How to Set Up CustomerDNA

CustomerDNA is the layer of MorningAI that turns audience guesswork into specific, observable customer behavior. The new persona engine writes like a top-tier ethnographer, not a generic AI summary. You give it a brand and a few details; it gives you back a real human you can build campaigns around.

This guide walks you through generating, editing, and exporting personas in the redesigned CustomerDNA experience.

Why CustomerDNA Matters

Most AI personas read like LinkedIn fluff: "values quality and convenience, motivated by self-improvement." That tells a marketer nothing. The rebuilt CustomerDNA engine is grounded in behavioral realism rules: it reports specific brands, platforms, daily routines, and the gap between what customers say and what they actually do. That gap is where the marketing insight lives, and now MorningAI surfaces it natively for every persona you create.

Step-by-Step Setup

1. Open CustomerDNA. From the sidebar, hover DNA and click CustomerDNA. You will land on a grid of persona cards for your active brand.
2. Create a persona with AI. Click + New Persona to open the AI Create Persona modal. You will see a free-text prompt box, a B2C / B2B category toggle, and an Attachments slot that accepts up to three files (research PDFs, customer interview transcripts as .docx or .txt, or reference images).
3. Write a focused prompt. The more concrete you are, the better the output. Skip generic descriptors and lean into context. A good prompt looks like "Mid-30s working dad in suburban Dallas. Drives a Honda Pilot. Buys most things on Amazon Prime. Listens to Joe Rogan and ESPN on the way to drop kids at school." A weaker prompt is "A typical American dad."
4. Attach source material (optional). Drop in a customer interview transcript, a survey export, or even competitor screenshots. The AI uses them as evidence so the persona is grounded in your actual audience, not a stereotype.
5. Generate. Click Create. In a few seconds you will see a new persona card on the grid with an alliterative name (think "Aspiring Andrew" or "Shopping Susie") and a one-line summary that captures their core identity.
6. Open the persona detail modal. Click any card to open the full-screen detail view. Every field is editable inline. The default layout flows: Persona Name, Age, Gender, Location, Income, Values & Motivations, Needs & Pain Points, Attitude & Behaviors, Buying Triggers, Decision-Making Authority, Internal Stakeholders, Vendor Selection Criteria.
7. Add a face. Drag and drop an image into the persona avatar slot, or click to upload. A photo makes the persona feel real to the team using it.
8. Refine with AI Update. At the bottom of the detail modal is an AI Update prompt box. Type a natural-language instruction like "Make her more aspirational about fitness but admit she only works out twice a month," or "Add more detail about her LinkedIn behavior and the influencers she follows," and the persona regenerates with that change baked in.
9. Export to PDF. Click Export PDF in the detail modal. You get a polished, brand-ready document you can drop into a campaign brief, agency handoff, or leadership deck.

What a Great Persona Looks Like

The engine is instructed to report observable behavior, not stated preference. Expect outputs like:

• Persona summary: "A 28-year-old Brooklyn creative who aspires to minimalist fashion but impulse-shops fast fashion deals. Lives for Instagram validation and uses retail therapy to cope with dating app burnout."
• Specific brands and platforms: TikTok and Facebook (not "social media"). DoorDash and Instacart (not "delivery apps"). Mountain Dew and Monster Energy (not "beverages").
• Aspiration vs reality: "Pins healthy meal-prep boards on Pinterest but her cart is frozen pizzas and 2-liters of Dr Pepper."
• Emotional drivers and coping mechanisms: what gives them comfort, what they are escaping, what triggers an impulse buy.
• Daily routines, media diet, guilty pleasures: the texture a real ethnographer would capture in a field report.

If a generated persona drifts into sanitized filler ("appreciates quality," "values wellness"), use AI Update to push it back toward observed evidence: "Drop the generic wellness language. What does she actually buy on Sundays?"

B2C vs B2B Personas

Flip the toggle at the top of the AI Create modal to switch modes:

• B2C focuses on the individual: lifestyle, brands they buy, media they consume, emotional drivers, impulse triggers.
• B2B focuses on stakeholder roles such as Category Buyer, VP Sales, or Director of Marketing. Outputs include Decision-Making Authority, Internal Stakeholders, and Vendor Selection Criteria so you can map a real buying committee.

The behavioral realism rules apply to both. A B2B persona is still a human with a Slack habit, a podcast queue, and a coffee preference, not a job title.

Real-World Use Cases

Agency new-business pitch. Generate three personas for a prospective client, export each to PDF, and walk them through a strategy deck. Replaces tens of thousands of dollars of fieldwork with a 30-minute working session.

Campaign brief grounding. Before writing any creative, open the persona detail modal and read the Buying Triggers and Pain Points fields out loud to the team. Briefs that start here are sharper than briefs that start with a positioning statement.

Voice-of-customer audit. Drop a Zoom interview transcript into the Attachments slot. Generate a persona. Compare the AI's output to what the interviewee actually said. The places they diverge are where your assumptions were wrong.

Pro Tips

Start with the gap. The single most useful prompt instruction is some version of "capture the tension between what they say and what they do." That is where the marketing insight lives.

Iterate with AI Update. Treat the first output as a draft. Three or four AI Update passes ("add more detail about her weekend routines," "make him more skeptical of brand messaging") will produce a much stronger persona than trying to nail it in one prompt.

Use real evidence. The Attachments slot is the single biggest quality lever. Even one customer interview transcript will dramatically improve the realism of the output.

Name personas memorably. The alliterative names ("Aspiring Andrew") are not a gimmick. They make personas easier to reference in meetings and briefs months later.

Frequently Asked Questions

How many personas should I create? Most brands land on three to five. More than that and the team stops referencing them. Fewer than three and you risk over-generalizing.

Can I edit fields by hand instead of using AI Update? Yes. Every field in the detail modal is inline-editable. AI Update is for regenerating across multiple fields at once.

Will MorningAI generate personas for sensitive industries? Yes. The engine is instructed to allow all industries (alcohol, tobacco, cannabis, nightlife, firearms) without sanitization. The output respects the realities of the audience.

Can I share a persona outside MorningAI? Use the Export PDF button in the detail modal. The export is brand-ready and works as a deck slide, agency handoff, or campaign brief insert.

Does the persona affect content generated elsewhere in MorningAI? Yes. Briefs, Chat, and Studio pull persona context from CustomerDNA. A sharper persona produces sharper content downstream.

Next Steps

Open CustomerDNA in MorningAI and generate your first ethnographer-grade persona. When you are ready to put it to work, see How to Use CustomerDNA for Better Content.