Playbook
SaaS explainer videos: the five you actually need
Most SaaS companies buy one expensive video for the homepage and then have nothing for the eight pages where a prospect actually decides. This is about which videos to make, how long each one gets to be, and why the second question changes the first.
Published 10 September 2026 · 9 min read
The short version
- You need five kinds of video, not one. Homepage hero, feature explainers, onboarding, a sales walkthrough, and ad cutdowns. They have different lengths, different jobs and different budgets.
- Only one of them justifies an agency. The hero. The other four are short, numerous and replaced often — exactly the profile that makes per-video production sensible and a four-week studio cycle absurd.
- The feature pages are where deals are actually decided, and they are usually the pages with no video at all, because the budget went on the homepage.
- Length is set by the page, not by the script. 90 seconds on the homepage, 40 on a feature page, under 30 in product, 6 to earn a paid-social view.
Why one hero video is the wrong shape for SaaS
A SaaS product is not one idea. It is one promise and then eight to twenty capabilities that each need explaining to a different person at a different moment — the evaluator on the pricing page, the admin worrying about SSO, the new user on day one, the champion forwarding something internally.
A single 90-second film cannot serve all of those, and the attempt produces the video everybody recognises: a soft opening about modern teams, a montage of abstract shapes, and no specific claim anybody could repeat afterwards. It is generic because it is being asked to be relevant to everyone.
Explainer videos are among the most-produced formats in B2B marketing — Wyzowl’s annual survey put them second behind social video in 2026, on a sample of 266 respondents, which is worth knowing and too small to build a strategy on. The useful question is not whether to make one. It is which five to make.
The five videos, and what each one is for
| Video | Length | Its one job | Route worth paying for |
|---|---|---|---|
| Homepage hero | 60–90s | Make a stranger recognise their own problem and believe you have thought about it. | The one case for a studio. Judged by thousands of prospects, lives two years, amortises to nothing. |
| Feature explainer | 30–45s | Show one capability to somebody already interested in your category. | Per-video software. You need eight to twenty of these and they change when the feature does. |
| Onboarding / in-app | 15–30s each | Get a new user through one step without reading documentation. | Per-video software. Short, numerous, and rewritten every time the UI moves. |
| Sales walkthrough | 2–3 min | Answer the same five objections your team answers on every call. | Software, and re-made per segment. A studio version dates the moment pricing changes. |
| Ad cutdowns | 6–15s | Earn the next three seconds. Nothing more. | Software, in volume. You are testing hooks, so you need ten versions and not one good one. |
One agency video and four categories of software video is the shape that fits an actual SaaS company. The common alternative — five agency videos — is a $30,000 line item that arrives after the quarter it was meant to support.
How long each one gets to be
Length is a property of the page, not of how much you have to say. The same script cut to 90 seconds and to 40 is two different videos and both are correct in their own slot.
The homepage hero: 60–90 seconds
Long enough to name a problem, show the shape of an answer and make one specific claim. Past 90 seconds you are describing features to someone who has not yet agreed they have the problem.
The feature explainer: 30–45 seconds
The visitor already knows what your product is — they clicked a feature link. Re-introducing the company here wastes the only twelve seconds you were given. Start inside the feature.
Onboarding: under 30 seconds per step
Anything longer and the user tries to do the thing while the video is still playing, which is the correct instinct and means the video failed. One step, one video.
Ads: six seconds decides it
The rest of the ad only exists for the people the first six seconds kept. This is why ad video should be produced in tens: you are not making a video, you are testing hooks, and you cannot test hooks four weeks at a time.
The library is what changes the arithmetic
One video at agency prices is a defensible purchase. Twenty are not, and twenty is the real requirement. This is the whole argument in one table.
| Route | 1 hero | + 15 short videos | Time to change one |
|---|---|---|---|
| Mid-market studio | $5,000 | $52,500–$80,000 | 2–3 weeks |
| Freelancer | $2,500 | $25,000–$40,000 | 1–2 weeks |
| Subscription tool | — | $350–$1,200/yr | Hours |
| Per-video (ours) | — | €45 for 16 | Minutes |
The point of that table is not that software is cheaper. It is that the studio column makes the library impossible, so companies on that route quietly stop after the hero video and leave nineteen pages with nothing on them. The cost of an agency is not the invoice — it is the videos you then never make.
What a SaaS explainer has to show that a generic tool cannot
Two things separate a video about your product from a video about software in general, and both are places where tools quietly fail.
- Your interface. For a feature explainer, an abstract illustration of “a dashboard” is worse than nothing: it tells the viewer you either could not show the product or did not want to. Whatever you use has to be able to put the actual thing on screen.
- Your numbers, unaltered. “Cuts onboarding from 6 days to 4 hours” is the only sentence in the video anybody repeats. Be careful here: tools that generate charts from a prompt will happily invent plausible figures nobody supplied, and a wrong number burned into a video cannot be corrected after it ships. Ask what a tool does when the brief contains no numbers. The right answer is that it draws no chart.
Vertical, widescreen, or both
Widescreen for anything embedded in a page — homepage, feature pages, in-app, docs. Vertical for paid social, and for the cutdowns, where the whole point is to occupy a phone screen.
The mistake to avoid is cropping one into the other. A 16:9 video re-cropped to 9:16 loses the edges of every frame, which is where text sits, and a vertical video letterboxed into a page looks like a mistake. If a video needs to run in both places it should be laid out for both — which is a production requirement, and one worth checking before you commission anything.
Where ExplainVids fits
We are the per-video row: €3 for one finished film, up to two minutes, vertical or widescreen, no subscription. For the four categories in the table that are short, numerous and frequently replaced, that is the shape of the problem.
Paste a feature page’s URL and it reads the page and writes the script from what is actually on it, in your site’s own colours — which for a feature-explainer library is usually faster than briefing each one. Charts are drawn only from figures you put in the brief.
For the homepage hero of a product you have raised money against: hire a studio. That video is judged by more people than any other asset you own, and the honest split is the one above.
Pick your least-loved feature page, paste its URL, and see what a 40-second version of it looks like. Three euros, about two minutes, and it is the same decision the rest of the library turns on.
Make one and seeQuestions
How long should a SaaS explainer video be?
It depends entirely on where it sits. A homepage hero has about 60 to 90 seconds before the viewer leaves; a feature-page explainer should be 30 to 45 seconds because the visitor already knows what your product is; an in-app onboarding video should be under 30 seconds per step; a paid-social ad has roughly 6 seconds to earn the rest. The single most common mistake is making one 2-minute video and putting it in all four places.
How much does a SaaS explainer video cost?
A custom animated one from a studio runs $3,500 to $7,000 for 60 seconds in the mid-market, and $7,000 to $15,000 from a premium B2B agency. Software you drive yourself is two orders of magnitude below that — from around €3 for a single finished video. The full breakdown, with every figure attributed to the source that published it, is in our explainer video cost article.
What makes a B2B explainer video different from a consumer one?
The viewer is usually not the buyer, and the video has to survive being forwarded. A consumer video persuades the person watching it; a B2B video has to make the person watching it able to explain your product to someone else — their manager, their security team, their CFO. That means naming the problem in the first eight seconds, being specific about what the product does, and not relying on tone or humour to carry the message.
Should a SaaS explainer video show the actual product?
For a feature explainer or an onboarding video, yes — abstract illustration of a UI is worse than the UI. For a homepage hero, not necessarily: at that point the visitor is deciding whether the problem you name is the problem they have, and screenshots of an interface they cannot yet read do not help with that. Show the problem on the homepage and the product on the feature pages.
Do I need a different video for each feature?
If you have feature pages, yes, and this is where most SaaS video budgets are spent wrong. One expensive hero video and no feature videos is a worse outcome than a plainer hero and eight 40-second feature explainers, because the feature pages are where a prospect who is already interested goes to be convinced.
Can AI make a SaaS explainer video that is actually usable?
For feature explainers, onboarding steps, changelog videos and internal walkthroughs — the short, specific, numerous ones — yes, and the economics are not close. For the homepage hero of a funded product that thousands of prospects will judge you on, hire people. The honest line is that AI tools are good at the videos you need many of, and not yet good at the one you need to be perfect.
What should a SaaS explainer video say in the first ten seconds?
The problem, in the customer's own words, and not the product name. A viewer who does not recognise their own situation in the first two sentences has no reason to watch the third. The product should arrive as the answer to something already stated, which is also the structure that survives being forwarded to somebody who has never heard of you.
Related: what an explainer video costs for the prices behind the table above, and explainer video software compared for which tool makes which kind of video.