A planning deadline and no way to describe the audience.
The company knew its existing buyers. It had no way to describe the ones it needed next.
No segmentation, and a database too small to build one
A spirits company was rebuilding brand strategy across a multi-brand portfolio ahead of the next planning year. Leadership had a fixed presentation date and had to decide where marketing investment would go across the portfolio.
Underneath that decision was a gap. There was no formal audience segmentation, and the customer data the company held was too small to represent the wider category. Anything built on it would have described the buyers it already had.
Two brands from two different distilleries were the focus, both in the top investment tier. One had won distribution but had not converted it into repeat buying.
Two halves, each correcting the other.
Personas describe who. Need states describe when.
Segmentation work usually fails in one of two ways. It produces personas that feel true but cannot be found in any media plan, or audiences that can be bought but say nothing about why anyone buys.
The way out was not to pick one. The same person shops completely differently depending on whether they are replacing their everyday bottle, buying a gift, or hunting something rare. A persona alone cannot tell you which message to serve. A need state alone cannot tell you how to say it.
A need state you cannot size is an observation. A need state you can size is a line in a plan.
Selection set the ceiling on everything after it.
How the study ran
Screening
An in-depth screener ran on the ethnography platform and drew close to 300 respondents. It did two jobs: it qualified participants, and it produced a quantitative dataset of its own, which we analyzed and reported alongside the field work rather than discarding once recruiting closed.
Selection
This step set the ceiling on everything after it. Every submission was reviewed, including video responses, against several criteria at once rather than a single quota. The final 24 were approved with the client before the study opened.
Five paced missions
Each had to be completed before the next unlocked, and each was written against the learning agenda rather than against curiosity.
- Visit their local liquor store and document it
- Shop online across three different sites
- Show us their liquor cabinet
- Talk about ordering a drink out versus drinking at home
- Make a drink for us
The online shopping mission is worth calling out. It is a usability study running inside a brand study, and it produced the digital experience findings that later became the website recommendations.
Live probing
The study was monitored in near real time. When someone submitted something interesting we came back with follow-up questions while they were still in that frame of mind, rather than reading it cold two weeks later. That is the difference between a diary study and a survey with photographs attached.
Quantitative validation and sizing
Need states were rebuilt as custom audiences using filters chosen to approximate each group's habits, motivations, and psychographic profile, then analyzed for media engagement, motivators, and the effect the pandemic was having on drinking occasions. Each was sized against the adult online population so the audiences could be ranked against each other.
Category check
Syndicated category research ran alongside as a test of whether what we saw in 24 homes matched what the category was doing at scale.
The shape of the study is the argument. Depth first, so the audience definitions came from watching people rather than from demographic cuts the company already believed in. Scale second, so the strategy could be costed instead of admired.
Out of behavior, not a demographic grid.
Three ways of being a bourbon drinker
These came out of what people did in their own homes and stores, and each one buys for a different reason.
The Purist
Drinks within one category and understands flavor complexity. Hunting what is rare and hard to find. Recognition for his expertise matters as much as the bottle does.
The In-Betweener
Wants what is new and on trend. Alcohol runs through his social life, his buying moves with the season and his mood, and he will sip something premium and mix with something cheaper on the same night.
The Experimenter
Discovers through blogs, influencers, reviews, and bartenders. Buys to share, cares about packaging, and needs price comparison before committing to something unfamiliar.
Six situations, sized and ranked.
Six reasons to reach for a bottle
Each was validated in the qualitative work, rebuilt as a targetable audience, and sized. They differed by orders of magnitude in reach and, more usefully, in how predictable the behavior was. The three largest accounted for about two thirds of the total opportunity, which is what let the client focus a planning conversation that could easily have tried to cover all six.
Replace or replenish
Their go-to bottle, part of a routine. Triggered by running low. The most predictable behavior of the six.
Special occasion
Hosting, and wanting every base covered. Motivated by trust, universal appeal, and not embarrassing themselves in front of guests.
Collecting
Not any bottle will do. Driven by scarcity, hype, and the thrill of acquiring something other collectors would recognize.
Treating yourself
A milestone, a promotion, or just a hard week. They want a reason why this bottle is worth being the reward.
Gift giving
Thoughtful matters, standing out matters more. The tension is between a safe gift everyone enjoys and one with a bit of mystique.
Exploring
Time to try something new. Willing to take a risk, but needs a safety net: a trusted recommendation or a reason to believe.
Both axes are ordinal. No figures are shown, and none can be read off this chart. The useful part is where the pattern breaks. Collecting is one of the least predictable occasions and still one of the largest, which makes it worth planning for rather than reacting to. Gift giving is highly predictable and comparatively small, so it earns a seasonal push and not a year-round budget line.
The same act, for three completely different reasons
Persona × need stateCrossing the two is where the value was. Three people standing in the same aisle, replacing the same everyday bottle, need three different messages, three different channels, and three different pieces of creative.
"When I am at a store to replenish my regular brand, I want to be thrilled by finding something rare, so I enjoy the big find."
- What he does
- Works several local stores looking for allocated bottles, and researches online while standing in the aisle.
- How you reach him
- Geotargeted media carrying anticipated releases and inventory news.
"When I am replenishing, I want to be inspired by something that matches my mood, so I see myself in the brand."
- What he does
- Goes to his usual store for his usual bottles, then walks the aisle to see what catches his eye.
- How you reach him
- On-pack and shelf-level prompts that open into the story behind the process and the people.
"When I am replenishing, I want to be intrigued by different tasting notes and packaging, so I find something new to experiment with."
- What she does
- Asks a friend or a store associate, and looks up other brands while she is still in the store.
- How you reach her
- Review and tasting-note placements where she already goes to check before buying.
This is what turned a set of audience portraits into a targeting brief. Personas describe who someone is. Need states describe the situation they are in. Laid across each other, every cell has a person, a moment, a motivation, and a media decision attached to it.
Eleven patterns cut across everything.
What held true across all 24
Eleven behavioral themes ran through the whole study. The ones with the most commercial consequence were these.
A usability study hiding inside a brand study.
Participants shopped the sites and narrated as they went
The online mission put people on the two focus brands' sites alongside competitor and delivery sites in the same session, talking through what they were doing.
They could not find purchase paths. They could not locate reviews and left the site to search for them elsewhere. They hit missing price information and loading problems, and in one case could not tell whose site they were on. Set against the competitor experiences they shopped minutes earlier, the gaps were specific and fixable rather than a matter of taste.
Delivered into brand planning.
Three personas, six audiences, and a targeting framework
The work delivered three personas grounded in observed behavior, six sized and targetable audiences, a persona-by-need-state targeting framework, eleven behavioral themes with brand opportunities attached to each, and a set of website recommendations grounded in real shopping sessions rather than a heuristic review.
It went into the client's leadership brand planning presentation and fed the agency's media, CRM, and communications planning for the following year.
Client, brand, competitor, and participant details are generalized. Audience sizing figures come from a licensed third-party platform and are described in relative terms only. Participant quotes are paraphrased and carry no identifying detail.
Depth and scale, not one or the other.
Twenty-four people cannot tell you how big anything is
The ethnography alone was not the valuable part. The value was in refusing to choose between depth and scale.
The qualitative work explained why people buy. The quantitative work said how many of them there are and what it would cost to reach them. Neither half would have survived a leadership meeting on its own.
That is still how I work. Go and watch what people actually do, then find out how many of them there are.