How long does video approval actually take? The most direct published answer: eight days and more than three versions, per Filestage's survey of 366 marketers and creatives (2023). This page collects the citable statistics on video production, review friction, engagement, AI adoption, and the freelance market - every figure carries its source and year, and every source below was fetched and read in November 2026 before being cited. Where research is vendor-commissioned, it says so. Figures that circulated widely but could not be verified on a live source are not here - including one of the most-quoted video statistics on the internet, which turns out to come from a retired forecast (see the traffic section).
Adoption: who is making video
| Statistic | Source, year |
|---|---|
| 91% of businesses use video as a marketing tool | Wyzowl State of Video Marketing, 2026 edition (n=266, fieldwork late 2025) |
| 67% of non-users plan to start using video in 2026 | Wyzowl 2026 |
| 93% of video marketers call video an important part of their strategy | Wyzowl 2026 |
| 92% of marketers plan to spend the same or more on video in 2026 (49% more, 43% same, 8% less) | Wyzowl 2026 |
| 59% make videos in-house; 42% use an external creator in some capacity | Wyzowl 2026 |
| 76% of teams make at least one video per month | Wistia State of Video 2026 (900+ professionals surveyed; 13M+ videos and 79M viewing hours analyzed) |
| 943,305 business videos were created on Vidyard in 2024 - 88% more than the prior year; 37 videos per user on average, up 241% | Vidyard Video in Business Benchmark, AI Edition (Jan-Dec 2024 data) |
Effectiveness and budget
| Statistic | Source, year |
|---|---|
| 82% of marketers say video has given them a good ROI - down from 93% in the 2025 report | Wyzowl 2026 (the drop is stated on their own page) |
| 21% of marketers say short-form video delivers the highest ROI of any format | HubSpot State of Marketing, 2025 |
| ~40% of companies spent under $5,000 producing videos last year; just over 30% spent more | Wistia 2026 |
| 40% of teams plan to increase video spend - down from 57% in 2023; nearly half are keeping budgets flat | Wistia 2026 |
The 82%-down-from-93% ROI line deserves a beat: adoption rose while perceived ROI fell in the same survey year. Vendor surveys measure their own ecosystems, but the two series moving in opposite directions is the most honest sentence in the dataset.
Engagement: what happens after publish
| Statistic | Source, year |
|---|---|
| Videos under a minute average a 52% engagement rate | Wistia 2026 (platform data) |
| Videos under one minute hold 65% of viewers to the end; over 20 minutes, only 20% remain | Vidyard Benchmark, 2024 data |
| Videos of 5-30 minutes average 9% click-through; engagement drops 11% once videos cross 30 minutes | Wistia 2026 |
| Lead-gen forms inside 5-30 minute videos see 53-55% CTR; inside 60+ minute videos, 67-75% | Wistia 2026 |
| Videos under a minute average 16 seconds of watch time | Wistia 2024, via HubSpot's roundup (updated Mar 2025) |
| 3-5 minute videos average 43% engagement - but how-to videos of that length hit 74% | Wistia 2024, via HubSpot |
| 81% of teams share videos on LinkedIn; YouTube follows at 76% | Wistia 2026 |
| Vertical HD uploads grew 24% year over year | Wistia 2026 |
Consumption: the traffic numbers, corrected
The most-cited video statistic on the internet - Cisco's "video will be 82% of internet traffic" - comes from a forecast series Cisco retired; quoting it today means citing a projection that has already expired. The live, measured numbers come from Ericsson's Mobility Report, which measures actual carrier networks:
| Statistic | Source, year |
|---|---|
| Video is around 50% of mobile traffic (up to 60% on downlink); social video is 70-80% of all mobile video traffic | Ericsson Mobility Report article "Short-form dominates video traffic," Nov 2025 (measured on four sampled European networks) |
| Mobile data traffic: 146 exabytes/month at end-2025, forecast to reach 328 EB/month by 2031; 22 GB per smartphone per month globally | Ericsson Mobility Forecast, June 2026 edition |
| Mobile traffic grew 23% between Q2 2025 and Q2 2026; half of it rides 5G, forecast at 85% by 2031 | Ericsson Mobility Report, June 2026 |
The review loop: where production time actually goes
This is the section most "video statistics" pages skip, and it is where the operational pain lives. Two vendor surveys dominate - both flagged as vendor research, both with disclosed methods:
| Statistic | Source, year |
|---|---|
| Average creative sign-off takes eight days and over three versions | Filestage State of Creative Collaboration (survey of 366 marketers and creatives, 2023) |
| Time to approval by org type: freelancers 4 days, startups 9 days, brands 10 days | Filestage, 2023 |
| 90% of account managers say waiting for feedback slows them down | Filestage, 2023 |
| 75% of creative collaboration happens remotely; 65% of teams are hybrid, up 33% year on year | Filestage, 2023 |
| 48% spend at least five hours per month chasing feedback | Ziflow State of Creative Workflow, with the American Marketing Association (2023) |
| 60% frequently spend time explaining to external stakeholders how to give feedback | Ziflow/AMA 2023 |
| 57% say 3-5 versions occur before a project is done; 25% produce more than six | Ziflow/AMA 2023 |
| 43% say feedback on outdated versions is an occasional or regular problem | Ziflow/AMA 2023 |
| 30% absorbed extra project costs from bad feedback or missed deadlines; 82% say at least 5% of projects end because of bad feedback | Ziflow/AMA 2023 |
| Only 28% devote over half their workday to actual creative work - the rest is coordination | Ziflow/AMA 2023 |
| 70% work on more than seven projects or campaigns per week | Ziflow/AMA 2023 |
The two surveys agree from independent samples: rounds are counted in threes and fives, and the bottleneck is not creative capacity - it is feedback latency and coordination. That is the quantitative case for the round-discipline practices in the review checklist and the version-count expectations in the revision-rounds guide; the numbers say the loop, not the editing, is the cost center.
What platform vendors claim about fixing it
| Claimed result | Source, year |
|---|---|
| 461% three-year ROI and $11.0M NPV (medium scenario) for Adobe creative solutions including Frame.io | Forrester Total Economic Impact study, commissioned by Adobe, Jan 2025 (11 interviewees, 6 organizations) |
| Reduces time spent reviewing and fixing assets by as much as 75% | Same Forrester TEI (Jan 2025) |
| One interviewee avoided four to five feedback cycles on variant content | Same Forrester TEI (Jan 2025) |
Read commissioned research as an upper bound. Forrester TEI methodology models a composite organization from interviews with customers selected by the commissioning vendor; the numbers describe a ceiling, not a median. Note also that the frequently-quoted 2021 Frame.io TEI is no longer publicly reachable - the January 2025 Adobe study is the live citation.
AI in video production
| Statistic | Source, year |
|---|---|
| 63% of video marketers have used AI to create or edit video, up from 51% - the biggest year-over-year jump the Wyzowl report measures | Wyzowl 2026 vs 2025 |
| 86% of creators actively use generative AI; 55% for editing/upscaling/enhancement, 52% for generating new assets | Adobe Creators' Toolkit Report with The Harris Poll (16,000+ creators, Sept 2025 fieldwork) |
| 76% of creators say AI accelerated their business or follower growth; 70% are optimistic about agentic AI | Adobe/Harris Poll, 2025 |
| About two-thirds of podcasters and video creators have used generative AI in production; 78% are likely to keep using it | Descript survey with Ipsos (1,004 creators, updated Jul 2026) |
| More than a third of video teams already use AI in their workflow; daily producers are 58% more likely to use it | Wistia 2026 |
The freelance video market
| Statistic | Source, year |
|---|---|
| Film and video editors: $75,420 median annual pay; 39,400 jobs; 4% projected growth 2025-2035; ~5,600 openings per year | U.S. Bureau of Labor Statistics, Occupational Outlook Handbook (wages May 2025; page updated Aug 2026) |
| 29,240 film and video editors employed; $66,600 median; 90th percentile $154,480 | BLS OEWS, May 2023 (SOC 27-4032) |
| 39% of U.S. workers freelanced in 2026, up from 35%; freelancers earned $1.5 trillion in 2024 | Upwork Future Workforce Index 2026 (fieldwork Mar-Apr 2026; earnings figure from Upwork's Apr 2025 report) |
| Freelancers adopt AI agents at 41% versus 24% for full-time employees; 89% are excited about AI reshaping their services | Upwork 2026 |
The editor market context feeds directly into the rates conversation: the editor-rates guide uses the same BLS anchoring, and the staff-vs-freelance comparison builds its break-even math on these bands.
How to read video statistics honestly
Four rules that separate usable numbers from blog-fodder:
- Check the sample and the date. Wyzowl's 2026 edition surveyed 266 businesses; Filestage's surveyed 366; Adobe's polled 16,000+ creators. A percentage without an n and a year is decoration.
- Know the species. Annual surveys (Wyzowl, Wistia, Upwork), platform-data benchmarks (Vidyard, Wistia engagement), carrier measurements (Ericsson), commissioned economic models (Forrester TEI), and government data (BLS) carry very different weight - this page flags each.
- Beware the retired citation. Cisco's "82% of traffic" forecast series is retired, yet the number still tops search results. If a stats page cites it without noticing, its other figures deserve suspicion too.
- Treat vendor research as directional. Wyzowl, Wistia, Ziflow, Filestage, Adobe, and Upwork all research their own markets. The findings are still usable - disclosed methods, consistent series - but the honest framing is "vendor-observed," not "independent."
And the operational takeaway for teams living inside these numbers: the review loop is measurable (days to approval, versions per project, hours chasing feedback), which means it is improvable. The remote review guide covers the loop itself, and AI feedback summarization covers the newest lever on the coordination cost those surveys keep measuring.
FAQ
How long does video approval take on average?
The most direct published figure is eight days and more than three versions to reach sign-off, from Filestage's 2023 survey of 366 marketers and creatives. The same survey found freelancers approve fastest (four days) and brands slowest (ten days).
How many revisions does a typical video go through?
Ziflow's 2023 report with the American Marketing Association found 57% of teams say three to five versions occur before a project is done, and 25% produce more than six. Filestage's independent survey agrees on "more than three versions" as the median experience.
What percentage of businesses use video marketing?
91% per Wyzowl's 2026 survey (fieldwork late 2025, n=266), with 67% of non-users planning to start. The series has run for twelve years, which makes it the standard adoption citation.
What percentage of internet traffic is video?
For mobile networks - the honest, measured scope - Ericsson's November 2025 Mobility Report article puts video at around 50% of mobile traffic (up to 60% of downlink), with social video at 70-80% of mobile video traffic. The famous "82% of all internet traffic" figure comes from Cisco's retired forecast series and should no longer be quoted as current.
Are video marketing statistics reliable?
The major annual surveys (Wyzowl, Wistia, Vidyard, HubSpot) are reliable as directionally useful vendor research with disclosed methods and sample sizes - not as independent measurement. Forrester TEI figures are commissioned models and belong in any deck labeled as an upper bound. Government data (BLS) and carrier measurements (Ericsson) are the most neutral numbers in the set.