A brand runs a quiz for eight months, collects 40,000 responses, and sends the same Tuesday campaign to all of them. The answers are sitting in Klaviyo. Nobody has written a single segment against them.
This happens because the setup work and the payoff work look like the same job and they are not. Connecting the quiz to your email platform takes an afternoon. Deciding which combinations of answers deserve a different message takes a conversation about your catalogue, and that conversation keeps getting postponed.
What follows is the second half of that job. What a quiz answer has to look like before you can segment on it, the three kinds of segment worth building, and worked definitions for eight categories.
There are two ways quiz data lands in your email platform. It can arrive as a text blob, something like a note on the profile that reads “dry skin, fragrance-free, budget conscious.” Or it can arrive as named properties with controlled values: skin_type set to dry, fragrance set to none, budget_tier set to under_50.
The first version is readable. It is not queryable. You cannot build a segment that says “everyone whose budget_tier is under_50” if the budget answer is buried inside a sentence.
Sensez syncs quiz answers to Klaviyo, Mailchimp, and 30 or so other platforms as attributes rather than free text, which is what makes everything below possible. The mechanics of getting that connection running are covered in how to send quiz data to Klaviyo, so this piece assumes it is already done.
Two things to check before you start writing segments:
The payoff for getting this right is measurable and it has been measured. Mailchimp compared about 11,000 segmented campaigns against the same senders’ unsegmented ones, across roughly 9 million recipients. Campaigns segmented on a database merge field got 54.79% more clicks than unsegmented sends from the same accounts, and campaigns segmented on declared interest groups got 74.53% more. Quiz answers are both of those things at once: declared interests, stored as structured fields.
Connecting the quiz to your email platform takes an afternoon. Deciding which answer combinations deserve a different message takes a conversation about your catalogue, and that conversation keeps getting postponed.
Most segment libraries sprawl because nobody decided what a segment is for. Three categories cover almost everything a mid-market brand needs.
Stated need comes from a single direct answer. The customer told you they want help sleeping, or they drink pour-over, or their dog itches. One field, one segment, no interpretation. These are the easiest to build and the easiest to write copy for, because you are repeating something the customer said about themselves.
Inferred state comes from a combination. A supplement customer who selects “energy” as a goal and “nothing currently” as their regimen is a different person from one who selects “energy” and lists four products. Same stated goal, opposite message. Combination segments are where quiz data starts outperforming purchase history, because purchase history cannot tell you what someone is already taking from a competitor.
Lifecycle position comes from answers plus time. Usage rate, pack size, and quiz date together tell you roughly where a customer is in their consumption cycle. That is the foundation for replenishment work, which is its own article.
Each of these uses fields a quiz can realistically collect in seven to ten questions. The logic assumes conditional branching, which is how Sensez quizzes route: later questions appear based on earlier answers rather than every customer walking the same path.
Fields: primary_goal, current_regimen, format_preference, purchase_reason
The segment that matters most here is first stack versus expanding stack. Someone selecting a goal with an empty current regimen needs education and a starter recommendation. Someone selecting the same goal with three products already in rotation needs to know what fits alongside what they take, and responds to different copy entirely.
Keep the language on goals and preferences. Segment names live in your ESP where nobody outside the team sees them, but they leak into copy more often than you would think, and supplement claim language is not somewhere to be casual.
Fields: life_stage, protein_sensitivity, pet_count, current_food_type
Life stage wording should follow the same conventions as your packaging, which means AAFCO nutrient profile language rather than invented tiers. If your bags say “adult maintenance,” your segments and your emails say it too.
Fields: brew_method, roast_preference, weekly_volume, grind_preference
The weekly_volume field is the one that makes replenishment timing possible later, so collect it even if you have no immediate use. Whole-bean buyers are an equipment and accessory cross-sell audience; an espresso plus light-roast pairing is a small, opinionated, high-LTV segment worth its own content.
Fields: primary_goal, format_preference, experience_level, time_of_use
New customers need an onboarding sequence with conservative framing and no advanced product. A topical buyer whose goal is recovery is an audience whose interests barely overlap with the tincture buyers. Compliance is the whole game here: segments are internal and can be blunt, but everything the segment triggers goes through the same review as any other CBD copy.
Fields: sweetness_tolerance, price_ceiling, occasion, varietal_familiarity
The price_ceiling field is the one most wine brands skip and the one that prevents the most unsubscribes. Recommending a $70 bottle to a $25 buyer reads as not listening. Gifting is an entirely separate calendar, driven by dates rather than consumption, and the exploring segment responds to discovery packs and mixed cases.
Fields: experience_level, solo_or_partnered, discretion_requirement, category_interest
A high discretion_requirement is a suppression rule as much as a segment. It should govern subject lines, sender name, image selection, and anything referencing packaging. It is the one segment on this list where getting it wrong does real harm to a customer rather than just annoying them.
Fields: activity_type, season_of_use, skill_level, trip_length
Activity type plus season of use drives a calendar rather than a cadence. A backcountry ski segment goes quiet in June and that is correct. The beginner day-tripper is the largest segment in most catalogues and the one most often ignored in favour of the expedition buyer.
Fields: room, existing_style, project_stage, budget_tier
A researching customer needs a long nurture with no discount pressure; a ready-to-buy customer is the segment worth spending on. Room plus existing style is the combination that makes a recommendation feel like it came from a person.
Every field above describes a customer at one moment. A puppy becomes an adult dog. A beginner becomes intermediate. A researching customer either buys or gives up.
This is not an argument against using the data. It is an argument for knowing how old it is and having a way to refresh it, which is what the check-in quiz covers. The practical rule: any segment built on a field that changes with time needs a date stamp alongside it, and a threshold past which you stop trusting it.
The failure mode is not too few segments. It is forty overlapping ones that nobody can remember, where a customer sits in nine of them and receives four emails on the same Tuesday.
A useful starting position for a mid-market brand is six to ten active segments, each with a clear owner in the campaign calendar and a stated reason to exist. If you cannot name the campaign a segment feeds, it is a saved search, not a segment.
Two housekeeping rules worth adopting. Every segment gets a written definition in plain language somewhere your team can read it, because six months from now nobody will remember why segment_11 excludes gummies. And decide explicitly what happens to people who match nothing, because most brands discover a quiet third of their list that falls through every rule and receives only the generic sends.
If you cannot name the campaign a segment feeds, it is a saved search, not a segment.
Pick one segment. Preferably the exclusion one, because those produce the fastest visible improvement and carry the least risk: the chicken-sensitive dog owners who stop receiving chicken formula emails, the capsule people who stop getting gummy launches.
Sensez auto-tags quiz respondents by their answers and pushes those attributes into your email platform, so the fields are likely already sitting there waiting. If you would rather have the quiz and the segment structure designed together rather than retrofitted, our Concierge team builds both.
