Introduction to System Analysis

Documentation & Communication Skills

18 min Lesson 8 of 10

Documentation & Communication Skills

A systems analyst who uncovers a perfect solution but cannot explain it clearly has not yet finished the job. Documentation and communication are the analyst's primary deliverables: they translate the raw findings of interviews, workshops, and analysis sessions into artefacts that decision-makers can approve, developers can build from, and testers can verify against. This lesson focuses on two critical competencies — writing for different audiences and facilitating meetings to present findings effectively.

Writing for Audiences: The Audience-First Principle

Every document an analyst produces has at least two audiences: a business audience (executives, department heads, end-users) and a technical audience (developers, DBAs, QA engineers). These groups have different vocabulary, different tolerances for detail, and different stakes in the outcome. Writing a single-version document that tries to serve everyone usually serves no one well.

The audience-first principle demands that you answer three questions before writing a single sentence:

  1. Who will read this? — Identify each reader group and what they need to decide or do after reading.
  2. What do they already know? — Match your vocabulary to their existing mental model. A clinic administrator knows "appointment slots" and "triage"; she does not know "foreign-key constraint" or "REST endpoint".
  3. What action does this document drive? — A feasibility report drives an approval decision; a requirements specification drives development; a test plan drives quality verification. Structure the document around that action.

Document Layers: Abstraction in Practice

A practical technique is to write the same finding at three levels of abstraction and publish the level appropriate for each reader group.

Three-layer document abstraction Executive Summary Audience: C-suite, board 1–2 pages, no jargon Business Requirements Audience: managers, users Functional language, examples System Specification Audience: developers, QA Clinic Booking System — Same Finding, Three Levels Executive: "The current booking process loses 18% of appointments; a digital system will reduce that to under 2%." Business: "Staff must be able to confirm, reschedule or cancel any appointment within 3 clicks from the main dashboard." Priority: High. Owner: Reception Manager." Technical: "PATCH /appointments/{id} must accept status enum (confirmed | rescheduled | cancelled) and return 200 within 400 ms (p95). Audit log entry required."
The same finding expressed at three abstraction levels for three different audiences in a clinic booking project.

Structuring Documents for Clarity

Well-structured documents share a common skeleton regardless of type:

  • Purpose statement — one paragraph that says exactly what the document is, who it is for, and what decision or action it supports.
  • Executive summary — key findings and recommendations, written last but placed first. Busy managers read this and stop.
  • Detailed body — findings, analysis, requirements, or specifications, organised by topic not by the order you discovered them.
  • Appendices — raw interview notes, full data tables, glossary. Keeps the body uncluttered.
Use consistent terminology. If you call it an "appointment slot" in the first section, do not switch to "booking window" or "time block" later. Inconsistent naming is one of the most common causes of developer misunderstandings and expensive rework.

Common Analyst Documents and Their Owners

For reference, the table below maps typical analysis artefacts to their primary audience and the decision each one drives. An online-store project illustrates each:

  • Project Charter / Vision Document — Audience: sponsor, steering committee. Decision: authorise the project. Example: "Launch a B2C electronics store by Q4."
  • Business Requirements Document (BRD) — Audience: business stakeholders. Decision: confirm scope and priorities. Example: "Customers must complete checkout in under 4 steps."
  • System Requirements Specification (SRS) — Audience: development team, QA. Decision: begin design and coding. Example: "Payment gateway must support Visa, Mastercard, and Apple Pay."
  • Process Model (swimlane diagram / BPMN) — Audience: process owners, analysts, developers. Decision: confirm understanding of the current or future process.
  • Meeting Minutes / Action Log — Audience: all participants. Decision: confirm agreements and track follow-up. Sent within 24 hours of the meeting.
Traceability matters. In a mature analysis practice, every requirement in the SRS can be traced back to a business requirement in the BRD, and every business requirement to a specific stakeholder need. This trace matrix prevents scope creep and makes impact analysis fast when requirements change.

Meeting Facilitation: Turning Talk into Decisions

Analysts spend a large share of their time in meetings — requirements workshops, review sessions, sign-off meetings, and progress check-ins. Poorly facilitated meetings are expensive: a two-hour workshop with twelve stakeholders from a logistics firm costs roughly 24 person-hours. If the meeting produces no decision and requires a follow-up, that cost doubles.

Effective facilitation follows a three-phase structure:

  1. Before the meeting — Distribute an agenda at least 24 hours in advance. State the objective (decision, review, or brainstorm — not "discussion"). Attach pre-reading so participants arrive informed. Identify a scribe (different from the facilitator).
  2. During the meeting — Open with the objective and time-box. Use structured techniques: round-robin for collecting opinions, dot voting for prioritising, parking lot for out-of-scope items. Summarise decisions aloud before moving on. Watch for dominant voices; actively invite quieter participants.
  3. After the meeting — Distribute minutes within 24 hours. Each action item must have an owner, a due date, and a clear definition of done. Follow up at the next meeting.
Meeting facilitation lifecycle Before Send agenda (24h) State clear objective Attach pre-reading Assign scribe During Open: objective + timebox Round-robin input Dot-vote priorities Parking lot out-of-scope Summarise decisions aloud Invite quiet voices After Minutes within 24h Owner + due date per item Definition of done Review at next meeting Facilitation produces decisions, not just conversations.
The three-phase meeting facilitation lifecycle: Before, During, and After. Each phase has specific outputs that drive the next.

Presenting Findings: Making Analysis Actionable

Presenting findings to stakeholders is different from presenting to peers. Stakeholders are making business decisions, not evaluating your analytical rigor. Structure your presentation around recommendations first — state what you recommend, then justify it. This is the Pyramid Principle (Barbara Minto): lead with the conclusion, support it with grouped arguments, back each argument with data.

In a logistics firm scenario, a weak presentation opens with: "We interviewed 12 staff and reviewed 400 shipment records over three weeks…". A strong presentation opens with: "We recommend replacing the manual dispatch log with an integrated TMS — this will cut average dispatch time from 47 minutes to under 8 minutes and reduce misfiled shipments by 90%. Here is the evidence."

Use the So-What test. For every slide or section, ask "so what?" — what does this mean for the audience? If you cannot answer immediately, that section is data without insight. Either add the insight or cut the section.

Handling Disagreement and Conflicting Stakeholders

Findings sometimes conflict with what powerful stakeholders believed or hoped. Skilled analysts separate the finding from the recommendation, cite evidence clearly, and avoid personal language. Instead of "the Finance team's process is broken", write "the current month-end reconciliation takes 3.5 days; industry benchmarks suggest 1 day is achievable with the proposed system." The evidence does the arguing; you remain neutral.

Avoid analysis paralysis in documentation. A 200-page requirements document that takes six months to perfect is not better than a 40-page document that reaches stakeholders in four weeks. Perfect documentation that arrives after decisions are made is worthless. Prefer concise, timely, versioned documents over exhaustive delayed ones.

Digital Communication: Async and Remote Realities

Modern analysis work spans time zones and remote teams. Much communication is now asynchronous — Slack, email, Confluence, Jira comments. Async writing must be more self-contained than live conversation: provide context, state the question or decision needed, and set a clear response deadline. A message like "Thoughts?" achieves nothing; "Please confirm by Friday whether the clinic system needs offline mode — this affects our architecture choice" drives a decision.

Version-control your documents. A requirements document that exists only as BRD_v3_FINAL_edits_v2.docx in email threads is a liability. Use a shared wiki (Confluence), a version-controlled folder (Git or SharePoint), or at minimum a naming convention with dates: BRD-ClinicBooking-2026-06-10.pdf. Always note what changed between versions.

Mastering documentation and communication transforms analysis from a private intellectual exercise into a shared, actionable foundation for the entire project team.