Comms Workflow Audit
A structured framework for mapping existing communications workflows, identifying inefficiencies and bottlenecks, and pinpointing where AI can add the most value — before buying tools or redesigning processes.
What it is
AI adoption without workflow clarity is technology bolted onto chaos. Before you can improve a process with AI, you need to know what the process actually is — where it starts, what steps it goes through, who does what, where time gets consumed, and where things go wrong.
The Comms Workflow Audit maps the key recurring workflows in a communications function at two levels. First, it captures the current state honestly: how things actually work today, not how they’re supposed to work on paper. Second, it identifies where in each workflow AI could add the most value — not theoretically, but specifically, with reference to the tasks being done and the time they consume.
The outputs of this audit have two uses. Internally, they show the team where inefficiency lives and create a shared map of the work. Externally (for consultants working with clients, or leaders presenting to boards), they build the evidence base for AI investment: “here is where we spend our time, here is where AI would save it, here is what that’s worth.”
This is a diagnostic and mapping tool, not an implementation plan. It tells you where to act. The implementation detail follows in the Comms Workflow Audit’s sister templates: the AI Tool Evaluation Framework, the Capability Gap Analysis, and the AI Readiness Assessment.
When to use it
Use this template when:
- You’re planning to introduce AI tools into your communications workflows and want to start with the right ones
- You suspect significant time is being lost in certain workflows but haven’t mapped where
- You’re preparing a business case for AI investment and need to quantify the time cost of current processes
- You’ve just completed an AI Readiness Assessment and want to follow up on workflow-readiness gaps
- A new leader is trying to understand how the communications function actually operates
Don’t use this template when:
- You’ve already decided which tools to adopt and just need to implement (though going back and doing this first is still worth it)
- You’re mapping approval workflows specifically (use the Approval Workflow Mapper in Phase 3)
- You need a complete operating model redesign rather than a workflow-level audit
- You have a very small team (under 3 people) with highly varied work — mapping workflows works best where there’s repetition
Inputs needed
- Time to interview or survey the people who do the work — this shouldn’t be a desk exercise alone
- Honest estimates of time spent on tasks (rough and rounded is fine — the goal is pattern, not precision)
- Information on which tools are used at each stage (even if it’s just “email” and “Word”)
- Willingness to describe actual current practice, not ideal practice
- Optional: a week’s worth of calendar or time-tracking data to sense-check time estimates
The template
Comms Workflow Audit
Organisation: [Name] Function: [Communications team / department] Audit completed by: [Name] Date: [Date] Method: [Team workshop / Individual interviews / Self-completion / Mixed]
Step 1: Identify core workflows
List the main recurring workflows in the communications function. These should be activities the team does regularly (weekly, monthly, or for every campaign/project), not one-off tasks.
Core workflow inventory:
| Workflow | Frequency | Who does it | Typical time per cycle | Priority for audit |
|---|---|---|---|---|
| [e.g., Content creation: LinkedIn posts] | [Daily / Weekly / Per campaign] | [Role(s)] | [Hours] | High / Medium / Low |
| [e.g., Media monitoring and briefing] | ||||
| [e.g., Press release drafting] | ||||
| [e.g., Stakeholder reporting] | ||||
| [e.g., Content approval process] | ||||
| [e.g., Campaign planning and briefing] | ||||
| [e.g., Internal comms drafting] | ||||
| [e.g., Social media scheduling] | ||||
| [e.g., Media pitch development] | ||||
| [e.g., Analytics reporting] |
Select 3–5 highest-priority workflows for detailed audit below.
Step 2: Detailed workflow mapping
Complete one block per selected workflow.
Workflow: [Name] Frequency: [How often] Total estimated time per cycle: [Hours] Estimated annual time cost: [Hours × frequency]
Current workflow steps:
| Step | Description | Who does it | Time estimate | Tools used | Quality/difficulty |
|---|---|---|---|---|---|
| 1 | [What happens] | [Role] | [mins/hrs] | [Tool(s)] | High / Medium / Routine |
| 2 | |||||
| 3 | |||||
| 4 | |||||
| 5 | |||||
| 6 |
Where time is disproportionately consumed: [Which 1–2 steps eat the most time relative to their value?]
Where quality most often suffers: [Which steps are most prone to inconsistency, error, or rework?]
The main bottleneck in this workflow: [Where does the workflow get stuck, slowed, or blocked? What causes it?]
Current pain points (from the people doing the work):
AI opportunity assessment:
| Step | AI potential | Specific AI capability needed | Readiness for AI | Risk level |
|---|---|---|---|---|
| [Step 1] | High / Medium / Low / None | [e.g., research synthesis, first-draft generation, data extraction] | Ready / Needs prep / Not ready | High / Medium / Low |
Best AI integration point in this workflow: [The single step where AI would have the highest impact-to-effort ratio]
Redesigned workflow with AI (sketch): [Briefly describe how this workflow would look with AI integrated at the key step(s). What steps would change? What would the human still do?]
Estimated time saving if AI integrated: [Best estimate of time saved per cycle — be conservative; real savings typically come with learning curves]
[Repeat block for each selected workflow]
Step 3: Cross-workflow patterns
After mapping individual workflows, look for patterns across them.
Most common time-intensive task types:
| Task type | Appears in which workflows | Total estimated annual time | AI potential |
|---|---|---|---|
| [e.g., First-draft writing] | High / Medium / Low | ||
| [e.g., Research and fact-gathering] | |||
| [e.g., Data analysis and reporting] | |||
| [e.g., Formatting and layout] | |||
| [e.g., Approval routing] | |||
| [e.g., Scheduling and distribution] |
Where AI has the highest cross-workflow potential: [What task type, appearing across multiple workflows, would deliver the biggest aggregate time saving if AI-assisted?]
Where AI has the lowest potential (or is inadvisable): [Which tasks or workflow stages should remain human-led, and why?]
Step 4: Readiness assessment by workflow
For each workflow audited, assess the readiness for AI integration:
| Workflow | Readiness | Primary barrier | Recommended first action |
|---|---|---|---|
| Ready / With prep / Not ready | Skills / Tools / Process clarity / Governance | ||
Step 5: Prioritisation matrix
Plot each workflow against two axes: AI impact potential (high/low) and AI integration effort (high/low). The goal is to identify the high-impact, low-effort integrations to start with.
Prioritisation:
| Quadrant | Workflows in this quadrant | Recommended approach |
|---|---|---|
| High impact, low effort | Start here — Quick wins to build confidence | |
| High impact, high effort | Plan for later — Valuable but needs investment | |
| Low impact, low effort | Consider — Easy but limited return | |
| Low impact, high effort | Avoid/defer — Least valuable use of resources |
Summary and recommendations
Top 3 workflow integration opportunities:
| Opportunity | Workflow | AI use case | Estimated annual time saving | Priority |
|---|---|---|---|---|
| 1. | ||||
| 2. | ||||
| 3. |
Workflow foundations to build first: [Are there any workflows too poorly defined for AI to improve? List them and the prerequisite work needed.]
Governance considerations identified: [Any workflows where governance or risk considerations need addressing before AI integration]
Recommended first 60 days: [The specific, sequenced actions for beginning AI integration based on this audit]
AI prompt
Base prompt
I've mapped our key communications workflows and want help identifying where AI would add the most value and how to sequence adoption.
Our organisation: [DESCRIBE]
Team size and structure: [DESCRIBE]
Here are our top 3 workflows with time estimates:
Workflow 1: [NAME]
Steps: [LIST STEPS WITH TIME ESTIMATES]
Main bottleneck: [DESCRIBE]
Current tools: [LIST]
Workflow 2: [NAME]
Steps: [LIST]
Main bottleneck: [DESCRIBE]
Current tools: [LIST]
Workflow 3: [NAME]
Steps: [LIST]
Main bottleneck: [DESCRIBE]
Current tools: [LIST]
Please:
1. For each workflow, identify the 1–2 steps with the highest AI potential and what specific AI capability would help
2. Estimate the realistic time saving per cycle if AI were integrated at those steps
3. Identify which workflow integration is most ready to begin now vs. which needs preparation
4. Suggest what a redesigned version of each workflow would look like with AI integrated
5. Recommend a sequenced 90-day AI integration plan based on these workflows
Be specific. "Use AI for content creation" isn't useful — I need to know which step, which AI capability, and what the human still does.Prompt variations
Variation 1: Single workflow deep dive
I want to redesign one specific workflow to integrate AI effectively. Here's how it currently works:
Workflow: [NAME]
Frequency: [HOW OFTEN]
Who does it: [ROLES]
Steps:
1. [DESCRIBE STEP — time, tools, who]
2. [DESCRIBE STEP]
3. [DESCRIBE STEP]
[etc.]
Current pain points:
[DESCRIBE]
Constraints:
- Governance requirement: [Any content that requires human review before publication]
- Quality standard: [What does a good output look like]
- Existing tools we must use: [LIST]
Please redesign this workflow to integrate AI in the most valuable way. For each AI-assisted step:
1. What specifically does the AI do?
2. What does the human do (before, during, and after the AI step)?
3. What is the realistic time saving?
4. What could go wrong and how is that managed?
Show the redesigned workflow as a step-by-step process, not just a description.Variation 2: Time-cost analysis
I need to build a business case for AI investment. Here are the workflows I've mapped:
[DESCRIBE WORKFLOWS WITH TIME ESTIMATES]
Team details: [SIZE, AVERAGE SALARY BAND OR DAY RATE]
Please:
1. Estimate the total annual time cost of these workflows at current efficiency
2. Model realistic time savings from AI integration at each workflow (use conservative estimates — what AI consistently delivers in practice, not best-case scenarios)
3. Calculate the potential annual ROI in time and, where appropriate, cost
4. Identify which single investment would have the highest ROI
Format this as a business case table I could present to a finance audience.Variation 3: Bottleneck diagnosis
Our communications team is consistently overloaded and I'm trying to understand where the time is actually going.
Here's how the team spends its time (rough estimates):
[DESCRIBE: by workflow or task type, with approximate % of time]
The team's main complaints are:
[LIST THE MOST COMMON FRUSTRATIONS]
We currently use these tools:
[LIST]
Please:
1. Based on this picture, identify the most likely sources of inefficiency
2. Diagnose which bottlenecks are workflow design problems vs. skills/capacity problems vs. technology gaps
3. Recommend whether AI is the right solution for each bottleneck, or whether something else (better briefing, clearer process, additional resource) would help more
4. Prioritise the top 3 changes with the highest impact on team capacity
I want honest diagnosis, not just an argument for AI tools.Human review checklist
- Real practice mapped, not ideal practice: The workflows reflect how the team actually works, not the official process
- Time estimates from practitioners: Time estimates came from the people doing the work, not management assumptions
- Bottlenecks genuinely identified: The identified bottlenecks are where things actually slow down, not where they theoretically should
- AI recommendations are specific: “Use AI for this step” includes what specific AI capability is needed, not just a general endorsement
- Human role preserved: Each AI integration recommendation specifies what the human still does — AI doesn’t remove the human entirely
- Governance steps maintained: Any content that requires approval or review still has that step in the redesigned workflow
- Effort estimate realistic: Time saving estimates are conservative, accounting for prompt iteration and output review time
- Prioritisation based on evidence: The recommended starting points are genuinely high-impact and high-readiness, not just exciting
- Workflow foundations checked: Where workflows are too poorly defined for AI to improve, this is flagged rather than ignored
- Next steps specific: The 60-day plan names specific actions, not just “begin AI integration”
Example output
Comms Workflow Audit Organisation: Brightfield Housing Association | Audited by: Faur Consulting Date: April 2026
Workflow 1: Monthly stakeholder newsletter (current state)
| Step | Time | Who | Pain point |
|---|---|---|---|
| Gather updates from 6 departments | 3 hrs | Senior Comms Manager | Chasing responses; inconsistent input format |
| Draft content | 2.5 hrs | Comms Manager | Starting from scratch each time |
| Edit and format | 1 hr | Senior Manager | Multiple rounds of minor edits |
| Approval from CEO and 2 Directors | Variable (1–5 days) | Admin | Approval often delayed; no clear deadline |
| Publish and distribute | 30 mins | Comms Coordinator | Straightforward |
Total: 7–9 hours per cycle | Monthly | ~96–108 hours per year
Redesigned with AI:
| Step | Change | AI role | Time (new) |
|---|---|---|---|
| Gather updates | Standardised input form with consistent fields | No AI; process fix | 1.5 hrs |
| Draft content | AI first-draft from input form data | AI: synthesis and draft generation | 45 mins |
| Edit and format | Human review, voice/brand alignment, factual check | Human-led | 45 mins |
| Approval | Approval process unchanged | No change | Variable |
| Publish | Unchanged | No change | 30 mins |
Estimated time saving: 4–5 hours per cycle | ~50–60 hours per year
Related templates
- AI Readiness Assessment — Complete before this audit to understand overall readiness context
- AI Tool Evaluation Framework — Use the workflows identified here to evaluate which tools best serve those specific needs
- Approval Workflow Mapper — Specific template for mapping approval and sign-off workflows
- Capability Gap Analysis — Assess whether the team has skills to use AI in the workflows identified
- Objectives & Measurement Framework — Set measurable objectives for workflow improvement
Tips for success
Map real workflows, not aspirational ones The most common mistake is describing how a workflow is supposed to work. The interesting data — where time is lost, where quality suffers, where things go wrong — only appears if people describe what actually happens. Create a safe environment for honest description.
Talk to the people who do the work, not just the people who manage it Managers often describe idealised versions of workflows because they’re not in them day to day. The people doing the work know where the friction is. Run the audit with practitioners, validated by management, not the other way around.
Quantify time before making AI claims “This workflow takes too long” is not a business case. “This workflow takes 8 hours per cycle, runs weekly, and AI could reduce it to 3 hours, saving 260 hours per year across the team” is. Rough estimates are fine; the order of magnitude matters more than precision.
Look for the repetition AI adds most value to tasks done repeatedly, with similar inputs and expected output structures. One-off creative projects are harder to improve with AI than templated, recurring processes. Find the repetition in the work first.
Redesigned workflows still need humans Every AI-integrated workflow should clearly specify what the human still does. AI removes steps that are time-intensive and low-judgement; it doesn’t remove the need for expertise, editorial judgement, or accountability. If your redesigned workflow has the human only clicking “approve”, you’ve probably over-automated.
Common pitfalls
Mapping the exception, not the rule Every workflow has outlier cases — the press release that took three weeks instead of three days, the stakeholder report that required five rounds of changes. Map the typical flow, not the worst case or the best case. Outlier handling can be noted separately.
Confusing workflow redesign with tool selection “We should buy Jasper” is not a workflow redesign. The workflow audit tells you what needs to change. Tool selection (using the AI Tool Evaluation Framework) comes after you know what you’re trying to improve. Don’t start with tools.
Ignoring the approval steps Approval and review steps often consume as much time as production steps, but they get less attention because they feel like governance rather than “workflow.” They’re part of the workflow. Mapping them reveals opportunities for both AI assistance (drafting review notes, tracking approval status) and process improvement.
Treating all time saving as equal value Not all hours are equivalent. Saving 5 hours of a senior strategist’s time is more valuable than saving 5 hours of coordination admin. When prioritising, think about whose time you’re freeing up and what they’d do with it.
Not revisiting the audit Workflows evolve. A team that adopts AI will have different workflows in 12 months than it has today. Re-run the audit annually, or whenever the team structure, tools, or work type changes significantly. The map goes stale.
Related templates
Need this implemented in your organisation?
Faur helps communications teams build frameworks, train teams, and embed consistent practices across channels.
Get in touch ↗