How to assess written communication for a distributed role
The fairest way to assess written communication for a distributed role is a short, role-related work sample with a clear rubric. Ask the candidate to write something the job actually requires, then score clarity, context, escalation, audience awareness, and actionability, and leave accent, personality, and personal writing style out of the score.
A good exercise should take about 20–30 minutes, use the same prompt for comparable candidates, and make the expected audience and outcome explicit. The Equal Employment Opportunity Commission has stated that employment laws apply to AI and other technologies used in recruiting and selection, including when a seemingly neutral practice creates unjustifiable disparate impact. [1]
Build the prompt from the role
Distributed work creates a writing requirement because decisions, handoffs, and blockers often move through written channels. Test that work directly.
Use a short prompt with four parts:
Situation: what happened and why it matters
Audience: who will read the message
Constraint: what is known, missing, or sensitive
Action: what the reader needs to decide or do next
Examples:
Finance: “Write a 150-word variance note to the controller. Revenue is 6% below plan in one region, two invoices are still unposted, and the source report is not final. State what you know, what you would verify, and what decision or follow-up you need.”
Sales: “Write a follow-up to a prospect after a discovery call. The prospect has a deadline but has not confirmed the buying process. Summarize the business problem, ask one useful question, and propose a next step without overstating certainty.”
Support: “Write a response to a customer whose report is delayed by a known issue. Acknowledge the impact, explain what is confirmed, avoid promising an unapproved date, and state the escalation path.”
These prompts test communication as it happens on the job, using scenarios kept separate from real company deliverables.
Use a simple evaluation rubric
Score each dimension from 0 to 2. Define the rubric before reviewing submissions.
Dimension | 0 | 1 | 2 |
|---|---|---|---|
Clarity | Hard to follow or missing the point | Understandable with avoidable ambiguity | Direct, organized, and easy to act on |
Context | Repeats facts or invents certainty | Includes some relevant context | Selects the facts the audience needs |
Escalation | Hides risk or escalates everything | Notices a risk but gives no path | Names the risk, owner, timing, and next step |
Audience and respect | Blunt, confusing, or mismatched to reader | Adequate tone with uneven focus | Respectful, specific, and appropriate to the role |
Actionability | No clear request or next step | Request is implied | Reader knows exactly what to decide or do |
Use the score to guide the conversation alongside your judgment. Record one strength, one concern, and one follow-up question. Keep comments tied to the submitted work.
Make the exercise fair and accessible
Tell candidates the time limit, expected length, audience, and whether outside tools are allowed. If the role permits AI-assisted drafting, let candidates use it openly. Ask them to identify what they used, what they changed, and how they checked the result; evaluate the final communication and the candidate’s judgment against the role.
Offer a reasonable accommodation route separately from the hiring manager when possible. The federal ADA regulations allow employers to use criteria that are job-related and consistent with business necessity, and allow an applicant to describe or demonstrate how they would perform job-related functions with or without reasonable accommodation. [2]
In practice, that may mean an alternate response format, assistive technology, additional time, or a different way to demonstrate the same essential communication outcome. Keep diagnoses out of the exercise entirely. Score only the grammar conventions the role requires, and judge clarity by whether the reader can act on the message, whatever the accent, idiom, or level of polish.
If automated scoring is involved, use it only as a flag for human review. The EEOC’s 2024 guidance notes that AI-enabled recruiting, screening, and hiring tools can create discrimination risk, including when disability-related differences affect how a system interprets speech or other signals. [1]
Close with a short follow-up
After the writing sample, ask one or two questions:
“What assumption did you make?”
“What would you verify before sending this?”
“When would you escalate this, and to whom?”
“How would you change the message for a customer, peer, or executive?”
The follow-up tests whether the candidate can explain decisions and respond to context. Keep it short and structured so it measures reasoning.
Keep a reusable assessment record
For each candidate, retain:
The exact prompt and allowed resources
The submission and time window
The rubric score by dimension
One evidence-based note per score
Any accommodation provided, stored confidentially and separately from ordinary evaluation notes
The decision and the job-related reasons for it
This creates consistency across interviewers and makes the process easier to improve. It also helps the team distinguish a writing issue from a missing process skill, such as failing to identify the owner of an unresolved finance exception.
FAQs
Should the exercise be timed?
Use a reasonable time window that reflects the role. State it in advance and consider accommodations that change how the candidate demonstrates the same essential skill.
Should we test grammar?
Only where the job requires a defined level of copy accuracy or editing. For most distributed roles, prioritize meaning, context, judgment, and a clear next step.
Should candidates be allowed to use AI?
Match the exercise to the actual job. If AI use is allowed in the role, allow it transparently and assess how the candidate reviews and improves the output. If it is not allowed, say so before the exercise.
What is a strong passing signal?
The reader can understand the situation, trust the boundaries of what is known, see the next step, and tell who owns it.
Let our talent team run the screen
Crossbridge’s talent team personally interviews every candidate, including how they communicate in writing for distributed work. You receive 3 to 5 vetted profiles for your role, with no upfront cost and a 60-day replacement guarantee.
Get candidates for your open role or explore Nearshore Staffing.