Character Design for 2D Explainer Video

How to design a character that can carry a series, what the animation budget actually depends on, and where characters help learning.

A character in an explainer video is doing a specific job: giving the viewer someone to follow through material that is otherwise abstract. Designed well it makes a process comprehensible and a series coherent. Designed carelessly it becomes a generic figure gesturing at diagrams, which adds production cost and nothing else.

The first design decision is how much the character has to do. A figure that only needs to stand, point and react can be built simply and animated cheaply. One that must walk, handle objects, turn and express a range of emotion requires a fuller construction and a rig to match, and the cost difference across a series is substantial. Deciding the required range before designing prevents building either too little or too much.

Silhouette determines recognisability and should be tested at small size before any detail is added. A character identifiable as a black shape will work as a thumbnail, at the edge of a busy frame, and on a phone. One that depends on facial detail or colour for identification will not, and most of a character's screen appearances are small.

Style should follow the brand rather than the medium's conventions. The default explainer character, rounded, flat, cheerful, minimal features, is so widely used that it now signals template rather than design. A character drawn in the brand's own visual language, using its palette and line quality, distinguishes the content immediately. Jonauskaite et al. (2020) documented consistent patterns of emotion associations with colours, which is a reason to derive the character's palette from the brand rather than defaulting to the friendly pastels the format has accumulated.

The research on animated figures in instructional contexts is more supportive than the format's reputation suggests. Li et al. (2022) found that animated pedagogical agents enhanced learning outcomes during learning, and Liew et al. (2022) examined how an agent's emotional presentation influenced mental effort and learning performance in a multimedia learning environment. Both indicate that a designed presence is functional rather than decorative when it is given a role.

That role should be explicit. The character can be the learner, discovering the process alongside the viewer. The guide, explaining it. The customer, experiencing the problem. Or the operator, demonstrating. Each produces a different film, and a character with no defined role becomes a figure standing beside information, which is the version that adds nothing.

Cognitive load is the constraint that decides how much the character should be doing at any moment. Beege and Ploetzner (2025), studying learning from interactive video, examined how design and cognitive load influence what viewers take from video material, and Ludwig et al. (2026) found that instructional design and cognitive load affect knowledge acquisition. A character animating energetically while a diagram builds competes with the diagram, and the diagram is usually the point.

Reuse economics are what justify the design cost, and they only work at volume. A character costs meaningfully more for its first appearance because design and rigging have to be built, and meaningfully less for its tenth. A company producing a series benefits; a company producing one explainer should use a simpler treatment and spend the difference on the explanation.

Representation deserves deliberate thought in a multi ethnic market. A character that reads as belonging to one community may narrow the content's reach, while one designed to be entirely unplaceable can feel evasive. The workable position for most Malaysian corporate content is a small cast rather than a single figure, which allows the material to reflect the audience without any individual character carrying that burden.

The archive requirement is the part that gets forgotten and determines whether the character survives a change of supplier. The design files, the rig, the colour specification, the expression set and the usage rules should be delivered and archived as brand assets alongside the logo. A character whose production files are inaccessible has to be rebuilt, at which point the recognition the brand accumulated was attached to something it did not control.

References

Jonauskaite, D., Parraga, C. A., Quiblier, M., & Mohr, C. (2020). Feeling blue or seeing red? Similar patterns of emotion associations with colour patches and colour terms. i-Perception, 11(1), Article 2041669520902484. https://doi.org/10.1177/2041669520902484

Li, W., Wang, F., Mayer, R. E., & Liu, T. (2022). Animated pedagogical agents enhance learning outcomes and brain activity during learning. Journal of Computer Assisted Learning, 38(3), 621–637. https://doi.org/10.1111/jcal.12634

Liew, T. W., Tan, S. M., & Kew, S. N. (2022). Can an angry pedagogical agent enhance mental effort and learning performance in a multimedia learning environment? Information and Learning Sciences, 123(9/10), 555–576. https://doi.org/10.1108/ILS-09-2021-0079

Beege, M., & Ploetzner, R. (2025). Learning from interactive video: The influence of self-explanations, navigation, and cognitive load. Instructional Science, 53(1), 99–119. https://doi.org/10.1007/s11251-024-09693-5

Ludwig, S., Rausch, A., & Taub, M. (2026). Effects of instructional design, instructional preferences, and cognitive load on problem solving and knowledge acquisition in a computer-based office simulation. Learning and Instruction, 101, Article 102255. https://doi.org/10.1016/j.learninstruc.2025.102255