How to Make Animated Map Videos (No Face, No Filming)
How to make an animated map video with no face and no camera: the three map styles, the one-question script, and the label and number rules that keep a map readable.
An animated map video explains a place by showing it. Borders fill in, a route draws itself across an ocean, a counter rolls up to 105 million while a voice explains why almost everyone in Egypt lives on about 4 percent of the land. There is no face on camera and nothing to film. That is why the format keeps showing up on faceless channels: the picture is data, and the whole performance is a voice.
You recognize the family by its titles. "Why 105 Million Egyptians Live on 4% of Their Land". "Why Norway and Sweden Split in 1905". "Why Iceland Has Almost No Trees". This is a breakdown of that format: what an animated map video is, the three ways to make one, the script shape that holds one together, the label and number rules that decide whether the map reads clearly or confuses people, and what a render involves. The worked example throughout is a demo we made with Unfaced, embedded under the script section.
What an animated map video is#
An animated map video is a narrated video in which a map does the explaining. Where people live, how a border was drawn, how far a ship has to sail, why one country controls an ocean zone: the map shows it while the narration says it. The map fills, empties, rewinds and counts. Names land on the places they belong to. When the story moves to 1905, a year marker rolls to 1905 and the map turns to paper.
Two things separate it from a slideshow with a voiceover. First, the map is continuous. One map stays on screen and changes while the story stays in one part of the world, so the viewer keeps their bearings instead of re-reading a new picture every ten seconds. Second, the figures in the narration become objects on the map. "105 million people" is a counter over the country. "Nine in 10" is a share panel. "Eight times as much" is a bar rising eight times. Each object lands on the word that names it.
For a faceless channel the format removes the usual production problems. Nobody has to be on camera, no room has to look good, and no footage has to be shot. The assets are map data, a voice and the script. Everything else is layout and timing.
The three map styles (story, explainer, documentary)#
Three styles cover the format, and picking one is the first decision because it decides what the viewer sees and what you have to supply.
| Style | What the viewer sees | Best for |
|---|---|---|
| Map explainer | One continuous map, no faces. Every figure becomes an object on it: coins, shares, calendar pages, a country rising. Relief for today, sepia for the past. | Geography, population and economics questions. |
| Map story | A bright map where countries have faces and their own lines, with flags, a year ticker and small drawn figures. | History told fast: purchases, borders, splits, wars. |
| Map documentary | Real footage of the place cut with an explainer map. The map says where and how much; the footage shows what it looks like. | Place documentaries that need the real view and the geography in one video. |
The map explainer is the quietest of the three and the easiest to read. One narrator, no characters, and a rule that keeps the picture honest: if the narration does not say a number, the map does not draw one. You choose a look (relief, atlas or night for the present; the past turns sepia), how many numbers appear, and whether ships and flags show.
The map story lets the map talk. Countries speak their own lines, a year ticker rolls through the dates, and small figures stand where the story happens. It fits history because history has sides, and sides are easy to color. You pick one of six map looks and decide whether countries or flags get faces.
The map documentary is the one style that brings outside pictures: real footage of the place, its streets, its port, its landscape, cut back to the map whenever the narration says where or how much. It is also the style where licensing matters, because footage is an asset you either own, license or find under terms that allow the use. The map itself needs the same check, and the last section covers where map data usually comes from.
All three are open styles in Unfaced, and the demos on the animated map videos page show each one on the same format.
Script structure: one question, one map#
The question is the title, stated as an odd fact the viewer wants resolved: why 105 million people live on 4 percent of the land, why a country's shape looks the way it does, why a border runs straight. One map carries the whole answer.
Here is the worked example, "Why 105 Million Egyptians Live on 4% of Their Land", made with Unfaced in the map explainer style:
Watch how it is built while you read the structure under it. The opening shows the paradox and its question: the people, the land, the gap between them. The middle takes the factors one at a time, and each factor is one beat that adds something the viewer has not seen: a region filling, a rain figure, a route, a share. The counters say what they count (105 MILLION PEOPLE, 4% OF THE LAND, 10 mm OF RAINFALL A YEAR), one number at a time. Egypt stays on every map in the video, so the viewer never has to find the subject again. The close comes back to the opening thing and weighs the factors.
The transferable rules are short:
- One question per video. A second question is a second video.
- One map while the story stays in one part of the world. Start a new map only when the story moves far away.
- Every beat adds at least one thing the viewer has not seen. Two beats that would look the same are one beat's job.
- Say the year when the story moves in time, in digits.
- Every number carries its unit and says what it counts, and no number appears that the narration does not say.
This is the same skeleton as the 15 beats of a high-retention faceless script, with the map taking the role the picture usually takes. If a blank page is the problem, the script outline generator builds the beat list with timings for your topic, and the script timer checks that the script fits the length you want before you render anything. The opening line is worth its own pass against how to write a hook, because the map has to pay off whatever the opening line promises.
Labels, numbers and the 3 map mistakes#
Labels and numbers are most of the on-screen text in an animated map video, and both have one rule that covers most of it. A label goes on the thing it names. A number says what it counts.
Labels: name the places the narration names, put the name on the place, keep one style of caps across the video, and let names fade after a few seconds so the map can be read again. A label that covers the subject of the beat counts as a defect.
Numbers: print the figure with its unit, exactly as the voice says it (45% of the land, 10 mm a year, $307 million), and one number at a time. If the narration compares two figures, draw them as two objects of different size rather than two labels. A figure the narration never says does not go on screen.
Everything else that goes wrong with map videos tends to land in three mistakes. We check our own demos frame by frame before they go anywhere, and these are the three the check keeps finding:
- A label in the wrong place. A city name inland when the city is on the coast. A country name over the wrong country. A legend naming one country while the shape below it is another. Viewers read labels before they listen, so one misplaced name outlives everything the narration says after it.
- A number that contradicts the narration. The voice says one figure and the counter shows another, or a bare number sits over a region with nothing saying what it counts. The fix is boring and effective: every figure in the script gets its unit and its noun before anyone draws it.
- Decoration with nothing behind it. A legend color with no matching shape on the map. A route drawn across land. A region filled in a color the viewer never sees explained. If a visual element is not carrying information the narration needs, it is taking attention from the ones that are.
The check that catches all three is cheap: go through the render one frame per second and compare labels and legends against the map, screen text against the narration's figures, and any real pictures against what the voice is saying at that moment. It is the step that separates a demo that reads from one that gets paused and mocked.
From script to render#
A render is the last step, and it is the shortest. The work is the script and the plan.
In a hand-built workflow you write the script, plan each beat's map layer (a filled shape, a route, a pin, a counter, a year), draw the map in an editor, record a voice and cut it together. The asset question is mostly about map data and, for the documentary style, footage. Natural Earth's borders and country shapes are public domain, with no permission needed and no credit required (terms). NASA's Blue Marble imagery is NASA material and carries no copyright. OpenStreetMap is different: its data is open under the ODbL, and using it in a video requires a credit that names OpenStreetMap and its license (copyright page). Stock footage brings its own license, so check what a clip's terms allow before it becomes part of a video you monetize.
In Unfaced the same pipeline runs from one sentence: pick a map style, put in the question or the topic, and the studio writes the script, plans the map layers against the narration's words and renders the video. You can watch the demos on the animated map videos page, and you can make one in the studio. Nothing is published until you approve it.
Whatever route you take, run the checks before the render rather than after it. The retention risk checker reads the script for the structural problems that lose people, and once the video is live, how to read a retention graph tells you which beat actually lost them.