Communication Bad Smells Catalog · Spell · DCC/UFLA
PT ← Spell
  1. Home
  2. Methodology
About the research

From code communication to people communication

This catalog is the output of a case study extending the concept of bad smells — coined by Kent Beck and Martin Fowler for code quality — to the socio-technical domain of communication in multidisciplinary remote agile teams.

Theoretical grounding

Grounding in the literature

Applying the bad smell concept to communication builds on an established body of literature about noise, clarity and communication effectiveness in distributed teams.

Shannon & Weaver (1949)

Ground the notion of noise in communication and loss of clarity.

 Fowler (2018)

Coined the concept of a bad smell in code — the metaphor this study extends to communication.

 Zahedi & Babar (2014)

Sharing technical knowledge through artefacts.

 Diel et al. (2016)

Communication strategies in distributed DevOps teams.

 Weger et al. (2022)

Communication effectiveness and reflexivity in teams.

 Kostin & Strode (2022)

Relate clarity to effectiveness in distributed Scrum teams.

 Rauf et al. (2023)

Structured communication in requirements elicitation on remote projects.

 Neeley (2021)

The choice of communication channels as a decisive factor for remote project outcomes.

The case study

The research was conducted in a real remote project developing a track-based training platform. The project ran for five years and involved 50 participants in total, organized into subteams by function and area, across six stages of interaction:

  1. 1

    Client → Requirements

    The client passes the demands on to the Project Manager and the Requirements Team.

  2. 2

    Requirements → UX/UI

    Demands are elicited and documented as Epics and User Stories, prioritized, and turned into high-fidelity screen prototypes.

  3. 3

    Management → Development

    The Project Manager distributes tasks to the Development Team according to scope (coding, testing, DevOps, content production).

  4. 4

    Development → QA/Test

    The software increment is validated against the definition of done (unit, functional and usability testing).

  5. 5

    QA/Test → DevOps

    The DevOps Team integrates the code into the infrastructure.

  6. 6

    Management → Client

    The manager receives the sprint deliverable and presents it to the client.

Methodology

A mixed-methods questionnaire was applied to nine teams, following the concurrent triangulation design of Creswell & Creswell (2017): 17 open and closed questions, organized into four evaluative topics.

  • Topic 1

    Tools

  • Topic 2

    Communication Channels

  • Topic 3

    Challenges (perceptions)

  • Topic 4

    Possible solutions (perceptions)

The qualitative data were validated with a word cloud and a Sankey diagram (correlating reported challenges with proposed solutions), while the quantitative data measured the frequency (scale of 1 to 5) and quality (scale of 1 to 6) of communication at each contact point in the project.

Step by step

From questionnaire to insights

The quantitative and qualitative tracks ran in parallel until they converged in the integration of the findings.

Apply Questionnaire

A mixed questionnaire of 17 open and closed questions, applied to nine teams across different areas of the project, covering tools, channels, challenges and perceived solutions.

Quantitative track

  1. Data Cleaning

    Organizing the closed answers before analysis, preparing the frequency (1–5) and quality (1–6) scales for tabulation.

  2. Statistical Analysis

    Computing average frequency and quality of communication for each contact point — teams, leadership, management and coordination.

  3. Descriptive Statistics

    Consolidating the study's headline figures, such as the 100% adoption of WhatsApp and the percentage distribution of responding teams.

  4. Inferential Test

    Comparing contact points to locate where communication is weakest — showing a horizontal structure whose links outside the team (coordination, governance, users) are the most fragile.

Qualitative track

  1. Coding and Themes

    Answers to the two open-ended questions — challenges faced and solutions suggested — are organized into recurring themes cited by participants.

  2. Thematic Analysis

    Coded themes are grouped into consistent patterns — the basis from which the five bad smells of the catalog emerged.

  3. Word Cloud

    Term-frequency visualization over the open-ended answers — project, information, WhatsApp and Discord are among the most cited.

  4. Content Analysis

    Qualitative reading relating each reported symptom to the solution suggested by the participant themselves — the basis of the Sankey diagram below.

Data Triangulation

Integrating the quantitative and qualitative findings, following the concurrent triangulation design of Creswell & Creswell (2017) — the crossing that made it possible to name and categorize the five bad smells.

Insights and Proposals

The outcome of the process: the five communication bad smells identified and their mitigation strategies, documented in this catalog.

Research questions

  • RQ1

    What are the main communication tools used in the project and how do they impact the effectiveness of communication?

    Finding: WhatsApp is used by 100% of respondents, becoming the de facto channel even though it is not a formal project tool — one of the findings behind the recommendation to move important decisions to traceable channels (e.g. Discord, Confluence).

  • RQ2

    How do the frequency and quality of communication vary between different contact points in the project?

    Finding: Communication between teams is the most frequent and the best rated (4.2/5 frequency, 4.5/6 quality); Coordination, Governance and Users fall well below (averages of 2.97 and 3.33) — evidence of a horizontal structure whose links outside the team are the weakest.

  • RQ3

    What are the main communication challenges faced by the teams and what solutions can be proposed to overcome them?

    Finding: The twelve symptoms reported in the open-ended answers (see the correlation diagram below) were grouped into the five bad smells of the catalog, each with specific mitigation strategies drawn from the participants' own suggestions.

  • RQ4

    How can communication be improved to increase team integration and overall project efficiency?

    Finding: The most cited suggestions — centralizing communication, integrating the teams better and establishing consistent processes — recur across almost every symptom, suggesting that gains in integration depend less on a new tool and more on team agreements and present leadership.

What the data showed

Participating teams (% of responses)

  • Testing/QA30.4%
  • Development21.7%
  • Digital Transformation13%
  • DevOps8.7%
  • Management8.7%
  • Requirements4.3%
  • UX/UI4.3%
  • Coordination4.3%
  • Administrative Support4.3%

Tools

100%

of the professionals use WhatsApp to communicate in the project.

Most cited communication points

  • Communication between teams78.3%
  • Communication between leadership47%

A predominantly horizontal structure, with an emphasis on intra-team interaction.

Communication frequency × quality by channel

Frequency (scale of 1 to 5) Quality (scale of 1 to 6)

  • Teams

    Frequency 4.2
    Quality 4.5

  • Leadership

    Frequency 3.8
    Quality 4.3

  • Management

    Frequency 3.2
    Quality 3.7

  • Coordination / Governance / Users (average)

    Frequency 3.0
    Quality 3.3
Qualitative triangulation

From reported symptoms to suggested improvements

In the two open-ended questions of the questionnaire, each participant described the challenges they face and the actions they would suggest to solve them. The diagram below reconstructs those associations answer by answer — the same triangulation technique that made it possible to identify the five bad smells and ground the mitigation strategies of the catalog.

Reported symptom Suggested improvement
From reported symptoms to suggested improvements Sankey diagram linking the twelve communication symptoms reported by participants to the improvements they themselves suggested. The thickness of each link is proportional to the number of answers. Difficulty defining a common schedule → Adopting new tools: 8 answers Difficulty defining a common schedule → Following ceremonies/meetings: 4 answers Difficulty defining a common schedule → Centralizing communication: 4 answers Difficulty establishing more organic communication → Integrating the teams better: 2 answers Difficulty establishing more organic communication → Leadership relaying crucial information: 2 answers Lack of awareness of the multidisciplinary team → Following ceremonies/meetings: 3 answers Lack of awareness of the multidisciplinary team → Integrating the teams better: 3 answers Lack of awareness of the multidisciplinary team → Leadership relaying crucial information: 3 answers Decentralization in communication → Centralizing communication: 4 answers Decentralization in communication → Integrating the teams better: 4 answers Decentralization in communication → Adopting new tools: 4 answers Decentralization in communication → Well-established processes: 4 answers Difficulty monitoring the project → More follow-up: 4 answers Difficulty monitoring the project → Integrating the teams better: 4 answers Difficulty monitoring the project → Well-established processes: 4 answers Difficulty monitoring the project → Daily/weekly written report: 4 answers Difficulty obtaining detailed information → Leadership relaying crucial information: 10 answers Difficulty obtaining detailed information → Integrating the teams better: 5 answers Difficulty obtaining detailed information → Centralizing communication: 5 answers Difficulty obtaining detailed information → Recording information: 5 answers Difficulty communicating between hierarchical levels → More communication between managers: 8 answers Difficulty communicating between hierarchical levels → Annual sync meetings/newsletter: 4 answers Difficulty communicating between hierarchical levels → Integrating the teams better: 4 answers Disorganization of project information → Centralizing communication: 6 answers Disorganization of project information → Adopting new tools: 6 answers Disorganization of project information → Better documentation: 6 answers Disorganization of project information → Integrating the teams better: 6 answers Disorganization of project information → Well-established processes: 6 answers Disorganization of project information → Recording information: 6 answers Difficulty maintaining a consistent process → Following ceremonies/meetings: 4 answers Difficulty maintaining a consistent process → More follow-up: 4 answers Difficulty maintaining a consistent process → Daily/weekly written report: 4 answers Difficulty maintaining a consistent process → Well-established processes: 4 answers Lack of knowledge of standards of other projects → Annual sync meetings/newsletter: 3 answers Lack of knowledge of standards of other projects → Integrating the teams better: 3 answers Lack of knowledge of standards of other projects → More communication between managers: 3 answers Lack of knowledge of the flow of information → Centralizing communication: 5 answers Lack of knowledge of the flow of information → Integrating the teams better: 5 answers Lack of knowledge of the flow of information → Adopting new tools: 5 answers Lack of knowledge of the flow of information → Daily/weekly written report: 5 answers Lack of knowledge of the flow of information → Well-established processes: 5 answers Lack of decision recording → Recording information: 4 answers Lack of decision recording → Centralizing communication: 4 answers Lack of decision recording → More follow-up: 4 answers Lack of decision recording → Better documentation: 4 answers Difficulty defining a common schedule: 16 answers Difficulty defining a common schedule Difficulty establishing more organic communication: 4 answers Difficulty establishing more organic communication Lack of awareness of the multidisciplinary team: 9 answers Lack of awareness of the multidisciplinary team Decentralization in communication: 16 answers Decentralization in communication Difficulty monitoring the project: 16 answers Difficulty monitoring the project Difficulty obtaining detailed information: 25 answers Difficulty obtaining detailed information Difficulty communicating between hierarchical levels: 16 answers Difficulty communicating between hierarchical levels Disorganization of project information: 36 answers Disorganization of project information Difficulty maintaining a consistent process: 16 answers Difficulty maintaining a consistent process Lack of knowledge of standards of other projects: 9 answers Lack of knowledge of standards of other projects Lack of knowledge of the flow of information: 25 answers Lack of knowledge of the flow of information Lack of decision recording: 16 answers Lack of decision recording Integrating the teams better: 36 answers Integrating the teams better Centralizing communication: 28 answers Centralizing communication Adopting new tools: 23 answers Adopting new tools Well-established processes: 23 answers Well-established processes Leadership relaying crucial information: 15 answers Leadership relaying crucial information Recording information: 15 answers Recording information Daily/weekly written report: 13 answers Daily/weekly written report More follow-up: 12 answers More follow-up Following ceremonies/meetings: 11 answers Following ceremonies/meetings More communication between managers: 11 answers More communication between managers Better documentation: 10 answers Better documentation Annual sync meetings/newsletter: 7 answers Annual sync meetings/newsletter

The thickness of each link is proportional to the number of answers associating that symptom with that improvement. The questionnaire was administered in Portuguese; these labels are translations of the participants' original answers.

Credits

Authors

The team behind the study and its presentation at WASHES 2026 (11th Workshop on Social, Human and Economic Aspects of Software).

  • Student César Lobato

    PPCA · Federal University of Pará (UFPA)

  • Advisor Prof. Dr. Carlos Portela

    PPCA · Federal University of Pará (UFPA)

    csp@ufpa.br
  • Co-advisor Prof. Dr. Mauricio Souza

    PPGCC · Federal University of Lavras (UFLA)

  • Manager Profa. Dra. Paula Palomino

    Faculty of Technology (FATEC)

  • Vice-chair Prof. Dr. Rodrigo Lisbôa

    Federal Rural University of the Amazon (UFRA)

  • Chair Prof. Dr. Thiago Cordeiro

    PPGI · Federal University of Alagoas (UFAL)

Contributions

  • Empirical validation of a taxonomy of communication bad smells, extending the established concept of code smells to the human and social dimensions of software development.
  • Practical, context-aware mitigation strategies tailored to remote work constraints — rotating meeting times to accommodate time zones, and centralized knowledge bases to break down expertise silos.
  • A methodological approach demonstrating the value of combining quantitative communication metrics with qualitative insights to provide a comprehensive view of team dynamics.

Limitations and future work

  • The single-case study design suggests a need for broader validation across different organizational contexts and team structures.
  • Longitudinal research could help determine whether the proposed mitigation strategies lead to sustained improvements over time.
  • Emerging technologies such as AI-powered communication analysis may offer new opportunities for automatically detecting and addressing communication smells before they significantly impact team performance.

How to cite

Bad Smells of Communication in Multidisciplinary Agile Teams of a Remote Software Project. A study aligned with GranDSI-BR 2016–2026 (Boscarioli et al., 2017), on the challenge of collaboration, coordination and information sharing in complex and distributed socio-technical systems.