Personas Overview

Understanding Your Personas

Personas turn your customer list into a small cast of real, recognizable customer types — so you can see who your customers actually are and decide what to do next. This article explains what personas are, where they come from, and how to use them.

What is a persona?

A persona is a group of your customers who look and behave alike. Instead of thinking about thousands of individual buyers, personas boil your customer base down to a handful of distinct "types" — for example, Suburban Family Planners or Sneakerhead Collectors.

Every brand gets its own persona set: a collection of bespoke personas, built entirely from your own data. A persona set usually contains a small number of groups (typically between three and six), and each group represents a meaningfully different kind of customer.

Where personas come from

Decile creates your personas by looking at information from your own first-party data, and additional third-party enrichment attributes. For example:

• Your own purchase data — what your customers buy, how often, and how much they spend.
• Demographic and lifestyle data — details like age, household income, location, and family status, which Decile appends to your customer records.

Decile then looks across all of that information and finds the natural groupings that already exist in your customer base. Rather than you guessing "I think our core customer is a suburban mom," Decile discovers the real patterns in the data and turns them into groups.

Once your personas are built, every customer is placed into exactly one persona, and any new customers you acquire are automatically sorted into the right group as they arrive. This means your personas stay current over time without you having to do anything.

What each persona tells you

Open a persona and you'll see several layers of detail that describe who this group is and what they're worth:

• A name and short description. Each persona gets a friendly name, a descriptive phrase, and a few sentences that summarize who these customers are — something you can drop straight into a campaign brief or share with your team.
• Key attributes. The values that most represent this group from your other personas across key traits, such as their typical age range, income level, or regional location.
• Value metrics. How much this group is worth to you, including average lifetime value (LTV), average order value (AOV), and what share of your customer base they make up.
• Market size (TAM). An estimate of how many people in the wider U.S. population look like this persona — useful for judging how much room you have to find more customers like them.

How to read persona attributes

Personas describe the dominant characteristics of a group, not a rule that applies to every single customer in it.

For example, a persona labeled "Female" or "Suburban" means most of the customers in that group are female or suburban — not all of them. The label reflects the strongest, most distinguishing trait for that group. If a name ever seems off, it usually means the group is more mixed than the single label suggests, and looking at the full attribute breakdown will give you the complete picture.

Multiple persona sets and your primary set

Some brands may create more than one persona set in Decile — for example, an older set alongside a newer one, or sets built for different purposes. When that's the case, one set is marked as your primary persona set.

Each persona set is independent. A set is built from its own source group of customers and its own run of the model, so the personas in one set have no relationship to the personas in another. Names, groupings, and attributes don't carry over between sets, and a customer can sit in different personas depending on which set you're looking at. Because of this, it's best to treat each set on its own rather than trying to compare personas across sets.

Your primary set is the one Decile's AI features rely on. When you explore your customers through Luma or the Decile MCP, those tools read from your primary persona set only — your other sets aren't available there. So if you ask Luma about a persona and it doesn't recognize the name, it's likely defined in a non-primary set.

If you want a different primary persona set to power your AI experiences, please reach out to support@decile.com. Keeping a single, well-chosen primary set is the best way to get consistent, accurate answers from Decile's AI features.

Putting personas to work

Personas aren't just for understanding your customers — they're built to act on. Every persona is also an audience you can use elsewhere. Common ways teams use them:

• Find more of your best customers (acquisition). Spot which persona has the best mix of value and market size, then build acquisition campaigns aimed at people like them.
• Improve repeat purchases (retention). Route customers into tailored follow-up flows based on the persona they belong to, so your messaging matches who they are.
• Compare high-value vs. low-value groups. See what actually sets your most valuable customers apart from one-time buyers, and use that to guide strategy.

Because each persona is also created as an audience, you can sync it to your connected advertising and email platforms and target the customers within that persona directly.

Best practices

A few guidelines will help you get the most trustworthy, useful results from personas.

Avoid running personas repeatedly on the same group of customers. Persona sets are built by finding the best-fit grouping within a source population. Running the model again on that same group won't produce meaningfully new insight — because it's solving the same problem on the same data, each run may return slightly different results, even though they'll generally follow the same underlying patterns. Rather than regenerating, work with the persona set you already have.

Don't build personas on very small populations. Personas need a large enough group of customers to find reliable patterns. Running them on a narrow slice — for example, "customers who bought in January 2025" — tends to produce noisy, low-confidence results. If you want to understand a specific slice like that, a better approach is to take your existing personas and overlay them on that slice. For instance, look at January 2025 orders and see how they break down by persona, then compare that split to your overall customer base. That tells you whether that period skewed toward certain personas — which is usually the real question — without the pitfalls of clustering a tiny group.

Keep a single, well-chosen primary set. Since Decile's AI features read only from your primary persona set, maintaining one clear primary set gives you consistent answers across Luma and the Decile MCP.

Frequently asked questions

Do I have to build my personas? No. Decile generates your persona set automatically from your data. There's nothing to set up or configure.

How many personas will my persona set have? It varies by brand. Decile will automatically choose the right number based on your data — usually 3-5, so the results stay easy to digest. If you want to have a specific number of personas, you can specify when kicking off a new persona run.

Do personas update as I get new customers? Yes. New customers are automatically assigned to the persona they best match, so your personas reflect your current customer base over time.

Can a customer be in more than one persona? No. Each customer belongs to exactly one persona at a time.

Can I use a persona to run a campaign? Yes. Every persona is also an audience, so you can sync it to your connected ad and email platforms and target that group directly.

Are personas available in Decile's AI features? Yes. You can explore your personas through Decile's AI features, including Luma and the Decile MCP. Just note that these features use your primary persona set only. If your brand has more than one persona set, the personas from your other sets won't appear in AI features until you make one of them your primary set.

A persona's name doesn't seem to match everyone in it — is that a mistake? Usually not. A persona's name reflects the group's most dominant traits, not a description that fits every member perfectly. Check the full attribute breakdown to see the complete makeup of the group.