Designing Annotations in Visualization: Considerations from Visualization Practitioners and Educators

Abstract: Annotation is a central mechanism in visualization design that enables people to communicate key insights. Prior research has provided essential accounts of the visual forms annotations take, but less attention has been paid to the decisions behind them. This paper examines how annotations are designed in practice and how educators reflect on those practices. We conducted a two-phase qualitative study: interviews with ten practitioners from diverse backgrounds revealed the heuristics they draw on when creating annotations, and interviews with seven visualization educators offered complementary perspectives situated within broader concerns of clarity, guidance, and viewer agency. These studies provide a systematic account of annotation design knowledge in professional settings, highlighting the considerations, trade-offs, and contextual judgments that shape the use of annotations. By making this tacit expertise explicit, our work complements prior form-focused studies, strengthens understanding of annotation as a design activity, and points to opportunities for improved tool and guideline support.

• The paper presents a qualitative study that elicits annotation design heuristics used by both practitioners and educators.
• It identifies six interdependent considerations, including audience tailoring, hierarchy, and density tuning for effective visualizations.
• The study highlights challenges in tool integration and pedagogical gaps, offering a framework for improved annotation authoring.

Annotation Design: Knowledge from Visualization Practice and Pedagogy

Introduction

This essay summarizes "Designing Annotations in Visualization: Considerations from Visualization Practitioners and Educators" (2604.07691), which conducts an in-depth qualitative analysis of annotation design practices in professional visualization contexts. The study systematically elicits annotation decision heuristics from both practitioners and educators, encapsulating their operational logic, design tensions, and boundary conditions. This discourse is predominantly situated in the context of two-dimensional visualization artifacts and reflections on annotation authoring, with a rigorous focus on tacit knowledge and empirical rationale driving annotation usage and configuration.

Empirical Methodology and Scope

The paper utilizes a two-phase, semi-structured interview protocol to elicit situated annotation design considerations:

• Phase 1: Ten practitioners with varied visualization experience articulate annotation design choices, referencing self-produced annotated artifacts.
• Phase 2: Seven educators with significant visualization teaching and research profiles critique and extend practitioner-identified considerations, reflecting on pedagogical framing and risks.

Data were thematically coded using reflexive thematic analysis. The participants span roles such as data journalists, analysts, consultants, and university educators, covering dashboards, narrative infographics, and scientific visualizations. The analysis foregrounds situated, production-oriented reasoning over prescriptive guidelines.

Six Annotation Design Considerations

The study distills annotation practice into six recurring, interdependent design considerations:

1. Audience Tailoring: Annotations are audience-dependent, with textual density, content, and placement modulated by anticipated reader knowledge, context loss risks (e.g., chart detachment or screenshots), and potential misinterpretation costs. Defensive annotation—embedding caveats or context directly on the chart—is adopted where redistribution risk is high, but educators highlight sustainability and maintenance trade-offs.
2. Annotation Hierarchy: Practitioners manage attentional focus by hierarchically encoding annotation primacy. Primary messages receive maximal salience; secondary commentary is visually subordinated or eliminated to reduce noise. This aligns visually with empirically established principles favoring restricted overlays and selective emphasis.
3. Proximity and Placement: The default stratagem is to attach annotation text directly to the target mark, falling back sequentially to leader lines and only to detached legends or keyed lists if adjacency induces clutter or in mobile/interactive scenarios. Legends are minimized due to their cognitive burden unless spatial or modular distribution warrants otherwise.
4. Association by Style and Color: Label-target association employs color and styling only when categorical discriminability is assured. Redundant visual cues such as proximity, connectors, or enclosures are routinely paired with color coding, in response to known perceptual and accessibility limitations.
5. Layer Identity: Stylization choices signal whether annotations are integrated as part of the base encoding (using uniform typographic and color styles) or visually distinct as author commentary. This is a pragmatic rhetorical decision, structured by communicative goals, maintenance cost, and anticipated chart reuse or reinterpretation.
6. Density Tuning: The number of visible annotations is bounded by scanability and available space, with "annotation budgets" flexed according to the communicative core of the artifact and technical affordances of the medium. Numeric rules are treated as local heuristics; practitioners often employ additive or subtractive workflows to reach acceptable density.

Educators systematize these considerations, emphasizing that all are context-contingent and resist codification as universal rules.

Contextual and Methodological Extensions

The paper extends the analysis in several important dimensions:

• Medium Dependency: Annotations in static artifacts are maximally explicit due to absence of interactive cues. In contrast, tooltips and staged/hybrid annotation in interactive contexts offload detail, but critical claims are still persistently surfaced to avoid reliance on interaction.
• Informal Evaluation: Iterative critique with peers and non-experts governs annotation adequacy. The principle metric is whether unaided readers infer the intended message from first glance, with removal or restyling of annotations that induce clutter or mislead.
• Authoring Toolchains: Practitioner workflows are multi-tool, commonly exporting charts from analytic tools to graphic editors (e.g., PowerPoint, Illustrator, Figma) for final annotation pass, driven by inadequate default control in BI or dashboard platforms. This engenders maintenance debt and data-linkage fragility, reinforcing the need for annotation-aware authoring systems and grammars.
• Pedagogical Integration: Annotation design is largely implicit in undergraduate and graduate curricula, receiving little dedicated instructional time. Educators see the study's considerations as candidate vocabulary for critique-centered pedagogic interventions and for scaffolding reflective student practice.

Implications and Boundary Conditions

The study provides a critical bridge between empirical taxonomies of annotation forms and the situated knowledge required for effective annotation production:

• Complex Interdependencies: Considerations interact non-trivially; audience tailoring constrains density and explanation depth, while placement and styling jointly determine both reading order and maintainability.
• Limits of Prescriptive Guidance: Both practitioners and educators emphasize that rigid numeric or sequential heuristics are rarely functional outside local design or organizational contexts.
• Tool and Systems Implications: The findings motivate development of visualization grammars and authoring systems that treat annotations as first-class, data-linked, and context-aware objects, supporting stable hierarchy, association, and reuse across static and interactive contexts.
• Research Directions: The six identified considerations provide a structured substrate for empirical annotation effect studies (comprehension, persuasion, trust), formal annotation grammar development, and design critique/teaching frameworks.

Conclusion

Annotation design in visualization is a multi-dimensional, context-dependent activity characterized by tightly coupled practical judgments. By systematically elucidating these decision logics and mediating practitioner-educator dialogue, the study (2604.07691) provides a transferable framework for understanding and supporting annotation design. These considerations offer not only design reflection scaffolds but also concrete targets for tool developers, educators, and researchers seeking to operationalize, automate, or empirically validate annotation strategies within the broader visualization pipeline.