5 min read
The real cost of a hand-made product demo
A five-minute demo video can take days to make, and the bill comes due again every time the product changes. Here is where the hours actually go.
By The Primordio team

Ask a team how long their last product demo took to make and you will usually hear a number like “an afternoon.” Ask the person who actually made it and the number is closer to a week.
A short demo video looks simple on screen. Someone clicks through a workflow and explains it. But the five minutes a viewer sees sit on top of a long chain of work, most of which never shows up in a plan or a budget.
Where the hours go
Planning and scripting
Before anyone presses record, someone has to decide what the demo covers, in what order, and what will be said at each step. A demo that skips this stage tends to wander, so the script usually goes through a draft or two and a review with whoever owns the feature.
Preparing the environment
A demo needs an account that looks real but isn’t, with believable data, no customer information, and nothing half-finished on screen. Building that environment, and resetting it between takes, is often the single biggest time sink. Every take that creates a record or changes a setting leaves the account in a different state from the one the script assumes.
Recording and retakes
Very few demos are recorded in one take. A typo in a form, a slow page load, a notification sliding in from the corner, or a stumble in the narration all mean starting the segment again. Recording also tends to happen in short bursts between other work, which adds setup time to every session.
Voiceover
Narration is either recorded live, which multiplies retakes, or recorded separately and synced afterward, which adds its own editing pass. Either way, someone needs a quiet room, a decent microphone and the patience to say the same sentence several times.
Editing
Editing is where an afternoon turns into a week. Cutting dead air, fixing the cursor wandering off target, zooming into the part of the screen that matters, adding captions, lining up narration with the action and exporting in the right formats all take time, and they take more time for someone who doesn’t edit video every day.
Review and publishing
The finished video goes back to the feature owner for a check, which usually produces a list of small fixes. Then it has to be uploaded, titled, described and linked from wherever customers will look for it.

The cost nobody budgets: doing it again
All of that is the cost of the first version. The more expensive part comes later.
Products change constantly. A button is renamed, a menu moves, a screen is redesigned. Each change can make a demo wrong, and a demo can’t be patched the way a paragraph of text can. Fixing a video usually means recording the affected section again, and matching the old voice, pacing and visual style closely enough that the seam doesn’t show. Often it is easier to remake the whole thing.
So the real cost of a demo is not one production. It is one production plus a partial or full remake every time the workflow it shows changes. For a fast-moving product, that can mean several remakes a year for every demo in the library.
The costs that don’t show up on a timesheet
Skilled people doing repetitive work. Demos are usually made by product managers, solutions engineers or support leads: people who understand the product well. Every hour they spend on retakes and editing is an hour not spent on the work only they can do.
Inconsistency. When demos are made by different people at different times, they look and sound different. Some have captions and some don’t. Some explain why, others only what. Customers notice, even if they can’t say exactly what feels off.
The demos that never get made. Because each demo is expensive, teams ration them. The flagship workflow gets a video. The setting that generates a steady trickle of support questions doesn’t, because there is never time.
The demos that quietly go stale. When remaking a demo means another week of work, the remake gets postponed. The old video stays up, slowly drifting further from the product, as described in why product docs go stale.
How to estimate your own number
You can get a rough figure for your team in a few minutes:
- Hours per demo. Add up planning, environment setup, recording, voiceover, editing and review for a typical demo. Be honest about retakes.
- Number of demos. Count the demos you have, plus the ones you would make if they were cheaper.
- Updates per year. Estimate how often each demo’s workflow changes enough to need a remake. For an active product, two to four times a year is common.
- Multiply. Hours per demo, times demos, times one plus the number of updates, gives you hours per year.
Multiply that by the hourly cost of the people doing the work and the number is usually larger than anyone expected. That figure is worth having when you decide which demos to make, which to retire, and how to produce the rest.
Where Primordio fits
Primordio was built to remove most of this chain. You describe the workflow in plain English, review the plan, and get back a narrated video with captions and a written guide from the same run, with no recording sessions or editing. When the product changes, the tutorial can be refreshed against the current interface instead of remade by hand.
A quick checklist
- You know roughly how many hours a typical demo takes, retakes included.
- You have a reusable demo environment with realistic, fictional data.
- Every demo has a named owner and a date it was last checked.
- You track which workflows change most often, since those demos cost the most to keep.
- You have a short list of demos you would make if production were cheaper.
- You budget for updates, not just for first versions.
A demo is one of the most useful things a product team can publish. It is also one of the most expensive to keep true. Knowing the real cost is the first step to deciding how to pay it.