IGNITE EXTENDED / LESSON 07
Which customer group fits the value you can actually deliver?
A segment-choice memo with an attribute matrix.
Customers with similar demographics can buy the same category for different reasons. The Calyx discussion distinguishes flower-buying situations and asks what matters in each. A personal decoration purchase, a gift and an urgent occasion can imply different priorities.
A benefit segment groups situations in which a similar set of attributes drives choice. Demographics may help locate the people after the segment is understood. Starting with demographics alone can conceal the reason they choose one alternative over another.
The archived marketing discussion converts catalog distribution into the cost of acquiring a new customer. A low purchase response means the cost of reaching many prospects must be recovered from the few who order. Revenue growth can therefore make losses larger.
The earlier teaching example uses $40 of first-order contribution and $74 of acquisition expense, leaving negative $34 on the first order. These are historical discussion values, not a current benchmark. Repeat purchasing can change lifetime economics, but it must be measured rather than assumed.
Demographics describe people. Benefit segmentation explains differences in what they seek in a buying situation. The same person can seek different benefits on different occasions.
Groups people by characteristics such as age or location. This can help reach people but may explain preferences poorly.
Groups buying situations by what people want to accomplish and which attributes they value.
Flower illustration: two people of the same age and city can choose differently because one needs an immediate gift and the other wants a long-lasting home display. Age may help locate buyers, but the buying situation better explains the relevant offer and alternative.
Use demographics where they help reach the target. Use evidence about desired benefits to explain why that target would choose the offer.
Segmentation connects a buying situation to desired benefits and to the company’s ability to deliver them. A demographic label alone leaves this logic incomplete.
Identify the occasion that triggers a choice. Ordering flowers for home decoration differs from buying an urgent apology gift, even when the buyer is the same person.
Discover the relative priorities in that situation. A home-display buyer may value longevity; an urgent gift buyer may prioritize immediate availability. Treat these priorities as hypotheses to investigate.
Compare what people actually choose and the benefits those alternatives deliver. A local florist offers a different combination of access, advice and presentation from a delivery-based offer.
Select a reachable situation where the offer has a useful fit and plausible economics. Explain who will be excluded and what evidence would cause the team to revise the choice.
The Calyx examples below apply this logic. The lesson is how benefits, alternatives and acquisition economics constrain the target, rather than a universal rule about flower buyers.
A full, open arrangement as pictured in the catalog beside a customer's photo of what arrived: tightly closed buds in plastic wrap on an office desk.
Calyx ships flowers in bud so they last longer, which means the product looks unimpressive on arrival. That is a strength for someone decorating their own home and a liability for a gift whose impact depends on the moment it is opened.
Mailing 100 catalogs to prospects costs ($0.68 + $0.06) x 100 = $74 and yields one customer with an $80 order at 50% margin, so $40 contribution and a $34 loss. Mailing 6 catalogs a year to a current customer costs $4.08 and, at a 4.5% response per catalog, yields $21.60 revenue, $10.80 contribution and $6.72 profit.
The $74 is the customer acquisition cost. At $6.72 a year from a retained customer, Calyx needs about five years of continued buying just to recover the $34 first-order loss, before discounting or churn. Repeat purchasing does not rescue broad prospecting at these rates; better targeting has to.
Situations from class (Mother's Day, a small-token gift, a gift for a spouse, an apology, support for someone in hospital, brightening one's own home) beside priorities (impact and presentation, price and spontaneity, reliability, longevity, variety, picking flowers yourself).
The same person appears in several situations. Priorities attach to the occasion, which is why age or income would predict choice poorly here.
Importance ratings for impulse, planned and home-decor purchases: longevity (low, low, high), variety (low, med, high), presentation (high, high, low), spontaneity (high, low, low) and remote delivery (low, high, high).
Home decor is the only segment that values longevity and variety highly and does not care much about presentation or spontaneity. The segments are defined by what matters, not by who the buyers are.
Calyx rates high on longevity, variety and remote delivery but low on presentation and spontaneity; grocery is high only on spontaneity; the local florist is high on variety, presentation and spontaneity; 800-Flowers is high only on remote delivery.
The local florist is the strongest all-round competitor, but it cannot deliver longevity or remote delivery as well. Calyx's distinctive combination is long life plus delivery, and only one segment wants exactly that.
Place valued attributes in rows. Represent segments by how much each attribute matters and competitors by how well they provide it. This makes the logic of a target choice visible: the company should offer something important to the segment that alternatives do not provide as well.
The matrix is a reasoning aid, not a measurement merely because it contains numbers. Scores need a common scale and credible evidence. A useful gap may disappear when customers explain that two attributes trade off or when a competitor can close it cheaply.
A company with limited resources cannot optimize the product, message and channel for every buying situation. Choosing a target concentrates work on a coherent set of needs. It also identifies customers the company will serve less well.
Evaluate fit, reachable potential and defensibility together. A large segment may have strong incumbents and expensive access. A smaller segment can be useful if it offers a workable path to adoption and learning. Revisit the target when evidence contradicts the attribute priorities that justified the choice.
Historical Calyx slide: $40 contribution − $74 prospect mailing cost = −$34 before other costs.
Choosing a target directs product, communication and channel decisions. It also means accepting that some customers will be better served by another offer.
Specifies a buying situation and the benefits the business can serve well.
Tries to satisfy incompatible priorities, making the offer and acquisition approach less clear.
Illustrative application: a delivery business prioritizing planned home displays can design around longevity and reliable replenishment. Trying simultaneously to promise an urgent bespoke gift may require a different inventory, service process and route. Averaging those priorities can make both offers less convincing.
State the excluded situation and the consequence. Focus is useful when the chosen combination can be served well and reached economically.
Brands plotted on powerful-to-gentle and prestige-to-economy axes, with circles marking segments: house proud (Omo, Fab), manual/children (Cold Power, near Drive and Ajax), value (Surf, Spree, near Dynamo and generic) and woolens (Lux, Softly).
The map is the visual form of the matrix: segments are clusters of desired benefits and brands sit where buyers perceive them. Gaps and crowding become visible, but only on the two axes chosen, so pick axes that actually drive choice.
The combined matrix with the home-decor segment and Calyx highlighted; their ratings line up on all five attributes (high, high, low, low, high).
This is what fit looks like: the segment's high-importance attributes are exactly Calyx's strengths, and Calyx's weaknesses fall on attributes that segment rates low. It also tells Calyx where its acquisition spending is wasted: catalogs reaching impulse and gift buyers.
Demographics are sometimes easy to address but weak predictors of preference; target segments should be identified from attitudes and behaviors.
Use benefits to choose the segment and demographics, or any other observable marker, only to find and reach it. Reversing the order targets people who are easy to buy media against but do not want what you offer.
Investors want to see focus, including a plan to win most of the target market; first figure out how to delight the target customer, then how to find them cost-effectively.
The source's '80 percent' is a rhetorical bar for dominance of a narrow segment, not a benchmark. The sequence is the point: fit first, then reach, which is exactly where Calyx's broad catalog mailing went wrong.
Selected reading: PDF pages 2, 3, 4
Calyx’s direct-from-grower delivery creates a particular benefit profile. The marketing decision requires matching that profile to buying situations and examining the economics of reaching those customers.
The case lists gifting, special events and decoration. Desired benefits can differ even when the person buying flowers is the same.
Fewer distribution steps shorten the interval between cutting and receipt. Longevity and freshness compete with other priorities such as immediacy and arrangement services.
Catalog response differs across existing customers, recipients and rented lists. Similar demographics do not guarantee similar buying behavior.
The company considers additional catalogs, mass media and Internet advertising while lacking the resources to pursue every approach together.
Selected reading: PDF pages 2, 3, 4
A claim of fresher flowers matters most where customers value that difference. An urgent delivery occasion can place greater weight on other attributes.
A response rate alone leaves order contribution, acquisition spending and later purchasing unresolved. Keep the denominator and time horizon visible.
The case’s channel shares, prices and mailing results describe its period. They supply a decision exercise rather than current market benchmarks.
The teaching case contrasts flowers selected for long life and variety with purchase situations where immediate availability or gift presentation matters more. A distinctive offer can still struggle when acquisition spending reaches people who prioritize different attributes.
Which buying situation offers the strongest fit, and how could it be reached?
Brand, Israeli and Ngwe found that LLM-generated survey responses sometimes match human willingness-to-pay estimates but are often inaccurate and occasionally wrong-signed. Fine-tuning on prior surveys helped within a known category, but did not improve estimates of differences between customer segments. Use synthetic respondents to draft hypotheses and interview guides, not to rate segment priorities in a matrix.
Newer work proposes calibrating a large set of LLM responses against a small set of human responses to correct the bias. Naively pooling AI and human answers, or substituting one for the other, biased estimates. For Relay that means real front-desk and owner interviews remain the anchor.
Livingstone Partners estimates roughly 37,000 US outpatient PT clinics, about 75% independently owned, with no operator holding more than 10% share and the largest under 6%. Private-equity platforms continue to acquire independents, so a clinic that buys as an independent today may be part of a group next year. Relay should record ownership changes as a segment-migration signal.
Sources: Brand, Israeli and Ngwe, Using LLMs for Market Research (HBS Working Paper 23-062) · Wang, Zhang and Zhang, Large Language Models for Market Research: A Data-augmentation Approach · Livingstone Partners, The Physical Therapy Tracker
Use the simplified Calyx packet: the offer is strong on longevity and variety, weaker on immediate pickup; an initial order produces $40 contribution while the prospect mailing costs $74 per acquired buyer under the case assumptions.
Choose a buying situation, explain the attribute fit, propose a more focused acquisition test and state why the first-order loss is not a full lifetime-value conclusion.
Relay's pilot runs at 3 clinics and Daniel is building v1 with an integration to one EHR. Maya must choose who v1 is sold to first. The candidate segments: independent single clinics; multi-location PT groups (5 to 60 locations); hospital outpatient departments. Founder-led sales cost about $2,400 per clinic; price is $399 per location per month at about 80% gross margin. Relay can only fill slots at clinics running the integrated EHR, and hospital buyers require long procurement and heavy IT and security review. There are roughly 38,000 US outpatient PT locations (illustrative).
Write a one-page segment-choice memo: a market-structure matrix (segment importance and alternative performance on 5 to 7 attributes), the chosen first target with fit, reach and economics, the segments deliberately excluded, and the evidence that would make you revisit the choice.
Calyx could buy catalog lists cheaply, but most recipients wanted presentation or spontaneity rather than long-lasting flowers, so response was 1 in 100. Choose the segment on benefits first and find an observable marker to reach it second.
At $6.72 a year from a retained Calyx customer, recovering a $34 first-order loss takes five years. Compute payback from measured repeat behavior, not from hoped-for loyalty.
Promising independents a no-IT setup and hospitals a full security and integration program pulls product, sales and onboarding in opposite directions. Name the excluded segment and the reason in the memo.
A Relay segment-choice memo with a market-structure matrix, a first target (independents and small groups on the integrated EHR), deliberate exclusions, CAC payback of 7.5 months, and written triggers for revisiting the choice.
Next lesson: Positioning & the value proposition
Check your understanding
Question 1
Calyx spends $74 to acquire a customer whose first order yields $40 of contribution. A retained customer then generates $6.72 of profit a year from catalogs. Ignoring discounting and churn, about how many years of retention does Calyx need to recover the first-order loss?
First-order loss = $40 - $74 = -$34. Years to recover = 34 / 6.72 = 5.06. 'About 11 years' wrongly divides the full $74 by $6.72, ignoring the $40 the first order already contributed.
Question 2
Relay's founder-led sales cost $2,400 per clinic. Each clinic pays $399 a month at 80% gross margin. What is the CAC payback period?
Monthly gross profit = 399 x 0.80 = $319.20; payback = 2,400 / 319.20 = 7.5 months. 6.0 months uses revenue rather than gross profit; 4.4 is the LTV/CAC ratio, not a payback period.
Question 3
A Relay advisor suggests targeting 'clinics in metro areas with more than 250,000 people' because the list is easy to buy. What is the main weakness of this segment definition?
Geography, like demographics, helps with reach but is a weak predictor of preference. A useful segment groups clinics by the benefits they seek and the situations they face, such as owner-run clinics feeling every empty slot.
Question 4
In the Calyx matrix, home-decor buyers rate longevity, variety and remote delivery high and presentation and spontaneity low. Calyx rates the same way. The local florist rates high on variety, presentation and spontaneity. What follows?
Fit means the segment's important attributes are the firm's strengths and its weaknesses fall on unimportant ones. Matching the florist on presentation would raise cost without serving the segment Calyx fits best.
Question 5
Why should Relay deliberately exclude hospital outpatient departments from its first target, even though each hospital system could buy many locations?
Target choice weighs fit, reach and economics together. A segment can be large and valuable yet a poor first target if the cost and time of reaching it outrun the startup's resources and delay learning.
Question 6
A founder argues: 'Independent clinics are 75% of the market, so they must be our best target.' What is wrong with this reasoning?
A large segment can have poor fit, expensive access or high churn. The choice needs evidence on fit, reach, defensibility and economics. Relay later finds that groups retain far better despite being fewer.
Workbook
Module output: An evidence log, problem frame, prototype, segment choice, positioning and route to market.
Sources
Original pages and full documents are in the Library edition.