Small-Agency Brand Voice: Review Three Cold Email Drafts With One Rubric

A small agency can produce three polished email drafts that sound like three different businesses. One is grandiose, another is oddly familiar and a third never explains why it is relevant. A shared rubric makes those differences visible before a reviewer approves the message.

OutreachBox’s inspected Brand Brain interface includes tone, person, reading level, instructions for what to do and avoid, banned phrases and evidence notes. The exercise below turns those controls into editorial decisions for a fictional agency.

Set the voice before comparing the drafts

Suppose the agency chooses a straightforward first-person plural voice, short sentences and specific observations. Its rules prohibit invented results, exaggerated praise and language implying an existing relationship when none exists.

These are example rules, not a description of every agency’s brand. Choose rules that reflect the actual business and the services it can substantiate. “Friendly” alone is too vague to resolve a disagreement between reviewers.

Write down the intended reader and the purpose of this first message. An email asking where to send a relevant observation needs a different ending from a proposal following an established conversation.

Draft one: polished but unsupported

Illustrative opener: “We are blown away by your incredible growth and know we can transform your conversions.” The sentence is enthusiastic, but neither the growth claim nor the promised transformation has evidence in this fictional brief.

A reviewer should return it for revision. Changing “incredible” to “impressive” would not fix the underlying issue. The agency needs a supported observation and a modest statement of what it offers.

Draft two: familiar without context

Illustrative opener: “Hey! Been following you forever and thought we should chat.” This fails if nobody at the agency has actually followed the recipient’s work. It also gives the reader no concrete reason to consider a conversation.

The problem is both factual and editorial. Even if the casual tone fits the brand, it does not justify an invented relationship or a vague request for time.

Draft three: specific and reviewable

Illustrative opener: “Your public services page links to a booking route that returned an error when we checked it today. Our agency reviews booking flows. Would you like the page and steps we checked?”

This is the strongest of the three only if those checks actually occurred and the sender offers that service. With no supporting observation, it fails too. The sample demonstrates a structure, not evidence about a real business.

Apply the same five questions

• Is every external factual claim supported by an attached source or observation?
• Does the wording match the chosen voice and person?
• Are any prohibited phrases, promises or invented relationships present?
• Can a new reader understand the service and its relevance?
• Is the next step proportionate to the evidence and relationship?

Use a short pass, revise or reject note for each question. A numerical style score can conceal a serious factual failure; do not average away a fabricated claim because the grammar is good.

Keep settings separate from verified enforcement

The inspected Brand Brain interface exposes editorial controls, but inspecting a screen does not prove every generation or sending path enforces them. Review the actual output and test the workflow you intend to use.

OutreachBox combines brand context with research and draft-review workflows. Create an account, save an agreed rubric and use it on the next three drafts before any recipient receives them.

Original paper-art illustration; not a photograph, site plan or product screenshot.