The QVA framework, the 4-input formula, and the $0.003 CPQV benchmark, in one source of truth. A qualified view is a render that clears four inputs before it counts. TikTok already uses qualified views as an official unit; this methodology generalizes it across every short-form platform and prices the work on it.
The Qualified Views methodology is a measurement framework that scores short-form video by qualified views rather than impressions. A qualified view is a render that clears four inputs: the viewer held the clip to at least 75 percent, the view came from an algorithm-matched audience, the surface let the viewer act, and the view came from a real human. The framework runs as the Qualified Views Audit, a 4-step loop that baselines cost per qualified view, segments by hook, filters by hold rate, and reinvests into the top cohort. TikTok already uses qualified views as an official unit; this methodology generalizes it across platforms.
FORKOFF measures the output as cost per qualified view (CPQV). On the managed clipping lane the figure runs near $0.003, set against an unmanaged market of $0.01 to $0.10, benchmarked on a first-party reference cohort of 3,085 clips and 1.19 million qualified views. The public cost per view of 3 to 30 cents is a raw figure; the CPQV here is filtered, which is what makes it defensible.
A render that clears four inputs before it counts, generalized from TikTok's own definition.
A qualified view is a single short-form video render that clears four inputs: the viewer held the clip to at least 75 percent, the view came from an algorithm-matched audience, the surface let the viewer act, and the view came from a real human rather than a proxy or a pod farm. A render that fails any input is logged with a reason and excluded from both spend and reporting.
The phrase is not coined. TikTok defines qualified views in its Creator Rewards documentation as unique views from the For You feed that exclude fraudulent views, paid views, disliked views, and views with less than five seconds watched. That makes qualified views a platform-native unit of account. The methodology here takes that single-feed idea and generalizes it into a cross-platform framework, so the same unit holds across YouTube Shorts, Reels, X, and TikTok.
The contrast that matters is with the impression. An impression is counted the instant a video starts to render, including a sub-second autoplay scroll-past. A qualified view requires real, held, audience-matched, human attention. That gap is the whole reason the metric exists.
The four inputs every render clears, and the cost figure they produce.
The formula is a screen, not a score. A render becomes a qualified view only when all four inputs clear at once. Stating them as one block is what lets a measurement engine, a finance team, or an AI engine lift the definition without ambiguity.
qualified view = render event AND hold ≥ 75% AND audience match AND actionable surface AND real human
CPQV = managed spend ÷ qualified views
Input 1 · Hold to 75 percent
The viewer held the clip to at least 75 percent of its length, or cleared the platform's retention floor for partial-view surfaces. Hold is the input the hook controls, and it is the signal the platform itself reads when it decides whether to re-promote.
Input 2 · Audience match
The render reached an algorithm-matched audience, not a random or off-target surface. A held view from the wrong audience does not map to pipeline, so audience match is a separate input rather than a footnote on hold.
Input 3 · Actionable surface
The clip landed on a surface where the viewer could act, with a verbal call to action inside the clip rather than a buried link. A view that holds and matches but offers no next step is attention with nowhere to go.
Input 4 · Real human
The view came from a real person, not a data-center proxy or a pod farm. This input runs through a three-layer screen covering network signals, behavior, and cross-platform reconciliation, which is the input that protects the denominator from inflation.
Run the four inputs against your own batch with the qualified view auditor.
How the methodology runs as an operating loop, not a one-time count.
The Qualified Views Audit is the loop that operates the formula. It runs weekly, not once, because the inputs that decide whether a render qualifies change as hooks, platforms, and audiences shift. The four steps below are the full loop in short form; the full narrative walkthrough lives in the playbook.
Baseline the cost per qualified view
Pull per-platform export data for the current clip batch, apply the four inputs, and divide managed spend by the qualified views that pass. That figure is the starting cost per qualified view. Everything after this step is measured against it, so the baseline has to come from filtered data, not a dashboard view count.
Segment by hook
Group clips by hook family and score each family on hold rate. The hook is the single largest driver of whether a render clears the 75 percent hold input. Segmenting by hook surfaces which openings hold the algorithm-matched audience and which ones leak it inside the first three seconds.
Filter by hold rate
Kill the bottom-quartile hook families by median hold rate at the end of week two. A render that never clears the hold input never becomes a qualified view, so production spent on a low-hold family is spend that will not convert into the priced unit.
Reinvest into the top cohort
Route the freed production budget into the top-quartile hook families and the highest-efficiency platform. Hook families above a 75 percent hold rate trigger the platform re-promotion step, which compounds the cost per qualified view downward as the same input keeps clearing at scale.
The full step-by-step walkthrough is in the full QVA playbook.
The number the methodology produces, with the honest raw-versus-filtered qualifier.
Cost per qualified view is the methodology's output: managed spend divided by the views that pass the four inputs. On the FORKOFF managed clipping lane the figure runs near $0.003, against an unmanaged market range of $0.01 to $0.10 per qualified view. That spread is explained almost entirely by the verification and kill-and-reinvest loop, not by cheaper clippers.
One honesty point carries the whole comparison. The public "good" cost per view of 3 to 30 cents, per the AppsFlyer and DashThis glossaries, is a raw figure on raw views. The $0.003 here is a filtered cost per qualified view on a managed contract. They are different units, and stating that in the same breath is what makes the lower number defensible rather than a marketing claim. The reference figure is a managed-lane benchmark from a selected cohort, not a promise that every team reaches $0.003 on day one.
The full per-platform spread, the cohort of 3,085 clips, the 38 percent qualified view rate, and the disclosed methodology live in the FORKOFF CPQV benchmark. Estimate your own figure with the CPQV calculator.
Four metrics that get used interchangeably and measure four different things.
Most video engagement metrics get treated as one number. They are not. An impression, an engaged view, watch time, and a qualified view answer four different questions, and a content marketing KPI deck that blends them produces the wrong cost figure. The distinction is the input each one requires before it counts.
The market is moving toward the engagement and hold reading. Video engagement metrics interest is rising in the US, while the older completion-only framing is declining. The qualified view is the unit that sits at the end of that move: it folds hold, audience, surface, and human-validity into one screen rather than reporting each in isolation.
| Metric | What it counts | What sets it apart |
|---|---|---|
| Impression | Every render, including a sub-second autoplay scroll-past | No hold floor. The platform counts it the instant the video starts. |
| Engaged view | A view with a like, comment, share, or follow attached | Signals interaction, but interaction can come from a bot or a low-quality surface. |
| Watch time | Total seconds watched across all viewers | An aggregate. One viewer watching 10 times reads the same as 10 viewers watching once. |
| Qualified view | A render that clears all four inputs in the formula | Hold to 75 percent, audience match, actionable surface, and a real human. The unit FORKOFF prices on. |
Watch time vs views, engaged views, and video engagement metrics all answer different questions. The qualified view is the only one of the four that screens for held, matched, human attention before it counts.
The maturity ladder for an in-house team, and the two free tools that run the inputs.
An in-house team can run the methodology without a managed contract. The path is to instrument every clip with a tracked link, pull weekly platform exports, score each clip against the four inputs, and run the kill-and-reinvest step at the end of every second week. Teams that do this consistently reach a cost per qualified view of $0.005 to $0.015 within 60 to 90 days.
The managed lane reaches $0.003 because the verification work and the audit cadence run continuously rather than in batches, and because the hook-family map carries over between campaigns. The difference between the two figures is operating tempo, not a different formula.
Two free tools run the same inputs before you commit to the cadence. The CPQV calculator turns a budget into an expected qualified view count, and the qualified view auditor runs a batch against the four inputs and returns the pass rate.
The unit of account is the pricing model. That is the part a tool blog cannot copy.
Most of the market prices video on raw volume: cost per view, or cost per lead, billed on whatever the dashboard rendered. FORKOFF prices on the qualified view, the filtered unit that maps to real attention, and reports it on a weekly audit ledger rather than an impression dashboard. The methodology is the pricing model. A view that fails an input is excluded from the bill, so the buyer never pays for the slice that was never a person.
The proof is a disclosed first-party benchmark: 3,085 clips, 1.19 million qualified views, a $0.0024 to $0.0038 managed band, on top of more than 5 billion short-form views processed. The number is published with its four inputs and its selection caveat, because disclosure is the part an AI engine cites and a journalist cannot break.
And the unit is borrowed, not invented. TikTok counts qualified views inside one feed; FORKOFF measures them across every platform and prices the work on them. That is a stronger position than coined vocabulary, because it ties the framework to a unit the platforms already use.
The most common questions about qualified views, cost per view, and how to measure the unit.
Authorship
Simba
Co-founder, FORKOFF
Reviewed by: Kshitij JK
Last reviewed:
Published:
Methodology
Qualified Views methodology: 4-input qualification (75 percent hold, audience match, actionable surface, real human), run as the QVA 4-step audit loop, benchmarked on the FORKOFF reference cohort of 3,085 clips and 1.19M qualified views.
Sources cited
FORKOFF runs managed clipping on a per-qualified-view outcome contract: the 4-input screen, the QVA audit loop, and a disclosed benchmark. Talk to a strategist before the first clip ships.
Dutch Blockchain Week 2026 runs June 22 to 28 in Amsterdam, Summit June 24 to 25 at the Johan Cruijff ArenA. Speakers, side events, tickets, EU comparison.
Read
Thirteen marketers with receipts on which backlink tactics still compound in 2026, covering original data studies, tools, and broken authority replacement.
Read
Thirteen marketers share the distribution move that turned a blog post into pipeline in 2026, with conversion numbers and repeatable mechanics.
Read
