Two Customers in a Trench Coat
The Hidden Buying Power of Family Units
The Kebab Shop Problem
I was sitting in a kebab shop recently, thinking about something that the kebab shop itself probably does not care about.
Customer data.
Not in the creepy Silicon Valley sense. Not “let’s build a machine learning model to predict whether someone wants extra garlic sauce.” More basic than that. The kind of data a small food business already gets without even trying: a name on the order, a card payment, maybe the last four digits of the card, maybe a rough timestamp, maybe the items ordered.
That alone is enough to start asking interesting questions.
How many people come back? How often do they return? How much do they spend? Are they buying the same thing every time? Are the regulars keeping the business alive? Is the shop growing because of loyal customers, or is it surviving on random walk-ins?
For a kebab shop, repeat customers are probably one of the clearest signs that the product is good. If people keep coming back for the same kebab, same snack pack, same late-night feed, something is working. If nobody returns, that tells a very different story.
But then I had a second thought.
What happens when the person paying is not really “the customer”?
Say my girlfriend walks into the shop and buys kebabs for both of us using my card. The shop might see her name attached to the order, but the card details are mine. Or maybe I order, but she is the one who decided we were eating kebabs in the first place. Or maybe I pay, she chooses, and we both consume the product.
So who is the customer?
Me, because it was my card?
Her, because her name was on the order?
Both of us, because we made the decision together?
Or are we not two customers at all, but one tiny household unit showing up in the data as fragmented individual transactions?
The Problem With Individual Customer Tracking
This is where a simple kebab-shop thought becomes a much bigger marketing question: are businesses actually targeting the right person, or are they confusing the payer with the buyer, the buyer with the decision-maker, and the decision-maker with the user?
In marketing theory, these roles are often separated. The initiator is the person who first suggests the purchase. The influencer shapes the decision. The decider makes the final call. The buyer completes the transaction. The user consumes the product. Sometimes those are all the same person. In a family, couple, workplace, friendship group, or household, they are often different people.
That distinction matters because most customer data systems are built around individuals.
One email equals one customer. One phone number equals one customer. One card equals one customer. One account equals one customer.
But real life does not work like that.
Real life is messier. A family might share cards. A couple might alternate who pays. A parent might buy something because a child asked for it. A husband might pay for something his wife chose. A wife might pay for something her husband uses. A group of friends might have one person who always orders, even though the decision belongs to the group.
The transaction says one thing.
The household reality says another.
This is the hidden buying power of family units.
A business looking only at individual transactions might see three average customers. But underneath that, there might actually be one very valuable household. The husband buys lunch there on weekdays. The wife orders dinner there once a fortnight. The kids ask for it on weekends. The data may show three separate customers with average value, when the reality is one family unit with high lifetime value.
That creates a problem for marketing.
If the business only targets the person who paid, it might miss the person who influenced the purchase. If it only rewards the person whose card was used, it might undervalue the family unit. If it splits the value across too many individual profiles, it might underestimate the true loyalty of the household. But if it merges people too aggressively, it can become inaccurate, invasive, and just plain weird.
This is why the kebab shop example is funny, but also useful.
A kebab shop probably will not build a customer identity graph. It probably does not need to. But the same problem shows up everywhere once the product becomes more expensive, more emotional, or more tied to household decisions.
Cars. Holidays. Furniture. Insurance. Kids’ toys. Groceries. Restaurants. Streaming services. Education. Health products. Home improvement. Financial services.
In all of these categories, the person who pays is not always the person who decides.
Who Actually Makes Purchasing Decisions?
The gender dynamic makes this even more interesting, although it needs to be handled carefully. Historically, in many heterosexual households, men were more likely to be positioned as the primary income earner, while women often managed more of the household purchasing, planning, care work, and day-to-day family decision-making. That pattern is changing, but it has not disappeared.
Pew Research found that in the United States, 55% of marriages still have a husband as the primary or sole breadwinner, while 29% of marriages have both spouses earning roughly the same amount and 16% have a wife as the breadwinner. Fifty years earlier, husbands were the breadwinner in 85% of marriages, so the pattern has shifted dramatically, but the old model still exists in many households.
At the same time, women’s influence over purchasing is enormous. NielsenIQ reports that women control an estimated $31.8 trillion of worldwide spending, are projected to control 75% of discretionary spending in the next five years, and have 70–80% influence on all consumer spending. BCG similarly describes women as managing around $32 trillion in global spending and argues that many companies still fail to design products and services around women’s actual needs.
So the old lazy assumption of “market to the man because he pays” is not just outdated. It is commercially dangerous.
But the opposite assumption is also too simple. You cannot just say “market to women because women decide.” That still flattens the household into a stereotype. The smarter approach is to ask: what role does each person play in the purchase?
For a car, one person may care about price and finance. Another may care about safety, comfort, and practicality. The kids may influence the choice more than anyone wants to admit. For toys, the child may be the initiator and user, while the parent is the buyer, payer, and gatekeeper. For food, the decision might be driven by convenience, cravings, routine, budget, or who complained loudest that night.
Children are also not passive in household buying. Research reported by Marketing Charts, based on an NRF report, found that 87% of surveyed parents said their children influence purchase decisions. Almost half said children influence purchases specifically for the child, while more than one-third said children influence purchases for the whole household.
The Family Unit as an Economic Entity
That means “family purchasing power” is not just about couples. It is about households as decision-making ecosystems.
And this is where customer lifetime value becomes more complicated.
Normally, customer lifetime value is attached to an individual. How much does this customer spend over time? How often do they buy? How likely are they to return?
But in household-driven categories, individual CLV can lie.
Imagine a kebab shop has four customer records:
David: buys once a month.
Sarah: buys once a month.
Sarah using David’s card: appears as either Sarah or David depending on the system.
A family order: appears as whoever placed the order.
A basic report might say, “These are average customers.”
A household-level view might say, “This couple or household buys from us every week.”
That changes how valuable they are. It changes how you market to them. It changes whether you offer them a loyalty reward. It changes whether you should target them with individual meals, couple deals, family bundles, or catering offers.
The unit of value is not always the person.
Sometimes the unit of value is the household.
Two Customers in a Trench Coat
This is obvious in B2B. Companies do not usually treat every employee as a totally separate buyer. They understand the “account.” They know that a business customer has multiple stakeholders: the user, the buyer, the approver, the finance person, the executive sponsor.
But in consumer marketing, we often forget the same thing happens inside homes.
A household is basically a tiny buying committee.
The issue is that tracking this is hard.
The most basic idea is to use payment information. But even that has problems. A merchant should not be treating a visible card number as a customer identity system. PCI guidance recognises masking and truncation standards, with the first six and last four digits commonly accepted as the maximum display format across payment brands.
The last four digits alone are also a weak identifier. They can collide across different cards. Cards expire. Cards get replaced. People use Apple Pay, Google Pay, different accounts, business cards, partner cards, and shared cards. Payment tokenisation can help because it replaces sensitive card numbers with a unique token, but even then, a token usually tells you about a payment instrument, not the full human or household behind it.
So if a business wants to understand family-unit value, it needs to be careful.
There are a few possible models.
The first model is to keep everything individual. Whoever pays gets the value. This is simple, clean, and easy to report on. But it underestimates household behaviour. It might split one loyal family across multiple “average” customers.
The second model is to split value across known participants. If two people are known to be part of an order, each gets partial attribution. This can be more accurate, but it creates practical problems. What if I buy kebabs with my girlfriend today, alone tomorrow, and with a friend next week? What if the person I was with changes over time? What if the system keeps assigning influence to someone who is no longer part of my life?
The third model is to create a household or family unit above the individual customer profiles. This is probably the most useful for serious businesses. You still keep individuals separate, but you also recognise that some purchases belong to a shared unit. The household gets its own lifetime value, while each person keeps their own preferences, behaviour, and marketing profile.
That gives you three layers:
The individual.
The transaction.
The household.
This is much closer to reality.
But it only works if the data is collected ethically and transparently. In Australia, the Australian Privacy Principles govern the collection, use, and disclosure of personal information for organisations covered by the Privacy Act. The OAIC defines personal information broadly as information or an opinion about an identified or reasonably identifiable individual, and even inferred tastes and preferences from credit-card purchases or web browsing can be personal information.
In other words, “we are just doing analytics” is not a free pass.
The non-creepy version is not secretly guessing who lives with whom based on cards and order patterns. The non-creepy version is building systems where people willingly identify themselves because there is a clear benefit.
A restaurant might offer a family loyalty account.
A grocery app might let households share lists, points, and offers.
A streaming service might have profiles under one account.
A car brand might design messaging for different decision roles rather than assuming the man is the buyer and the woman is a passenger.
A toy brand might market fun to the child, trust to the parent, and value to the household.
A finance brand might stop only speaking to the “main account holder” and instead think about the couple or family making the decision together.
The strategic question becomes: who needs to believe in this purchase for it to happen?
Not just who pays.
Who initiates it?
Who researches it?
Who vetoes it?
Who uses it?
Who complains if it is wrong?
Who has to be emotionally comfortable before the purchase goes ahead?
That last one is underrated. In family purchases, the veto vote can matter more than the payment method. One person may technically be able to buy the thing, but if their partner hates the idea, the purchase dies. One person may not pay a cent, but their approval is essential.
That is why marketing to households requires more than demographic targeting. It requires role targeting.
A person is not valuable only because they spend money. They may be valuable because they unlock someone else’s spending. They may be the recommender. The organiser. The planner. The gatekeeper. The emotional decider. The person who turns “maybe” into “yes.”
Why This Matters More Than You Think
This is where businesses can make better decisions.
For low-cost, frequent purchases like kebabs, coffee, takeaway, or groceries, household tracking can reveal patterns around routine and occasion. A single person buying one kebab is different from a couple buying dinner every Friday. A family buying every Sunday night is different again. The marketing should reflect the occasion: solo convenience, couple treat, family meal, late-night craving, work lunch, post-gym meal.
For high-cost purchases like cars, appliances, renovations, or holidays, the business should assume multiple decision-makers until proven otherwise. The campaign should not only speak to the person with the card or the finance application. It should speak to the practical user, the emotional influencer, the safety-conscious parent, the budget keeper, and the person who needs to feel heard.
For child-related purchases, the user and buyer are usually different. That means the product needs two sales pitches at once. The child needs to want it. The parent needs to trust it. The household needs to justify it.
For subscriptions and recurring products, household identity can be the difference between misunderstanding churn and understanding value. One account might have multiple users. One cancellation might not mean one disappointed customer; it might mean a whole household has moved on.
The danger is that most reporting systems simplify all of this into clean but misleading numbers.
Customer A spent $400.
Customer B spent $180.
Customer C spent $90.
But what if A, B, and C are the same household?
What if B influenced most of A’s spending?
What if C is the child who determines where the family eats?
What if the person with the lowest direct spend is actually the most important person in the buying decision?
Data Is Fun. Reality Is Messy.
This is why data is fun.
It starts with a kebab shop and ends with the realisation that “customer” is not as obvious as it sounds.
A customer can be a person.
A customer can be a card.
A customer can be an account.
A customer can be a household.
A customer can be an invisible decision-making unit made up of people with different personalities, different motivations, different levels of influence, and different relationships to the final purchase.
The businesses that understand this will market better.
They will stop treating every payment as a single-person event. They will stop assuming the payer is the decision-maker. They will stop undervaluing the household member who influences but does not transact. They will build loyalty systems around real buying behaviour, not just clean database rows.
And maybe the kebab shop still will not care.
Fair enough.
But once you move beyond kebabs and into categories where the purchase is bigger, more emotional, more expensive, more shared, or more family-driven, this becomes extremely important.
Because sometimes your best customer is not one customer.
Sometimes it is two customers in a trench coat.
Sometimes it is a couple.
Sometimes it is a family.
Sometimes it is a household.
And if your data cannot see that, your marketing probably cannot either.