An explainer does not explain everything; it removes exactly one confusion that is currently stopping a buyer from asking for a demo. This guide is written for marketing, brand and founder teams in Pune who are about to commit real budget and want to make that decision from evidence rather than from a showreel. Everything below comes from projects run in this city — the same IT parks, the same shop floors, the same permission queues and the same twelve-hour call sheets.
The failure mode we see most often is writing a feature tour instead of a single mechanism, and pacing it at 190 words per minute so nothing lands. It is rarely a competence problem. It is a sequencing problem: decisions that cost nothing on paper get deferred until the point where they cost a shoot day, and by then the only available fix is money. The chapters below put those decisions back in the order that keeps them cheap.
By the end you should be able to write or judge a 90-second explainer script against a five-beat structure and a real word budget. Where numbers appear they are real Pune ranges — for explainer video production, the honest band is ₹35k to ₹2.5L, and the guide explains what actually moves you between those two ends rather than pretending a single figure exists.
The mistake this guide prevents
writing a feature tour instead of a single mechanism, and pacing it at 190 words per minute so nothing lands
Beat one: the problem, in the viewer's words
The first fifteen words decide whether the rest is watched. They should describe a situation the viewer recognises immediately, using the vocabulary they would use themselves, not your category language. 'Reconciling three spreadsheets before every board meeting' beats 'fragmented financial data workflows'. This beat runs eight to twelve seconds and roughly twenty-five words. Its only job is recognition — the viewer thinking 'that is my Tuesday'. Skip it and everything that follows is heard as a sales pitch by someone who has not agreed there is a problem.
Beat two: the stakes, quantified once
Recognition without cost produces no urgency. Attach one number to the problem — hours lost, revenue leaked, error rate, time-to-onboard. One number, not three, and preferably one your own customer data supports. Ten seconds, twenty words. This is also the beat where most Indian SaaS explainers over-write, stacking statistics until the viewer stops counting. A single credible figure carries more weight than a cluster of impressive ones, because the viewer can hold it in their head until the end of the film.
Beat three: the mechanism, not the feature list
This is the longest beat, thirty to thirty-five seconds and about eighty words, and it answers one question: how does this actually work? Show the flow — input, transformation, output — in a way a smart person outside your industry could repeat. Resist naming modules. Resist showing the settings screen. Animation is the right medium here precisely because it can abstract the interface into a diagram, and a diagram is what the viewer needs before any UI makes sense. If your mechanism cannot be drawn in three frames, the product story is not ready to be filmed.
Beat four: the proof, from outside your company
Twelve to fifteen seconds. A named customer, a measurable outcome, a certification, a scale number — anything sourced from outside your own marketing team. This beat converts interest into permission to believe. It works better as a specific and unimpressive-sounding fact than as a superlative: 'cut month-end close from nine days to two at a 400-person manufacturer' outperforms 'trusted by industry leaders'. Superlatives are free, which is exactly why nobody believes them.
Beat five: one action, stated plainly
Eight seconds, fifteen words, one instruction. Book a demo, start the trial, download the spec sheet — never all three. The end card holds the action on screen for at least four seconds and repeats it in the voiceover, because a meaningful share of viewers are listening while doing something else. This is also the only place a URL belongs. Everything before it earns the right to be here, and stacking three options at the end reliably converts worse than one.
Pacing, voice and the silent viewer
Write to 140 words per minute, not 190. Read the script aloud with a stopwatch before it is approved — scripts that look tight on paper routinely run eighteen seconds long in the booth. Design for mute: every key line needs an on-screen equivalent, because a large share of social and site views start silent. Choose a voice that matches your buyer's seniority rather than the most 'energetic' option, and if you serve Maharashtra directly, budget Marathi and Hindi versions from the same animation build — the artwork is reused, only the timing sheet changes.
What you should be able to do now
write or judge a 90-second explainer script against a five-beat structure and a real word budget