, and international proceedings, including United States Federal District Court and the Grand Court of the Cayman Islands. I take plaintiff and defense work and I write rebuttal reports against other experts in this field. Related expert witness practices are described separately: domain names, social media, and digital marketing. A deeper reference on search evidence and expert testimony — the rules, the case law, and the methods — is maintained at searchengineoptimizationexpertwitness.com.

The list of matters in which I testified at trial or by deposition during the previous four years, as Rule 26(a)(2)(B)(v) requires, is maintained and provided on retention, together with a full curriculum vitae and the ten-year publication list required by Rule 26(a)(2)(B)(iv).

Engaging me as an SEO expert witness

The first call is a conflict check and a short conversation about what the dispute actually turns on. It costs nothing and it is frequently the most useful half hour in the matter, because a meaningful share of prospective search cases die there — the data is gone, or it never supported the theory, and it is far better to learn that before a complaint is drafted than during a deposition.

If we proceed, I would normally start as a consulting expert, work the data, and tell you honestly whether there is a defensible opinion in it. Fees are hourly against a retainer, set out in an engagement letter, with separate rates for deposition and trial days. Note that under Rule 26(b)(4)(E) the party taking an expert’s deposition ordinarily pays the expert’s reasonable fee for the time spent responding.

Frequently asked questions

This page describes the evidentiary environment in which I work and the standards my testimony has to meet. It is general information about my practice, not legal advice, and nothing on it creates an expert or consulting engagement.