The declared ROI on video has just lost eleven points in a year. In 2025, 93% of video marketers said they got a good return on investment from it. In 2026, only 82% do (Wyzowl, State of Video Marketing 2026, 266 respondents). Eleven points is serious. And yet that figure does not say what many are about to read into it.<\/p>\n\n\n\n
It does not say that video is running out of steam. It says video is leaving its hype phase.<\/p>\n\n\n\n
The nuance changes everything about how you will spend your next production budget. Because the real question is no longer \u201cshould we do video\u201d, to which the figures still answer yes, but \u201chow many videos, and for what\u201d. And there, the answer has changed.<\/p>\n\n\n\n
Let us kill the misreading first. Video remains, by a long way, the best-performing proof format in B2B. 83% of marketers say it directly increased their sales, and 52% of B2B marketers cite it as the content type with the highest ROI compared with every other format (Wyzowl 2026 and Vidico 2025, via Levitate Media). A customer convinced to buy after watching a video is still the most reliable mechanism in digital marketing. Nobody serious will advise you to stop producing it.<\/p>\n\n\n\n
So why is the declared ROI falling?<\/p>\n\n\n\n
Because when a format becomes available to everyone, it stops being an advantage. That is mechanical. In 2018, an SME publishing a polished video stood out from competitors who had none. In 2026, your competitors publish too, your suppliers publish, your baker publishes. Video is no longer the differentiator, it is the minimum. The fall from 93 to 82% does not measure a loss of quality in video. It measures the end of the format premium, the moment when the mere presence of a video no longer makes the difference.<\/p>\n\n\n\n
The production data confirms this normalisation. Wistia’s State of Video 2026 report, drawing on more than 900 professionals and the analysis of over 13 million videos, shows a market that has plateaued: nearly 40% of companies spent less than 5,000 dollars per video last year. Most of the volume produced today is made on a contained budget, not a premium one. Many videos, made fast, made cheaply, all looking alike.<\/p>\n\n\n\n
It is precisely that context that reopens a space for those who choose the opposite.<\/p>\n\n\n\n
Here is the mistake the ROI drop should help you avoid, and which it will nonetheless push many companies to make.<\/p>\n\n\n\n
The reflex, when a channel yields less, is to do more of it. Video ROI is falling? Let us double the cadence. Move from two videos a month to eight. Flood LinkedIn with short formats. That reasoning has an apparent logic: more content, more chance of reaching someone. Above all it has a flaw, which is that it confuses activity with results.<\/p>\n\n\n\n
Multiplying weak videos does not produce a strong result. It produces eight weak videos, a scattered budget, and a team exhausted feeding a machine that measures nothing. The view counter goes up, the qualified-enquiry counter does not move, and by the end of the quarter nobody can say which video did what.<\/p>\n\n\n\n
The brief before the brief:<\/strong> if a supplier offers you a \u201c12 videos a year package\u201d without first asking what each one is supposed to trigger, they are selling you a cadence, not a strategy. Cadence is easy to invoice. Results are not. That is precisely what separates an order executed from a project properly scoped.<\/p>\n\n\n\n Volume as an objective is a comfort metric. It reassures because it can be counted. But something being countable does not make it count.<\/p>\n\n\n\n The alternative rests on a discipline simple to state and demanding to hold: every video answers a precise, measurable business objective, decided before the shoot.<\/p>\n\n\n\n Take a case. A services SME struggling to turn inbound calls into signed contracts does not need eight LinkedIn clips. It needs a single, well-made video that answers its prospects’ number one objection at the exact moment it arises in the sales cycle. A video placed on the right page, sent at the right moment, calibrated on a single KPI: the rate at which leads convert into qualified meetings. One, not eight. But that one, we know why it exists and we will know whether it worked.<\/p>\n\n\n\n That is the difference between a deliverable and a result. A deliverable is not a result.<\/strong> A delivered video is a file. A video that moves a number on your dashboard is a lever. The first is ordered, the second is scoped.<\/p>\n\n\n\n In practice, producing right imposes three decisions upstream, before any camera:<\/p>\n\n\n\n That discipline has a useful side effect on the format itself. 71% of marketers judge videos of 30 seconds to 2 minutes the most effective (Wyzowl 2026). When each video has a precise job to do, it naturally becomes shorter and denser, because it is no longer trying to say everything. It is trying to do one thing well.<\/p>\n\n\n\n One legitimate objection remains, especially if you manage the budgets: producing fewer but better-calibrated videos, is that not an elegant way to justify more expensive productions?<\/p>\n\n\n\n Let us look at the figures rather than the adjectives. When 40% of videos are produced under 5,000 dollars and all look alike, scarcity no longer lies in having a video. It lies in the quality of the storytelling and the precision of the targeting. An impact documentary that genuinely tells a story, native subtitling in three languages that opens real markets, a motion design that clarifies a complex offer in ninety seconds: that is what low-cost volume does not make. Not because it is \u201cpremium\u201d in the label sense, but because it takes scoping, time and a craft that industrial cadence leaves no room to practise.<\/p>\n\n\n\n For an organisation with substantial one-off budgets, the calculation becomes even clearer. Rather than sprinkling the sum across a series of interchangeable pieces, concentrating it on a centrepiece that serves an identified objective, can be shown over time and adapted afterwards, produces an asset. Sprinkling produces flow. Flow is forgotten, an asset works.<\/p>\n\n\n\n This shift also has a channel logic. LinkedIn has become the leading place for sharing videos, cited by 8 teams out of 10, up more than 30% since 2024 (Wistia 2026). And LinkedIn is B2B ground, professional, where an audience of decision-makers does not reward easy virality but relevance. An audience that tells the worked-on from the dashed-off. Exactly the audience that penalises hollow volume and rewards the right video.<\/p>\n\n\n\n The end of the format premium is not bad news. It is a clarification. As long as video was an automatic advantage, having one was enough. Now that it has become the norm, you need to know what to do with it. Those who read the ROI drop as a signal to produce more will scatter themselves. Those who read it as a signal to produce better will take the ground left free.<\/p>\n\n\n\n Video has lost none of its power as proof. What has changed is that it no longer forgives a lack of scoping. A video with no quantified objective, no place in the journey and no measurement is a cost today. A video that answers those three questions is still the best investment in your arsenal.<\/p>\n\n\n\n Before ordering your next batch of videos, ask yourself the one question that really decides: which of these videos could you delete without anyone noticing? Do not produce that one. Put the budget on the one that would be missed.<\/p>\n\n\n\nProduce less, but right<\/h2>\n\n\n\n
\n
Why the high end becomes rational again<\/h2>\n\n\n\n
What it changes for your next budget<\/h2>\n\n\n\n