Brainstorm Broadcast Graphics: A Rundown-First Workflow

Published November 4, 2026 · Updated November 4, 2026 · 8 min read · by the Sakura5 team

To brainstorm broadcast graphics, start from the rundown rather than from a style frame: list every segment, write down what the viewer needs to know at each moment, map each of those needs to a graphic type — lower third, full-screen, over-the-shoulder box, ticker, bug — and only then explore looks. That order keeps the package useful on air instead of beautiful in a pitch deck and unused by the third show.

A quick note on the phrase itself. Brainstorm is also the name of a Spanish company that makes real-time graphics and virtual set software, so if you came here for a product manual, their own documentation is the place to go. This guide is about the other meaning: how a producer, designer or small team generates and organizes ideas for an on-air graphics package before anyone opens a design tool.

Below: why most graphics brainstorms go wrong, a six-step workflow, a full worked example for a local election night, the edge cases that catch teams out, and a short FAQ.

Why graphics brainstorms usually go wrong

The typical session opens with mood boards. Someone shares a reel from a big network, someone else likes a particular gradient, and within twenty minutes the room is debating typefaces. It feels productive. It isn’t, because nobody has yet agreed what the graphics are for.

Broadcast graphics are information design under time pressure. A lower third has a few seconds to tell a viewer who is speaking and why they matter. A results full-screen has to be read by someone half-watching from the kitchen. When the brainstorm starts with style, three things happen:

• Coverage gaps. The package has a gorgeous name strap and nothing for the moment a reporter needs to show a phone number, a correction or a breaking update.
• Template sprawl. Each idea becomes its own template, and the operator ends up with forty variations of the same box, none of which fits the long name that turns up at 6:58 p.m.
• Unbuildable ideas. Concepts that need data the newsroom doesn’t have, or animation the playout system can’t trigger live.

The fix is to separate three questions and answer them in order: what must the viewer know, which graphic type carries it, and only then, what does it look like.

How to brainstorm broadcast graphics: a six-step workflow

Set aside about ninety minutes for a single show, longer for a full channel package. It works alone or with a producer, a designer and whoever operates graphics on the night — that last person is the one most often left out and most often right.

1. Put the rundown at the center

Write the show name in the middle of a page and every segment around it as a branch: open, headlines, each story block, weather, sport, interview, close. If the show doesn’t have a stable rundown yet, use the last three episodes and merge them. Everything else hangs off these branches, so getting them right matters more than any idea you’ll have later.

2. List the viewer’s questions for each segment

Under each segment, write what a viewer would ask at that moment. “Who is this?” “Where is this happening?” “What’s the number?” “Is this live?” “What happens next?” Keep them as questions, not designs. This is the generating phase: no judging, no cutting.

3. Map each question to a graphic type

Now attach a carrier to every question. Identity usually means a lower third. Location means a locator strap or map. A single striking number might be a full-screen; a running total might live in a corner bug. Some questions don’t need a graphic at all — the anchor can just say it. Mark those explicitly so nobody designs for them.

4. Group, merge and count templates

Look across branches for repeats. You’ll find that “who is this” appears in six segments and “what’s the number” in four. Each repeated need becomes one template with variants, not several separate ideas. If your list has more than a dozen distinct templates for one show, it’s probably too many for an operator to run cleanly. Our guide to grouping ideas with an affinity diagram covers this clustering step in more detail.

5. Stress-test with real worst cases

For every template, write the ugliest real content it will ever carry: the longest guest name, the two-line job title, the number with a decimal and a unit, the location with an apostrophe. Check it against the safe areas your broadcaster uses — the EBU recommendation on safe areas for 16:9 production is the common reference if you don’t have a house spec.

6. Only now, explore the look

With a fixed list of templates and their worst-case content, style exploration becomes focused. Designers can try three directions against the same real content instead of against placeholder text that always fits.

Turning the map into a graphics list

The structure above is a tree: show, then segments, then viewer questions, then graphic types. A mind map is the natural shape for it, and it beats a spreadsheet at the generating stage because you can see gaps at a glance — a segment branch with no children is a segment nobody has thought about.

If you want a fast first draft, Mindmap AI can take a typed topic or a pasted rundown and build an organized starting map on an infinite canvas (an AI feature in the Pro tier). Tap any segment node to expand it with more sub-ideas, then delete what doesn’t fit your show. It won’t know your newsroom’s data sources or your playout system, so treat the generated branches as prompts to react to, not decisions. For more on that react-and-prune approach, see our piece on AI brainstorming with mind maps.

When the map settles, switch it to an outline view. That outline is effectively your graphics spec: segment, need, template, notes. Exporting it as a PDF gives the designer and the operator the same document to work from.

A worked example: local election night

Take a fictional regional station preparing graphics for a city council election night. The show runs two hours, there are six wards, each ward has between two and four candidates, and the team is three people: a producer, one designer and one graphics operator.

The rundown branches

The producer lays out seven segments: cold open, headline results, ward-by-ward block, reporter live shots from two count centers, analyst interview, turnout update and close. Those seven become the first ring of the map.

The questions

Under the ward-by-ward block alone, the team writes eight viewer questions in about ten minutes: who’s leading, by how much, how many ballots are counted, is this final, who held the seat before, what did they win by last time, is this a change, and where is this ward. Across all seven segments they end up with twenty-nine questions.

The merge

Mapping those twenty-nine questions to carriers and merging repeats leaves nine templates: candidate lower third, reporter lower third, analyst lower third, ward result full-screen, ward result over-the-shoulder box, counted-percentage bug, “projected” and “final” flags as a single template with two states, a ward locator map and a closing summary full-screen. Four questions get marked “anchor says it” and receive no graphic at all.

The stress test

The operator pulls the real candidate list. The longest name is twenty-seven characters with a hyphen, and one candidate’s party label is three words long. The first draft of the lower third breaks on both. The designer widens the strap and moves the party label to a second line before a single style frame is drawn, which saves a redesign the week of the election.

The finished map, exported as an outline, becomes a one-page brief: nine templates, the data each one needs, who triggers it and in which segment. The whole session took a little under two hours.

Edge cases worth planning for

Recurring shows versus one-off events. A daily show needs a small, durable set of templates that survive hundreds of episodes. A one-off event can afford a few bespoke pieces. Brainstorm the durable set first, then branch off the event-only ideas so they don’t leak into the permanent package.

Breaking news. Add a branch for the moment the rundown is thrown out. What does the operator need with no notice? Usually a generic headline strap, a “developing” flag and a correction template. If those don’t exist, someone will improvise one live.

Vertical and social cut-downs. If clips will be re-edited for phones, note on each template whether it must survive a 9:16 crop. A corner bug that works on air may vanish entirely in a vertical edit.

Accessibility. Contrast, minimum on-screen time and text size belong in the brainstorm, not in a late review. Put a single “readability” node on the map and check every template against it before you pick a style. A well-structured plan from your map makes that review a checklist instead of a debate.

FAQ

What are the main types of broadcast graphics?

The core set is lower thirds for names and titles, full-screens for data or quotes, over-the-shoulder boxes beside the anchor, corner bugs for persistent information like a live flag or score, tickers or crawls for running text, and locator straps or maps. Most shows need only a handful of these, each with a few variants.

How many templates should a graphics package have?

There’s no fixed number, but fewer than you think. Start from the viewer questions in your rundown, merge repeats, and you’ll usually land on a short list. If one show needs more than about a dozen distinct templates, check whether several are the same need with different styling and combine them.

Should the designer or the producer lead the brainstorm?

The producer should own the first half, because they know the rundown and what viewers need at each moment. The designer leads once the template list is fixed. Bring the graphics operator in from the start; they know which ideas can actually be triggered cleanly on air and which will fail under live pressure.

Can I use a mind map app to plan broadcast graphics?

Yes. The show-to-segment-to-question-to-template structure is a tree, which is exactly what a mind map shows well. An app such as Mindmap AI can draft the first branches from a rundown, then let you expand, prune and export the result as an outline that works as a graphics brief.

← All articles