Fake Views in Clipping: How to Spot Botted Views and Check If a Campaign Is Real

What counts as a fake view in a clipping campaign?

A clipping campaign buys distribution. You pay for a number of views delivered across creator-published short-form posts on TikTok, Instagram Reels, YouTube Shorts and X. A view is real when a human being was served the video in their feed. It is fake when the counter went up without that happening, or when it went up for a reason you did not buy.

Five distinct practices get lumped together under the phrase "fake views", and they fail in different ways.

Boosting is not fraud in every context. Plenty of campaigns legitimately include paid amplification. It becomes fraud when the buyer is told they are purchasing organic creator distribution and is charged a CPM for reach the clipper bought more cheaply elsewhere.

Why do clippers buy views in the first place?

This is the part most coverage skips, and it is the whole story. In a clipping campaign the clipper is paid per 1,000 views on the clips they post. That model is exactly why clipping works: it ties payment to delivered reach instead of hours worked, which is what lets the channel run at roughly $1 to $5 per thousand views while paid social commonly runs $15 to $40.

It also creates one entirely predictable incentive. If a clipper earns a few dollars per 1,000 views and a reseller will sell 1,000 views for a fraction of that, then buying views is a positive-margin trade. Making a genuinely good clip is slower, harder and far less certain than simply purchasing the number. The arithmetic does the rest.

Nothing about that is unique to clipping. Every pay-per-unit media model, from display impressions to affiliate clicks, has produced the same behaviour. What is specific to clipping is who is positioned to stop it, and here is the uncomfortable part: a provider that pays per view and does not actively police the number is not neutral. Fake views inflate reported delivery, and inflated delivery flatters the provider's own campaign report. Without enforcement, the provider's incentives point the same way as the botter's.

How can a buyer spot fake views without any special tools?

Most of this you can do from the public post and a campaign dashboard. The single most important rule: judge a clip against the posting account's own back catalogue, not against a universal benchmark. Ratios vary enormously by niche, format, platform and audience age, so a page's own history is the only honest control group.

Ratio ranges are directional heuristics for short-form video, not thresholds. A real clip can have low likes and a real page can have a slow week. One signal is noise; three signals together on the same post, or one signal repeating across every post from the same account, is a pattern.

The two signals that are hardest to fake

Velocity. Organic short-form distribution is bursty by design. The platform tests a clip on a small audience and, if retention holds, pushes it to progressively larger ones. That produces a curve with a steep front, a long decaying tail, and sometimes a second bump days later on a re-test. Bought views behave nothing like that: a delivery service fills an order at a roughly constant rate and then stops, so the curve is a ramp that flatlines at the purchased number. If your dashboard reports often enough to show the shape (Spade's updates every four hours), the shape is usually the fastest tell a buyer has.

Watch time. Bot traffic is optimised to trigger a counted view at the lowest possible cost, which in practice means it does not watch. So total watch hours stop tracking the view count. A clip with 500,000 views and an average watch time under a second is not a clip 500,000 people saw. Ask for average view duration alongside view totals on any platform where it is reported.

What do fake views actually cost a brand?

None of it shows up on the invoice, which is exactly why it persists.

• An inflated effective CPM. Take 5,000,000 delivered views at an illustrative industry CPM of $3, so $15,000 of spend. If 20% of those views are manufactured, you bought 4,000,000 real views for $15,000, an effective CPM of $3.75. You did not get cheap reach. You paid a 25% premium and were shown a discount.
• Polluted analytics and attribution. Bots do not click, save, search, stream, install or buy. When part of the delivery is fake, your view total rises while every downstream metric stays flat, and you cannot tell whether your creative missed or your reach never existed. Worse, those polluted numbers become the benchmark you plan the next campaign against.
• Platform penalties and reach suppression. Platforms detect inauthentic engagement and respond by removing views, limiting distribution or restricting the account. Views counted this week can be scrubbed next week, so a campaign can be reported as delivered and then quietly shrink. If the inauthentic activity clusters on your sound, your hashtag or your branded account, the suppression can attach to the asset rather than to the clipper who caused it.
• Brand-safety and reporting exposure. Manufactured engagement comes from accounts with no real identity and no moderation, which puts your brand next to spam networks. And if you are a label, an agency or an in-house team reporting campaign numbers upward, passing on fake views stops being a media problem and becomes a governance one.

What does a real defence against botted views look like?

Six components, in rough order of how much difference they make. None of them are exotic; what is rare is a provider doing all six.

1. Every clip reviewed, on a fixed cadence. Not a sample. Botting concentrates in a small number of accounts, so a sample can miss all of it and still come back clean.
2. Automated bot-pattern detection running alongside the humans. Posts analysed for bot patterns, fake engagement and suspicious velocity at both account and post level, because no team can eyeball velocity curves across tens of thousands of posts.
3. Rejection that removes the clip from billing. This is the only part that changes behaviour. If a rejected clip still counts toward delivered views, the buyer funds the fraud and the clipper keeps the upside. If rejection means the views do not count and do not get paid, the economics that made botting attractive collapse.
4. Post-publication monitoring and takedown. Views are often bought after a post goes up, so a single check at publication proves nothing. The campaign has to be watched while it runs, and clips have to be pullable after they are live.
5. Direct creator relationships instead of an open marketplace. A network of known, vetted clippers with a payment history is governable. An open marketplace where anyone can claim a brief is not: creating a new fake identity costs nothing.
6. Pricing on delivered and verified views. A CPM charged on views that survived review, with no retainer, points the provider at the same outcome as the buyer rather than at reported volume.

How Spade runs it: Spade's moderation team goes through every clip in a campaign once every day. A clip is rejected if it does not fit the brief and guidelines, or if it shows botted or boosted views. Rejected clips do not count toward delivered views, so the client does not pay for them. Campaigns are monitored while they run and bad clips can be pulled after they have gone live. Automated detection analyses posts for bot patterns, fake engagement and suspicious velocity. Across a network of 60,000+ vetted clippers and 80,000+ niche pages delivering 2B+ views per month, the measured authentic-view rate is 98.5%.

For context: Spade Group is a bootstrapped company founded in Vancouver in 2022, now 40+ people, with 1,000+ campaigns run for 300+ brands, including work with RCA, 10K Projects, Republic, Island, Interscope, Encore, Darkroom, Atlantic Records, Capitol Records, Warner Records, Epic Records and 88rising. Campaigns launch in 6 to 24 hours, pricing is CPM on guaranteed views delivered with no retainers and no upfront cost, and the dashboard updates every 4 hours.

What is the honest trade-off of policing fake views?

Policing costs volume, and any provider who tells you otherwise is selling. Every rejected clip is reach that does not appear on your report. A provider that rejects nothing will always be able to show a bigger delivered number for the same spend than a provider that rejects something. That gap is not performance. It is accounting.

So compare providers on verified views, not reported views, and treat a perfect score as a warning rather than a boast. An authentic-view rate of 98.5% is a number produced by looking. A claim of 100% is usually a number produced by not looking.

There is a second trade-off worth naming plainly. Daily review means a bad clip can be live for up to a day before it is caught. That is the real limit of post-publication moderation at network scale, and any provider promising that nothing bad ever reaches a feed is describing a system that does not exist. What actually matters is the interval between publication and review, whether the clip can be pulled once found, and whether you are billed for it in the meantime.

What should you ask any clipping provider?

These are vendor-neutral. Ask them of everyone you shortlist, Spade included.

1. How often is every clip in my campaign reviewed, and by a human, a script, or both?
2. What happens to a clip you reject: does it come off my invoice, or do I still pay for the views it generated?
3. What is your measured authentic-view rate across the network, and how do you measure it?
4. Can I see account-level reporting, showing which page posted what and how it performed, rather than a single aggregate total?
5. Which specific signals does your bot detection use?
6. Do you monitor clips after they are live, and can you get one taken down?
7. Where do your creators come from: direct relationships and vetting, or open sign-up?
8. Are boosted or promoted posts ever counted as delivered views, and how are they labelled?
9. What happens if the platform removes views after you have already reported them to me?

The answers matter less than whether the provider will put them in writing. A provider who commits to a review cadence, a rejection policy and a measured authenticity rate in a contract is telling you something real. A provider who answers all nine convincingly on a call and none of them on paper is also telling you something.

Bottom line

Clipping is the cheapest reach per unit available to most marketing teams, at roughly $1 to $5 per thousand views against $15 to $40 for paid social. That price advantage is only real if the views are.

The per-view model that makes clipping efficient is the same model that pays clippers to fake the number, so the useful question to put to a provider is not whether botting exists in their network. It does, in every network of any size. The question is how fast a botted clip is found, what happens to it, and whether you are still billed for it.

Frequently asked questions

How can you tell if clipping views are fake?

Check five things on the public post and in the campaign dashboard: the like-to-view ratio (under about 0.5% is a flag), the comment-to-view ratio and the quality of the comments, the view velocity curve (bought views arrive at a constant rate and stop dead at a round number, real ones spike then decay), average watch time against the view count, and whether audience geography matches the markets the campaign actually targeted. One signal on its own is noise. Three signals together on the same post, or one signal repeating across every post from the same account, is a pattern worth challenging.

Why do clippers buy views?

Because clippers are paid per 1,000 views. If a reseller sells views for less than the clipper earns per 1,000 views, buying views is a profitable trade, and it is faster and far more certain than making a clip that genuinely performs. That is the structural problem at the centre of clipping fraud: the payment model that aligns buyer and clipper around delivered reach also pays the clipper to manufacture it.

Are clipping agency views real?

It depends entirely on whether the agency polices its network and what happens when it finds a fake. Ask how often every clip is reviewed, what the measured authentic-view rate is, and whether a rejected clip comes off the invoice. Spade Clipping reviews every clip in a campaign once every day, rejects clips that show botted or boosted views, does not count rejected clips toward delivered views, and measures a 98.5% authentic-view rate across its network.

What do fake views cost a brand?

Four things. A higher effective CPM for reach that never existed: at 20% fake delivery, a $3 CPM is really $3.75 per thousand real views. Analytics and attribution you cannot trust, because bots never click, save, stream or convert. Exposure to platform penalties and reach suppression when inauthentic engagement attaches to your sound, hashtag or account. And brand-safety risk from adjacency to spam networks.

How does Spade Clipping stop botted views?

Spade's moderation team goes through every clip in a campaign once every day and rejects any clip that does not fit the brief and guidelines or that shows botted or boosted views. Rejected clips do not count toward delivered views, so the client does not pay for them. Automated detection analyses posts for bot patterns, fake engagement and suspicious velocity, campaigns are monitored while they run, and clips can be pulled after they have gone live. The measured authentic-view rate across the network is 98.5%.