Brand Voice Calibration (Onboarding)

Brand Voice Calibration (Onboarding)

1. 01

   Extract client brand guidelines from website and provided documents.

   Use the platform's training feature to ingest the client's URL, brand PDFs, and style guides. This seeds the AI with foundational tone and terminology.

2. 02

   Define a list of forbidden words and phrases specific to the client's industry.

   Include competitor names, overused jargon, and terms the client avoids. For example, a financial services client may forbid 'guaranteed returns'.

3. 03

   Generate a sample batch of 10 social posts using the AI and review for brand alignment.

   Check for tone consistency, factual accuracy, and adherence to the client's messaging pillars. Flag any generic or off-brand language.

4. 04

   Adjust the AI's temperature and style parameters to match the desired voice.

   Lower temperature (e.g., 0.3) for formal B2B clients; higher (e.g., 0.7) for casual consumer brands. Re-run the sample batch after each adjustment.

5. 05

   Create a brand voice reference document with before/after examples.

   Include 3-5 pairs of AI-generated drafts and the final approved versions. This serves as a training artifact for future iterations and new team members.

6. 06

   Set up automated content review rules for high-risk topics.

   Configure the platform to flag posts containing financial figures, health claims, or client-specific product names for mandatory human approval before publishing.

7. 07

   Document the calibration settings and share with the client for sign-off.

   Provide a one-page summary of the voice parameters, forbidden words, and review rules. Obtain written approval to lock the configuration for the first month of delivery.