Reactive practice

We find out what is actually there, then take down or correct what can be.

Something specific is happening and it is affecting decisions other people make about you. The first step is a dated record of every result, source, and machine-held assertion. The second is changing the ones that can be changed, in an order that holds.

What is actually happening

A single result is doing the work of a whole reputation.

Most people arrive with the same underlying problem in a different costume. Somewhere in the first page of results, or in the answer an AI system gives when asked about them, there is a piece of material that is outdated, one-sided, or simply false. It is not representative. It may be years old. And it is being read as the summary of everything.

The damage is rarely dramatic. It shows up as a deal that slows without explanation, a board seat that goes elsewhere, a candidate who takes another offer, a bank that asks one more question than it asked last year. By the time the pattern is obvious, it has usually been operating for a while.

The work has two halves. The first is establishing what is actually there: the specific set of URLs, statements, and machine-held assertions that produce the current picture, rather than what you fear is there or what a monitoring tool reports. The second is changing them in an order that holds.

The reactive practice

Six mechanisms, applied in the order the situation requires.

These are not packages. Which ones apply, and in what sequence, is the substance of the assessment. Running them in the wrong order can entrench the problem you are trying to solve.

01

Content Removal

The only permanent change.

Removal is the only intervention that does not have to be maintained. Everything else in the reactive practice manages position, and position can move again. When a page ceases to exist, the question closes.

It is also the intervention with the narrowest set of legitimate grounds, which is why the honest version of this service begins with a determination rather than a promise. Material comes down for specific reasons: it violates the platform's own published policy; it contains a factual error the publisher is willing to correct; it falls under a statutory right of erasure; it has passed the retention period the publisher itself set; or the person who published it is willing to reconsider when approached properly.

Each of those is a different process with a different counterparty and a different standard of evidence. A platform policy claim is a documentary exercise. A correction request to a publication is an editorial conversation, and it succeeds or fails on the quality of the evidence and the manner of the approach. Confusing the two, or sending the same aggressive letter to both, is how requests get refused and noted.

Where the limits are

Accurate, lawfully published, newsworthy material does not come down. Not for us, not for anyone, and not for a larger fee. A court order is the only mechanism that changes that, and obtaining one is a matter for the attorney you retain.

Removal requests also leave traces. A refused request can draw editorial attention to material that was quietly decaying, and some publishers report attempts. This is a real cost and we weigh it before acting, not after.

The right tool when

  • The material breaks a published rule

    Non-consensual imagery, doxxed personal data, impersonation, or content that violates the platform's own stated terms. The strongest and fastest category.

  • There is a demonstrable factual error

    A name confused with another person, a charge reported without its dismissal, a figure that was never correct. Publishers with editorial standards do correct these when shown the evidence.

  • The publisher's own retention period has passed

    Many sites, particularly aggregators and records databases, publish a retention policy and then fail to apply it. Pointing at their policy is often sufficient.

  • The original author is reachable

    A surprising amount of damaging material was posted in a bad week years ago by someone who no longer feels strongly. Approached properly, some of it simply comes down.

Asked about this mechanism

Can you remove a news article?
Almost never, and you should be suspicious of anyone who says otherwise. Legitimate news organizations do not remove accurate reporting. They do sometimes append corrections, add context, or update an old story to reflect a later outcome, which is often the more valuable result anyway.
What about the Wayback Machine and cached copies?
Archives operate under their own policies and are handled separately from the original publisher. Some honor exclusion requests, some do not. This is part of why we document the full footprint at the start rather than treating the source page as the whole problem.

02

Search Result Displacement

When material cannot be removed, position is the variable that remains.

Displacement works by publishing material with more genuine weight than the item you want moved, so that page one carries a whole record instead of a single episode. The practical difference between the second result and the twentieth is the difference between a fact everyone encounters and a fact almost nobody does. That gap is where this work operates.

Displacement is often described as though it were a technical trick. It is not. Search systems rank material by assessed authority, relevance, and durability, and the only reliable way to outrank something is to publish material that genuinely earns a higher assessment: substantive work, on platforms with editorial standards, about subjects you actually have standing to discuss.

This has an implication people find counterintuitive. The work that displaces a negative result is largely indistinguishable from the work of building a legitimate public profile, which is why this mechanism and the proactive practice overlap so heavily. It also means the timeline is measured in months: material has to be written, placed, indexed, and then accrue the signals that move it. Any proposal to compress that into weeks involves methods that create a larger problem than the one you started with.

Where the limits are

Displacement does not delete anything. A determined researcher, whether a journalist, an opposing counsel, or a diligence firm, will still find the material, and should be assumed to. What changes is the incidental encounter: the person who searches your name before a meeting and reads the first four results.

Everything we publish is substantive material you have standing to put your name to. Private blog networks, purchased links, spun content, and fabricated profiles are not used here: they work briefly, they are detectable, and the penalty lands on you rather than on the vendor.

The right tool when

  • Accurate but unrepresentative material ranks highly

    A true report of a single episode occupying the position that should belong to twenty years of work.

  • Removal has been attempted and refused

    The publisher has considered it and declined. Continuing to press rarely helps; changing the surrounding picture does.

  • The material is opinion rather than fact

    A review, a blog post, a forum thread. There is no factual error to correct, so the only available lever is proportion.

  • A namesake is being confused with you

    Someone else's record is being read as yours. The work is entity separation, making it unambiguous which person is which, as much as displacement.

Asked about this mechanism

Isn't this just gaming the algorithm?
It would be, if the material were fabricated. When the published work is substantive, accurate, and genuinely accepted on its merits by outlets that exercise editorial judgment, the result is a search index that reflects a fuller picture, which is what the index is supposed to do.
Will it work against a very high-authority news result?
Often not, and we will say so. A major national outlet's coverage of a genuinely newsworthy event carries authority that is extremely difficult to displace. In those cases the honest strategy is usually context and proportion rather than displacement.

03

Crisis Response

The first seventy-two hours determine the following two years.

Most of the lasting damage in a reputational crisis is not caused by the original event. It is caused by the response, which is usually improvised by exhausted people at speed. A crisis compresses judgment: something has broken publicly, the phone is ringing, and every instinct pushes toward doing something visible immediately. That instinct is responsible for most of the permanent damage we are later asked to repair.

The first question is not what to say. It is what is actually happening: what has been published, where, by whom, with what reach, and on what trajectory. A story with real velocity and a story that looks alarming but is confined to one thread require opposite responses, and the second is far more common than it feels from the inside. Responding loudly to a contained story is how it stops being contained.

We work alongside the attorney and communications people you already have. Our contribution is the digital dimension: what the searchable and machine-readable record will look like in six months, which is the part that outlives the news cycle and the part most crisis responses ignore entirely.

Where the limits are

We are not a law firm and we do not give legal advice. In an active crisis with legal exposure, your attorney leads and we work at their direction so that privilege is preserved.

We cannot make a true story untrue. Where the underlying facts are genuinely adverse, the work is accuracy, proportion, and the record of what you did next, rather than disappearance. And we will decline a crisis engagement where the objective is to conceal ongoing harm. That is not a negotiating position.

The right tool when

  • A story has just broken

    Hours old, trajectory unknown, and every decision made today will be visible for years.

  • A journalist has made contact

    There is a deadline and a decision about whether and how to engage. The wrong answer here is usually silence by default.

  • The event has passed but the record has not

    Coverage has stopped; the search results have not. This is the phase most crisis firms leave, and where durable damage actually accumulates.

Asked about this mechanism

Should we issue a statement?
Often not, and it is the most common reflex we advise against. A statement converts a story about an event into a story about your response, extends the news cycle, and creates a permanent document that will be quoted back. Sometimes it is necessary. It is rarely necessary as fast as it feels.
How quickly can you start?
Call 914.786.3544. Active crises are triaged ahead of everything else, and the initial assessment conversation happens the same day wherever possible.

04

Autocomplete Repair

A suggestion appears before a single result is read.

When someone types your name and the search box completes it with a word you would not choose, that word has already done its work, in under a second. The suggestion frames everything read afterwards, and many people never click past it; the suggestion itself was the answer they took away.

Predictions are generated from aggregate search behavior, weighted by recency and region and filtered by policies that vary by platform and change without notice. They are not editorial statements. But they read as though they were, because they appear in the same authoritative typeface as everything else the search engine says.

There are two legitimate levers. The first is policy: platforms remove predictions that violate their published rules, and where the ground exists this is frequently the fastest meaningful result available anywhere in the reactive practice. The second is behavioral: predictions reflect what people actually search, so sustained genuine interest in other associations shifts the distribution over time. This is slower, and it cannot be manufactured with scripted queries, which platforms detect and discount.

Where the limits are

Predictions are personalized and regional. What you see is not necessarily what a counterparty in another city sees, which is why assessment uses de-personalized, multi-region capture rather than what appears on your own screen.

Where a prediction reflects genuine, widespread, current public interest in a real event, it will not move. That is the system reporting reality accurately. Bot networks and paid query farms are not used here: platforms identify the pattern, discount it, and the prediction survives with a detection signal attached to you.

The right tool when

  • A prediction makes an allegation

    Your name completed with a criminal or ethical accusation. Where you are not a public figure, this frequently violates the platform's own published policy.

  • A company name draws a scam or fraud suggestion

    Frequently the highest-cost version of this problem, because it intercepts customers at the moment of highest intent.

  • A namesake's associations are attaching to you

    Someone else's record generating the prediction on your name. Entity separation is the substance of the work here.

Asked about this mechanism

Can autocomplete suggestions actually be removed?
Yes, when they violate a platform's published policy. This is one of the few places in the field with a formal, documented removal channel. Where no policy ground exists, the honest answer is displacement over time rather than removal.

05

AI Answer Correction

A model can state something false about you with complete composure.

An AI answer arrives in fluent prose, with no indication that it is guessing. There is no complaints department. A search result is a document with an author and a publisher, meaning someone to write to. An AI answer is a synthesis: no author, no publication date, no correction policy, and no address. The correction has to happen upstream, in the sources the answer was assembled from.

The errors take recognizable forms. Conflation: your record merged with a namesake's, sometimes across countries. Staleness: a matter that resolved years ago presented in the present tense. Invention: a plausible detail with no source at all, generated because the pattern fitted. And omission, which is often the most damaging: a substantial career summarized in one sentence about its single worst week.

None of these can be corrected by contacting the model. What can be corrected is the material the model draws on: the pages, databases, profiles, and structured records that constitute the available evidence about you. The verification discipline matters as much as the correction. Answers vary by model, by version, by phrasing, and by day, so we test across systems with multiple phrasings and document the distribution rather than the best case.

Where the limits are

No one controls what a model outputs. Not us, and not the companies that build them. Anyone claiming a guaranteed AI outcome is describing something that does not exist. Training data is also periodic: a model trained before a correction may continue to reflect the older picture until it is retrained, and that schedule is not ours.

We do not attempt prompt injection, poisoned inputs, or manufactured source material. They are detectable, they degrade the information environment, and a client caught doing it has a far worse problem than the one they started with.

The right tool when

  • A model states something factually wrong

    A role you never held, a company you never ran, an outcome that never occurred, delivered with complete confidence.

  • Your record is merged with a namesake's

    The most common failure mode. Two people with one name, one of whom has a problem, and the model has not distinguished them.

  • A resolved matter appears unresolved

    The allegation is in the sources; the dismissal or settlement is not, or is far less visible.

  • Different systems give contradictory answers

    A reliable signal that the underlying source material is inconsistent, which is a fixable condition.

Asked about this mechanism

Can you make a chatbot stop saying something about me?
Not directly, and no one can. What we can do is correct and strengthen the sources it draws on, which measurably changes what it says over time. We report what actually changed rather than what we hoped would.
How long until an AI answer changes?
Retrieval-based systems that search live can reflect corrected sources within weeks. Answers drawn from training data may not change until the model is retrained, on a schedule the developer controls and does not publish.

06

Litigation Support

Support for your attorneys, never a substitute for them.

Documentation, analysis, and digital strategy at the direction of the licensed counsel you retain. We are not a law firm and we do not give legal advice. We give your attorneys the evidentiary and technical material the digital dimension of a matter requires.

The work is largely evidentiary. Defamation matters turn on demonstrating publication, reach, persistence, and harm, each of which is a technical exercise in a digital context. When was this published, and can that be proven? How many people plausibly saw it? Where has it been syndicated, cached, archived, screenshotted, and re-posted? What did the search results actually look like on a given date, and can that be shown rather than asserted?

The second half is strategic coordination. Litigation is public. Filing creates a durable record that is itself searchable, and it frequently generates more visibility than the original material: the well-documented outcome where a plaintiff wins the case and loses the search results. That trade-off is a decision for you and your counsel, and it deserves to be made with the digital consequences quantified rather than assumed. We also work after judgment, because a court order requiring removal still has to be executed against platforms, indexes, caches, and archives.

Where the limits are

We do not assess the merits of a claim, recommend whether to file, or represent you in any proceeding. Those are matters for the attorney you retain. We are also not a substitute for a digital forensics examiner where formal forensic certification is required; where a matter needs that, we will say so.

Engagement through your attorney is strongly preferred. It affects how our work product is treated, and that protection is straightforward to establish at the start and impossible to add later.

The right tool when

  • Counsel is assessing a defamation claim

    The technical record of publication, reach, and persistence has to be documented before merits can be evaluated properly.

  • An order has been obtained

    Execution against platforms, indexes, caches, and archives, most of which require a process rather than a letter.

  • Counsel is weighing whether to file at all

    An honest estimate of the visibility litigation will itself generate, so the decision is made with that cost visible.

Asked about this mechanism

Should we sue?
That is a question for your attorney, and we will not answer it. What we will do is quantify the digital consequences of proceeding, including the visibility litigation itself generates, so the decision is informed.
Can you find who posted something anonymously?
We document the technical record that supports your counsel's process for pursuing identification. We do not undertake unauthorized investigative activity, and we decline requests to do so.

Sequence

Order is not a detail. It is most of the outcome.

  • Establish the baseline, in writing

    Every result, source, and machine-held assertion, documented and dated before anything is touched. Without one there is no way to prove later that anything actually changed.

  • Separate the removable from the permanent

    Some material has a legitimate basis for removal, and some does not. Establishing which is which before anything is spent is what keeps the work from becoming expensive activity with nothing to show for it.

  • Act on the removable first

    Removal at the source is the only change that is genuinely permanent. Everything else is a matter of position, which means it can move again.

  • Build weight against what remains

    Displacement is earned, not purchased: substantive material, published where publication is a decision someone else makes on merit.

  • Verify, then keep verifying

    Against the original baseline, on a schedule, including in AI systems that answer differently this quarter than they did last quarter.

Questions we are asked

Can you guarantee removal?
We can usually tell you within the first assessment whether a legitimate basis for removal exists: a policy violation, a factual error, an expired retention period, a reachable author. We pursue every one that does. What nobody can guarantee is the decision itself, because it belongs to a publisher, a platform, or a court. Where removal is not available, we tell you what is.
How long does this take?
It depends on the mechanism. A platform policy violation can resolve in weeks. Displacement through published work is measured in months, because it depends on material that has to earn its position. Anyone quoting you a fixed timeline before seeing your situation is quoting a sales figure, not an estimate.
Will taking action make it worse?
Sometimes, yes. A removal request can draw attention to material that was decaying quietly, and litigation creates a public record of its own. This is precisely why assessment comes before intervention, and why our advice is sometimes to do nothing at all.
Do you work with our attorneys?
Regularly, and we prefer it. We are not a law firm and do not give legal advice. Where a matter has a legal dimension, we work at the direction of the counsel you retain so that privilege is preserved and the strategies do not work against each other.

Tell us what is happening.

One conversation, in confidence, with an honest reading of what can be changed and what cannot. Under a mutual non-disclosure agreement if you would prefer.

If your matter is in litigation, or likely to be, have your attorney contact us instead. Communications routed through counsel are treated differently, and that protection cannot be added afterwards. Why this matters