Six Downloads Before the Lawsuit: the Streisand Effect, With the Real Numbers

What is the Streisand effect?

The pattern called the Streisand effect is what happens when an attempt to remove, hide, or legally suppress information hands it a much larger audience than it would have found otherwise. It is named after Barbra Streisand's 2003 lawsuit over one aerial photograph of her Malibu house, an image almost nobody had looked at until she sued to have it taken down. The attempt creates the audience, because a takedown, a lawsuit, or a vanished comment is a story in a way the original item was not.

For a marketer the question is narrower than the folklore: not whether the effect is real, but whether your response to one critical thread will be more interesting than the thread.

What actually happened in the Streisand lawsuit?

Barbra Streisand sued photographer Kenneth Adelman for $50 million in 2003 over one frame in the California Coastal Records Project, the aerial survey Kenneth and Gabrielle Adelman began in 2002 to photograph the California shoreline every 500 feet. The frame was Image 3850, filed on the site as "Streisand Estate, Malibu." She claimed invasion of privacy; Adelman was documenting coastal erosion and did not know whose roof was in the frame.

The numbers are why this case named the pattern. Wikipedia's article on the Streisand effect records six downloads of the image before the suit became public, two of those by Streisand's own attorneys. In the month after the case was reported, more than 420,000 people visited the site.

She lost. The case page at californiacoastline.org, published by the project itself, dates the threat letter to February 2003, the tentative anti-SLAPP ruling to December 3 2003 and the final ruling to December 31 2003, in Los Angeles Superior Court under case number SC077257. The court struck the suit under California's anti-SLAPP statute and awarded $177,107.54 in legal fees and costs, of which that page puts Adelman's share at $155,567.04. Streisand paid a photographer's lawyers to establish that four people who were not her attorneys had seen the picture.

Image 3850, Before and After

6 downloads, 2 by Streisand's attorneys

$50m suit filed, then reported

420,000+ visitors the following month

Who coined the term Streisand effect, and when?

Mike Masnick coined the phrase on Techdirt on January 5 2005, two years after the lawsuit, in a post that was not about Streisand. It is titled "Since When Is It Illegal To Just Mention A Trademark Online?" and covers a resort whose lawyers sent a cease and desist to Urinal.net, a site that publishes photographs of urinals, demanding it stop naming the resort. The lawyers then claimed their own letter was copyrighted and could not be republished.

Masnick's coinage is four words: "Let's call it the Streisand Effect." The sentence around it asks how long lawyers will take to work out that repressing something online is what gets it seen. Cite Masnick and the January 5 2005 post by name; most articles about the term skip both.

Why does trying to suppress something create the story?

Suppression is newsworthy in a way the original item is not. A photograph of a roof carries no conflict; a $50 million lawsuit over that photograph has a named plaintiff, a named defendant, stakes, and a party who looks like they have something to hide, which is the shape news desks and community front pages rank. The attempt supplies the missing story and proves someone with money cared.

A second mechanism runs alongside it on community platforms. Suppression turns a passive audience into a participating one, because readers who would have scrolled past a complaint will mirror one someone tried to erase. Assume your response is the story, then pick one you would be content to see quoted in full.

The Photo vs the Lawsuit

One aerial photograph

• No named plaintiff or defendant
• No conflict, no stakes
• Passive audience, almost nobody looked

A $50 million lawsuit

• Streisand v. Adelman, named parties
• Stakes, and a plaintiff who looks like she has something to hide
• A participating audience that mirrors what was suppressed

How much amplification are you risking?

Measure it as the ratio between your critic's audience and the audience your response created:

Amplification = audience after your response / audience before your response

Run the origin case through it: 420,000 visitors divided by 6 downloads is about 70,000x. Those figures count different things, downloads of one image against visits to a whole site, so treat 70,000 as an order of magnitude and not a measurement. The order of magnitude is the point, and the multiplier is set by how aggressive your response was, not by how wrong the critic was.

Get the starting number before you act. Read the thread's current traffic, which is what brand monitoring is for, and compare it to the audience you would escalate onto.

Which response to criticism backfires, and when?

Every response has a case where it is correct and one where it detonates, and the difference is rarely the criticism itself.

Response	When it works	When it backfires
Say nothing	Low engagement, no factual error	The false claim spreads unanswered into search and AI answers
Reply in the thread under your own name	You have a fix, a date, or a correction	The reply is defensive, templated, or posted from an account with no history
Correct an error with evidence	The error is checkable and you link proof	You correct a trivial detail and leave the main complaint standing
Ask the poster privately to edit	You caused a real error and offer a remedy	The message reads as pressure, and the poster screenshots it
Ask moderators to remove the thread	The post breaks a written rule you can quote	You are asking for removal of criticism, and mods say so publicly
Send a legal threat or DMCA notice	Real infringement, impersonation, leaked private data	The claim is opinion, and the letter becomes the headline
Publish a statement on your own site	You need one canonical version of the facts	You publish it instead of replying where the conversation is
Seed counter-posts from other accounts	Never	Anyone connects the accounts, turning a complaint into a scandal

The last row is not a judgment call. Coordinated posting to bury criticism is brigading, platforms treat it as vote manipulation, and it converts a customer complaint into a policy violation with your name on it.

What happens if you delete or legally threaten a community thread?

On Reddit, Hacker News and every other community platform, the screenshot exists before the deletion completes. The people in the thread are the ones most motivated to preserve it, and a removed post leaves a labeled gap that reads as confirmation to everyone arriving later. You do not remove the thread. You replace it with a thread about the removal, and that one has a plot.

A legal threat is worse in one way: it is a document, and the recipient decides whether to publish it. The Techdirt post that named this pattern exists because a resort's lawyers sent a letter and then claimed the letter was off limits. Moderators are the other exposure, since removal requests land in modmail and several communities publish them as policy. Reddiquette and the sidebar rules give you your only leverage, a written rule the post breaks. Without that rule to quote, you are asking a volunteer to take your side against their own community.

When is removal the right call?

Removal is right when the post contains something no version of the conversation needs: a home address, a leaked document, an employee's personal information, an impersonation account, or a stolen asset. The platform's reporting flow covers those, moves faster than a lawyer, and keeps the request out of public view. File once and stop.

Everything else is a communications decision that belongs in a plan you wrote before the bad week. Set the escalation ladder in your crisis communications plan, name who may approve a legal threat, and default everything below that line to a reply in the thread. An online reputation management policy nobody can override under pressure beats a fast response.

How do you find the thread before it needs a response?

You monitor the platforms where threads start, on an interval short enough to reply while one is still moving. A complaint answered in hour one is a comment. The same complaint answered in week two, after a failed takedown, is a case study.

RedReplier watches Reddit, X, Bluesky, Facebook and Hacker News together, ranks what it finds by buying intent, and explains with AI why each mention was flagged, which is how an online reputation management workflow gets ahead of a thread. Email alerts arrive on your plan's interval.

Frequently asked questions

How much did Barbra Streisand sue for, and did she win?

She sued for $50 million in 2003 over Image 3850 of the California Coastal Records Project, and she lost. Los Angeles Superior Court struck the case under California's anti-SLAPP statute, tentatively on December 3 2003 and finally on December 31 2003, under case number SC077257. The court awarded $177,107.54 in fees and costs, with Adelman's share at $155,567.04 per the project's case page.

How many times was the photo downloaded before the lawsuit?

Six. Wikipedia's Streisand effect article records six downloads before the suit became public, two of them by Streisand's own attorneys, leaving four by anyone else. More than 420,000 people visited the site in the month after the case was reported. Four unrelated viewers against 420,000 visitors is the whole argument for the term.

Who invented the phrase "Streisand effect"?

Mike Masnick, on Techdirt on January 5 2005, in a post titled "Since When Is It Illegal To Just Mention A Trademark Online?" It covered a resort trying to stop Urinal.net from naming it. His line is "Let's call it the Streisand Effect."

Can you get a negative Reddit thread deleted?

Only the poster or the subreddit's moderators can remove a thread, and moderators act on their own rules. If the post breaks a written rule, quote it in modmail and ask once. If it does not, the removal request becomes a story the mod team can publish, and the thread returns as a screenshot.

Does the Streisand effect apply to small brands with no press coverage?

Yes, because the amplification is relative to your starting audience, not an absolute one. A thread with 40 upvotes reaching its whole subreddit and then a competitor's newsletter is the same mechanism at a smaller scale. The multiplier follows how aggressive the response looks next to the original complaint.

Is a DMCA takedown safe if the post uses my logo or screenshots?

Criticism that includes a logo or a product screenshot is what fair use covers, and a notice filed against it converts a product complaint into a story about you abusing copyright law. File one when someone republishes your actual assets: paid content, source code, a full article. Have a lawyer decide which case you are in, not a marketer.

Why was Kenneth Adelman photographing the California coastline?

Kenneth and Gabrielle Adelman started the California Coastal Records Project in 2002 to photograph the California shoreline every 500 feet and document coastal erosion. Image 3850, filed as "Streisand Estate, Malibu," was one frame in that survey, and Adelman had no idea whose roof he had caught. The lawsuit turned a routine erosion photo into the case that named the pattern.

How do you calculate the amplification of the Streisand effect?

Amplification is the audience after your response divided by the audience before it. In the origin case, 420,000 visitors divided by 6 downloads works out to about 70,000x, though the two figures count different things: image downloads against site visits. Treat that number as an order of magnitude rather than a measurement, and get your own starting figure from brand monitoring before you act.

What is brigading, and why does it matter for a Streisand effect response?

Brigading is coordinated posting from multiple accounts meant to bury or drown out criticism, and platforms treat it as vote manipulation. In the post's response table, seeding counter-posts from other accounts is the one option with no case where it works, because anyone who connects the accounts turns a complaint into a scandal. A single reply under your own name carries none of that risk.

How does RedReplier help you get ahead of a Streisand effect situation?

RedReplier watches Reddit, X, Bluesky, Facebook and Hacker News together and ranks what it finds by buying intent, so a thread surfaces while it's still small. It explains with AI why each mention was flagged, with email alerts on your plan's interval. A complaint answered in hour one stays a comment instead of becoming a case study after a failed takedown.