Preparing Demo Data: Build a Demo Environment That Records Well

Most screen recordings fail before the record button is ever pressed. Not because of audio, lighting, or pacing, but because of what is on the screen: an empty dashboard, a customer named “asdf test 3”, a chart with a single data point, or a notification from a real colleague sliding in at the worst possible moment.

Demo data is set design. Viewers may never consciously notice good demo data, but they always notice bad demo data, and it quietly costs you credibility. This guide covers how to prepare an environment that looks real, records cleanly, and survives multiple takes.

Start With a Dedicated Demo Account

Never record from your working account. Your daily account is full of half-finished experiments, personal information, and states you cannot reproduce if a take goes wrong.

Create a separate account — or workspace, tenant, or profile — that exists only for recording:

• Neutral identity: a demo user with a plausible name and a generic avatar, not your own face and email.
• Clean permissions: whatever a typical new customer would see, so you do not demo admin-only features by accident.
• Reproducible: you should be able to return it to a known state in minutes.

If your product supports seeding or fixtures, write a script that creates this account from scratch. The five minutes you spend scripting it pays off the third time you need to re-record.

Name Your Sample Data Deliberately

Placeholder names are the single most common tell in a bad demo. “Test User 1”, “aaa”, and “Lorem ipsum project” all signal that nobody cared enough to prepare.

Good sample data follows a few rules:

• Plausible but clearly fictional. “Northwind Logistics” and “Harbor Coffee Co.” read as real companies without implying a real customer.
• Varied in shape. Mix short and long names, one-word and multi-word entries. Uniform data hides layout problems and looks synthetic.
• Free of real customer information. Never record a live account with actual client names, even blurred. If you need realism, base names on fiction, not on your CRM.
• Culturally neutral or intentionally diverse. If your audience is global, a list of names from one country looks narrow.

Check the strings that appear in truncated or overflowing form too — a sidebar that cuts off “Q3 Regional Perf…” on screen reads as unfinished.

Get the Data Density Right

An empty state is not a demo, and neither is a wall of 400 rows. Aim for the density a healthy, active account would show:

• Lists: enough rows to fill the visible area and suggest more below, without needing to scroll to find the interesting one.
• Charts: a trend with a story — growth, a dip, a recovery. Flat lines and single bars make the feature look useless.
• Dates: recent and relative. Data timestamped eight months ago instantly dates the video. If your demo environment supports it, shift all timestamps forward before recording.
• Numbers: realistic magnitudes. $12.00 in revenue looks like a test; $128,400 looks like a business.

Put the record you plan to click on somewhere predictable — near the top, but not always the first row, which looks staged.

Clean the Surrounding Environment

The demo account is only half the frame. Before you record:

• Close unrelated tabs and hide bookmarks bars containing personal links.
• Turn on Do Not Disturb so messages and calendar alerts stay out of the capture.
• Use a neutral desktop wallpaper and clear the desktop of files.
• Set the browser or app zoom so text is readable at the final export size, not just on your high-resolution display.
• Sign out of any personal accounts that might surface an avatar or autofill suggestion.

In Recorded, capturing a single window instead of the full screen removes an entire category of these problems at once. Window capture keeps the dock, menu bar, and second monitor out of the frame automatically.

Make State Resettable

Demos are destructive. You delete the item, you submit the form, you use up the invite code — and then your second take has nowhere to go.

Before recording, decide how you will get back to the starting state:

• Snapshot it. A database dump, a duplicated workspace, or a saved fixture you can restore in one command.
• Script it. A short seed script is more reliable than manually recreating fifteen records at midnight.
• Or record in the reverse order of destruction. Capture the “list with items” shot first, then the delete flow.

Knowing you can reset removes the tension from recording. You stop protecting the data and start focusing on the narration.

Plan Data Around the Story

The best demo data is designed backwards from the script. Walk through your outline and ask what has to already exist at each beat:

1. The opening shot needs an account that looks lived-in.
2. The middle needs a specific record to demonstrate the core feature — pre-create it, with a name you can say out loud.
3. The payoff needs a visible result: a completed status, a finished export, a chart that moved.

If a step requires waiting — a long upload, a background job, an email arriving — prepare the completed version in advance and either cut to it or speed up the wait in the editor.

A Quick Pre-Recording Data Checklist

• Recording from a demo account, not a personal one
• No real customer or employee names anywhere on screen
• Sample names are plausible, varied, and fictional
• Lists and charts have realistic density and a visible trend
• Dates and timestamps look recent
• Notifications silenced, unrelated tabs closed
• The record you will click on is pre-created and easy to find
• A reset path exists for take two

Spending fifteen minutes on this before you record saves an hour of retakes and makes the difference between a video that looks like a product and one that looks like a prototype.

Once the environment is ready, the rest is straightforward: capture the window, walk through the story, and let the editor handle the polish.