Persona Research For A Second Audience: Rank, Don't Average

Stage two of eight. The question it answers: how do we serve a second audience without breaking what works for the first?

A manufacturer's digital lead put the problem plainly: the website drives most of the business, almost all of it from homeowners, and the professionals who specify and install the product have been deprioritized for years. They want to serve the professionals without breaking what works for the homeowner. Their research on professionals is more robust than their research on homeowners, which is several years old.

That shape repeats. A membership organization built for its members now needs to recruit new ones. A university site built for undergraduates now competes for postgraduates. A consumer brand that grew up direct now has a trade channel worth as much as the storefront. In every case the primary audience is the one the structure was built around, and the secondary audience is the one leadership is now measured on.

The temptation is to average. One navigation that gestures at both. One tone somewhere between inspiration and specification. One journey with a fork halfway down. It feels fair, and it fails, because the two audiences arrive with different intents, different vocabularies, and different states of knowledge, and a structure that serves the midpoint serves nobody.

The alternative is to rank. Decide which audience the primary structure serves, decide what the other audience gets, and let that decision drive the research rather than follow it.

Ranking is a business decision, not a research finding

Research will not tell you which audience to rank first. It will tell you what each audience needs, and it will do so loudly. The ranking comes from the business: which audience carries revenue now, which carries growth, and what the brand's positioning says it is for.

On a global certification body's site, the ranking was made before the research was designed. Existing certified members were the large majority of the audience and needed speed: find a course, book a trip, renew, move on. Prospective members were a minority of the audience and the entire growth story, and they needed inspiring before they needed anything transactional. Professionals, the third group, needed functional efficiency and a set of tools that had nothing to do with either. The business's own language for its three journeys, start, keep, and teach, became the spine of the segmentation, which meant the research results landed in a vocabulary the organization already used.

That order dictated what the site's primary structure would optimize for and where the growth audience would get its own path. It also protected the research from the loudest finding. Prospects, in the research, wanted a low-barrier way to try the activity once. That was true and it was not the business's revenue driver. Because the ranking had been made, the finding became a design question for the journey stage rather than a reversal of the strategy.

The ranking changes how you recruit

Once the audiences are ranked, the research cannot recruit one pool and segment it afterward. Each audience needs its own source, its own screener, and often its own method, because they are not findable in the same place.

On the certification body's project, in research conducted with a partner, professionals were recruited from the organization's own professional list, existing members from the member database, and prospects, who by definition were not in any database, through a general panel screened for adjacent interests rather than for the activity itself. The prospect recruit was the hardest and the most important: it is the audience the business knows least and the one the redesign exists to win.

The manufacturer's version of this is the same shape. Homeowners exist in the lead database. Professionals exist in the account list and at the branch counter. Prospective homeowners who have never contacted a branch exist nowhere the business can see, and reaching them means going outside the customer base, which is where most redesign research quietly stops.

Sub-segments matter more than the persona

A persona is a summary. The decisions get made one level down.

Inside the certification body's prospect audience, the synthesis surfaced four sub-segments with different reasons for hesitating: people who had never considered the activity, people drawn to it but held back, people with a specific fear to overcome, and families deciding together. Each needed a different first piece of content and a different first step. A single "prospect" journey would have served the average of the four, which is nobody. There was also a crossover: people who had done an introductory experience and were deciding whether to continue, who appeared under both prospect and member depending on the moment.

The practical rule is that persona research is not finished when the personas are named. It is finished when each persona has been broken into the sub-segments whose first content differs, and when the crossovers between personas have been drawn. Those are the units the journey stage and the structure stage actually design for.

Anonymous and known are different audiences

There is a state most persona work ignores: the visitor the site cannot identify.

On the certification body's site, the anonymous state existed almost entirely for prospects. Members and professionals arrived logged in or with a history; their journeys only made sense identified. Prospects arrived cold, and the analytics could not segment anonymous visitors by persona because the persona is not observable, only behavior is. So the prospect journey had to work on behavior alone: what they looked at, where they came from, what they searched. That constraint shaped the personalization strategy later and the structure immediately: the entry points a prospect lands on have to do the sorting a login would otherwise do.

For a multi-location manufacturer the split runs the same way. Professionals with an account are known. Homeowners with a lead are known. The homeowner who has not yet contacted a branch, the one the site exists to convert, is anonymous, and the structure has to serve them without knowing who they are.

Sell the experience, then the product

One research theme recurs across audiences and industries: prospects buy an experience and are sold a specification. The certification body's research put it as selling experiences rather than technical products. The manufacturer's homeowner journey is described internally as inspiration-led for the same reason. A prospect does not want the product taxonomy; they want to see the outcome and be led to the product that delivers it.

The professional audience is the mirror image. They want the specification, fast, and every inspiration layer between them and it is friction. That asymmetry is the whole reason the audiences have to be ranked rather than averaged: the two entry experiences are not variations on one design. They are different designs that share a structure underneath.

Who owns the segments after the agency leaves

Segment definitions decay. On the certification body's project, the sub-segments were defined in a synthesis workshop, and weeks later the client's own team asked what the names meant. The definitions had lived in the agency's documents rather than in the client's vocabulary.

The fix is procedural. The segments get written into a document the client owns, in the client's language, with the recruiting source and the first content decision for each, and they get a named owner on the client side. Persona research that lives only in a readout is research the organization will redo in three years, which is roughly how old the manufacturer's homeowner research is now.

What the audiences stage hands over

• The ranking: which audience the primary structure serves, which gets its own path, and why, in one page the business signs.
• Personas with their sub-segments and crossovers, in the client's own journey vocabulary.
• The recruiting matrix: source, screener, and method per segment, so the research can be repeated.
• The anonymous-versus-known map: which audiences the site can identify, and what the anonymous journey has to do without identity.
• The first-content decision per sub-segment, which becomes an input to the journey and structure stages.
• The segment ownership document, held by the client.

What breaks when the stage is skipped

The site averages. Navigation, tone, and journey land between the audiences and serve neither, and the growth audience, which had the least voice in the research, is served worst.

The research recruits the known. Only the audiences already in the database get interviewed, and the site is redesigned for the customers it already has.

The persona becomes the unit of design. Sub-segments with different first steps get one journey, and the journey stage inherits an average.

Segments die in the readout. Three years later the research is redone from scratch.

Where this sits in the sequence

Stage two depends on stage one: the ranking has to be consistent with the two numbers the business is trying to move. It feeds stage three directly, because journeys are mapped per sub-segment, and stage four, because card sorting runs per persona. It also seeds the personalization layer: the anonymous-versus-known map is the first constraint on what can adapt for whom.

Questions we get asked about persona research

How is persona research different from audience segmentation?

Audience segmentation divides the market into groups with shared characteristics, usually for targeting. Persona research goes inside those groups to understand intent, vocabulary, hesitation, and the first step each needs, which is what a website has to design for. Segmentation tells you who; persona research tells you what to show them first.

How do you handle a primary and secondary audience on one website?

Rank them. Decide which the primary structure serves, give the other a deliberate path rather than a fork, and let the ranking drive the research recruiting. Then design the two entry experiences separately on top of a shared structure. Averaging the two is the failure mode.

How many user personas does a website redesign need?

Fewer personas than you expect and more sub-segments than you have. Three to four personas anchored on the journeys the business already names, each broken into the sub-segments whose first content differs. The sub-segments are the design unit.

How do you research an audience that isn't in your database?

Go outside it. Recruit through a panel screened for adjacent interests rather than for your product, through partner locations where prospects show up in person, or through a research vendor for the hardest segment. The audience the business knows least is usually the one the redesign exists to win.

Read next in the series

The full sequence is set out in the pillar guide, Digital Brand Strategy: Eight Stages From Baseline To Proof. Each stage has its own article.

• Stage 01, the target. What are we trying to move, and where does it stand today?
• Stage 02, the audiences. How do we serve a second audience without breaking what works for the first? You are reading it.
• Stage 03, the journey. Where does someone fall out between interest and a conversation?
• Stage 04, the structure. What do we call things, and what varies across every local site?
• Stage 05, being found. Strong nationally, weak on near me. How do we fix that specifically?
• Stage 06, the conversion. How does a visit become a qualified lead at the right location?
• Stage 07, the system. How do we get a design system your own team keeps building with?
• Stage 08, the proof. How will we know, with data, that any of this worked?
• The personalization layer. What should adapt, for whom, based on which signal?