How Digests, Sets, Living Cards and digital Albums can turn disposable messages into persistent customer memory
There is a peculiar silence in most customer relationships. It begins just after the transaction. The customer has bought, booked, invested, renewed or redeemed. The brand has recorded the event, credited the points and moved on to the next campaign. Then, until there is another reason to sell, very little happens.
Loyalty programmes were designed to make that transaction more valuable. They work extremely well when purchase is frequent and the reward is economic: airlines, grocery, fuel, hospitality. But for most customers of most brands, the time between purchases is far longer than the purchase itself. A points balance can sit for weeks or months without creating a reason to think, learn, open or return.
That gap suggests a different job. What if a brand could reward attention as deliberately as it rewards purchase? What if every useful interaction left the customer holding something — not another message consumed and forgotten, but a Card owned, a Set advanced, knowledge remembered, a Circle joined?
Points help a brand reward the transaction. Cards may help it own the time between transactions.
The mechanism is not a replacement for loyalty. It is a second rail. A recurring Digest earns periodic attention. A Set of Cards creates persistent memory between those moments. Over time, a shared, person-owned Album can aggregate many issuers into a habit no single brand has to sustain alone. The first two promises must stand on their own. The third is the compounding upside.
1
The Silence Between Purchases
Points-led loyalty is strongest when money changes hands.
Earn five points per rupee. Fly again and move up a tier. Spend a little more and unlock a voucher. The logic is clear because the behaviour is clear: reward the action that creates revenue.
Modern loyalty is more than points. The best programmes add experiential rewards, referrals, personalised offers and some engagement that needs no purchase, and the category is still growing. The argument here is not against loyalty programmes. It is against the limits of the points mechanic inside them. Even a good programme usually stores value as an abstract economic entitlement. Cards store attention as persistent objects and memory.
The trouble begins when the customer is not buying. A mutual-fund investor may hold for months. A car owner may not replace the vehicle for years. A jewellery customer may return around a wedding or a festival. Home improvement, insurance and travel all contain long stretches in which the customer is valuable but out of market. The default CRM rhythm is promotion, silence, promotion.
Points do not solve that silence well, because the balance is abstract, the reward is usually distant, and the programme rarely changes between transactions. A customer can hold 1,840 points and have no reason to think about the brand this morning. In Deloitte’s 2025 survey of 5,564 US loyalty-programme members, 40% said they sometimes forget to redeem rewards. Antavo’s 2026 loyalty report puts a number on the residue: about a quarter of the points US consumers earn are never spent, and roughly one in eight expire outright. That is not evidence that points fail. It is evidence that earning and remembering are different jobs. Unused balances can be one signal that the relationship has gone quiet.
Figure 1. A points programme speaks at each purchase. A Digest with a Set speaks twice a week, and what it leaves behind accumulates.
The accounting reinforces the distinction. Where loyalty points create a material right, IFRS 15 commonly treats that right as a separate performance obligation, with part of the transaction price deferred until the right is exercised or otherwise resolved. Breakage — rights customers never exercise — is an explicit part of the model. The treatment varies by programme structure, so “points are a liability” is too broad a claim. But points frequently carry an economic obligation that Cards need not.
I would change one position I have argued before. In my 2025 essay on Mu, I treated breakage as something an issuer could design for: expiry, friction, aspirational rewards and the economics of points left unused. That is not the right objective for an attention system. Breakage may improve an issuer’s economics, but it also means value promised to the customer was never experienced. For a system whose goal is repeated attention, unused value is a warning, not a victory. This essay supersedes that recommendation.
The objective is also not “daily brand recall”. No single brand needs, or deserves, to occupy the customer’s mind every morning. The better concept is mental availability: being easily thought of in relevant buying situations, as the Ehrenberg-Bass Institute describes it. The job of a between-purchase programme is to keep useful memory structures alive, so that the brand has a better chance of coming to mind when the category becomes relevant again.
2
From Points to Objects
A Card is visible. A Set is incomplete. An Album remembers.
The difference between a loyalty balance and a collection is not merely visual. It changes the unit of value. A point is a claim on something later. A Card is an object now. A Set turns that object into unfinished progress. An Album makes the progress persistent.
Most marketing content disappears after consumption. The sale email is read and archived. The social post slides down the feed. The newsletter item is useful for a minute and gone by lunch. Even most campaign gamification ends with the campaign. There is no residue in the customer’s hands.
A collection behaves differently. If three Cards arrive twice a week, the customer does not merely accumulate six opens. After a month there may be twenty-four objects, a Set two Cards short of complete, some things remembered and some still to learn. The twentieth interaction can be worth more than the first, because it sits on top of the previous nineteen.
That persistence suggests a useful inversion of CRM. CRM remembers what the company knows about the customer: purchases, clicks, segments, scores and predicted propensity. The Album remembers what the customer has taken from the relationship: Cards owned, Sets advanced, things learned, choices made, challenges mastered, gifts given.
CRM is company memory about the customer. The Album is customer-facing memory of the relationship.
The Album reads that memory back on three clocks. Hold is what has arrived, and it moves with every edition. Finish is how many Sets are complete, and it moves over weeks. Know is how much the customer can still answer for; it moves slowly, and it can fall when attention lapses. A points balance only goes up until it is spent. Know is the one number that tells anybody whether the relationship is working.
Table 1. Points and Cards solve different jobs
| Loyalty points | Sets and Cards | |
| Primary job | Reward purchase | Reward attention |
| What accumulates | A balance | Objects and progress |
| Earned by | Spending | Reading, predicting, recalling |
| Value | Mostly economic | Knowledge, progress, completion, identity |
| Between purchases | Often quiet | Designed to stay active |
| Social behaviour | Mostly private | Gift, Circles, later Trade |
| End state | Redeemed or expired | Finished, known, remembered |
| Brand role | Payer of benefits | Issuer and sponsor of a world |
This is why the comparison needs restraint. Cards are not a better discount. They do not beat an airline mile at buying a flight or a grocery point at reducing the bill. Their advantage appears in the much larger share of time when there is no transaction to subsidise. Points pay for purchases. Cards can pay for attention.
3
An Old Idea, Newly Economic
Brands issued collectible Sets for over a century. Physicality constrained them. Digital changes both the economics and the physics.
Cigarette cards began in the United States in the second half of the nineteenth century as packet stiffeners, pieces of card inserted to stop flimsy packets being crushed. The first dateable picture card appeared in 1879, and rival manufacturers soon followed with whole series. Many were educational — natural history, flags, ships, famous places — with a picture on one side, a description on the reverse, and the name of the set and its issuer printed on the card.
Read that last sentence with this essay’s vocabulary. A Front and a Flip. A named Set. An issuer who is not the subject. The grammar of a brand-issued collectible is more than a century old.
When the tobacco companies did not bring cards back after the Second World War, a tea company did. From 1954 until 1999, packets of Brooke Bond tea carried illustrated picture cards — British birds first, then wild flowers, space, transport, costume and inventors — fifty-nine series in the UK alone. The Science Museum Group holds a 1973 Brooke Bond album, History of Aviation, designed for fifty cards of important aircraft. Panini, founded in Modena in 1961, built an entire business on sticker collections and today describes more than a thousand collection launches a year. The appetite to complete, compare and exchange is not new.
These programmes did not fail for lack of appetite: Brooke Bond ran for forty-five years, and Panini’s physical business is still large. What held them back was physicality. Every card had to be printed, packed, inserted and shipped inside a product. The packet was the only channel, so the brand reached only people who were already buying. A card could never change once printed. And the issuer never knew who was collecting, what they held or what they had learned.
Digital removes each of those constraints. Distribution costs almost nothing per additional Card. AI makes production cheap. The issuer knows who owns what. And a Card can change after it is issued. The first two change the economics; the last two change the physics.
On the economics specifically: AI lowers the cost of production: researching, drafting, illustrating and personalising large numbers of Cards, provided humans remain accountable for the facts and the editorial judgement. Email lowers the cost of distribution: it can carry a rich, interactive object repeatedly, at scale, without paying a fresh media toll for every impression.
Email is not free. Deliverability, consent, reputation, infrastructure and production all cost money. But once that machinery exists, the incremental cost of another rich email is tiny compared with paid business messaging. From January 2026, Meta’s rate for a WhatsApp marketing message in India is about ₹0.86 before GST and platform charges. Three Cards twice a week to a million customers would cost roughly ₹9 crore a year in Meta fees alone. App push has different economics, but it reaches only people who installed the app and allowed notifications, and it normally pulls them back into the app for the experience. Email can carry the experience itself.
That leads to a clean division of channels. Email carries progression. The web carries archive, provenance and the full Album view. A brand app can show the same Set view to people who already use it. Person-to-person WhatsApp carries invitation and sharing, at no cost to the brand when one customer sends something to another. The cheapest rich channel does the repetition; the social channel does the spread.
Cheap production creates a second problem: too much content. The bottleneck shifts from throughput to taste. A Set still has to survive seven tests before a single Card is drawn: Name Three; twelve surprises rather than one surprise repeated; a legible gap; objecthood; rarity without ranking; meaningful relations between members; and, where possible, a live edge — the tests I set out in Got, Got, Need. A factory can make a hundred Cards quickly. It still needs judgement to decide whether the hundred deserve to exist.
This is also why I avoid “gamification” as the headline. Points, badges and leaderboards are one family of mechanics — the family this essay argues against. Collecting is older, quieter and more persistent. The game is not an ornament placed on a campaign. The collection is the memory structure that survives it.
4
The Machine
Six components answer six questions a CMO would ask.
Figure 2. Earn, accumulate, remember, return, persist and spread — with email underneath as the recurring distribution rail.
The Digest factory answers “Why open?” It publishes two or three short, useful items from the customer’s world even when the brand has nothing to sell. The job is not to smuggle a promotion into editorial clothing. It is to earn another voluntary visit to the inbox.
The Sets and Cards factory answers “Why come back?” A finite Set turns unrelated pieces of content into visible progress: twenty colours, fifty investing terms, twenty-four destinations. The reader can see what is held, what is missing and what can be finished.
The Commit gate answers “Why interact?” Every Card arrives Sealed. Before the Front is revealed, the reader makes a small commitment: Predict is the first step of the grammar. That moment of effort matters because the Card is not simply received; it is encountered.
Living Cards answer “Why reopen what I already own?” The Card can keep changing after it is acquired. The Album answers “Why does this get more valuable over time?” It keeps persistent state across editions. Gift, Circles and later Trade answer “Why bring someone else?” The collection acquires a social edge without a cashback budget.
Email, beneath all of it, answers “How can the brand afford to do this repeatedly?” It is the rail that places Read, Commit and Cards inside the same experience, with the brand’s website or app providing a persistent view rather than replacing the inbox habit.
Two rules govern what goes into a Set. The first tests quality: the Set must be worth collecting even if the logo disappeared; otherwise it is merchandise pretending to be culture. The second tests linkage: it must come from the brand’s world, and the customer must know who issued it; otherwise the customer may remember the album and forget the sponsor.
“The Colours of India — a Set issued by Asian Paints” passes both. Colour, architecture, craft, textiles and festivals belong to a world Asian Paints has permission to sponsor. “20 Great Asian Paints Products” fails the first test, because the brand has made itself the subject. A handsome “Great Inventions” Set from a paint company fails the second: the customer enjoys it and remembers nothing about the issuer. The brand is the issuer and sponsor, never the subject. An independent editorial team authors the Set; the brand gives it standing and distribution.
5
Aha and Wow
Reveal once; live forever.
There are two moments in the life of a Card, and they do different work.
Aha happens once. The Card is Sealed. The reader predicts. Then the Front appears, and the Flip explains the answer, adds context and offers the memorable detail. The surprise is earned because it follows a commitment. Prediction before a reveal can strengthen memory for what was predicted. That is the learning mechanism the Commit gate is designed to exploit. Without the gate, a Card becomes another pretty tile in an inbox.
Wow happens later. Something in the world changes, and a Card the customer already owns changes with it. A mission launches. A record falls. A rule changes. A price crosses a threshold. An investor-education Card on a regulation acquires a new date and status. The next email does not have to invent a reason to open; it can simply say: one of your Cards changed.
Figure 3. Aha creates the first memory; Wow gives the object a future.
This live edge is the sharpest difference between digital collecting and its physical predecessors. Printed cards could be beautifully designed, scarce and social, but they froze at the moment of printing. A loyalty balance changes only when the economics change. A Living Card can change because the world did.
Rewards then become optional rather than foundational. Completion produces status: Collector, Scholar, Circle Collector, Circle Masters. Mastery is always verified by the system, never conferred by another person. A Circle that finishes can earn a Card carrying its members’ names and completion date — worthless to anybody outside the group, and for that reason valuable to everybody in it. The Know number shows how much of a Set the person has retained, not merely held. In domains such as investor education, completion and mastery can support a credential. None of these needs to be cash-like.
I would keep Mu out of this B2B proposition at the start. Introducing a currency while arguing that attention should not be reduced to points would muddy the distinction. The first pilot should answer a cleaner question: will people return because the publication is useful and the collection is becoming theirs?
6
From Private Engagement to Earned Reach
A loyalty point is private; a missing Card starts a conversation.
The first social mechanic is Gift. A spare is minted through mastery or a challenge and passed on with one tap. “Maya is missing the Card you have two of” is a far more human prompt than “refer a friend and earn 100 points”. The object gives the invitation its meaning, and reciprocity stays open.
Circles come next. A family, an office group, a classroom or a group of friends can see its collective gaps and celebrate a shared finish. The Circle makes progress social without turning the Card into a price. Trade comes later, as the distinct movement verb it is. It should arrive after collection, Gift and Circle behaviour are proven, with no marketplace or exchange-rate logic in the first proposition.
This is also where the architecture asks the brand to give something up. There is one Album per person, not one per brand. Brands issue Sets into that person-owned container. A brand’s website or app can show the customer a view of its Set, but it should not recreate a separate album that the customer has to remember to visit. The analogy that holds all the way down is the stamp album: countries issue stamps; the collector owns the album.
The brand gives up the container and gains persistence inside it.
A stand-alone Set already improves the brand’s own Digest. The shared Album is what compounds later. If five issuers each send useful editions twice a week, the person may have something arriving or changing on most days, without any one brand pretending it can earn a daily open.
Figure 4. The promise hierarchy. One brand must win with Digest and Set; the shared Album adds frequency only after that works.
That distinction matters. The B2B promise is not “we will make customers think about you every day”. It is “we will help you stay mentally available between purchases by placing useful, persistent objects into a relationship the customer owns”. The network may create daily Album activity. The brand benefits from frequent presence inside it rather than demanding daily recall of itself.
What, then, does the brand buy? Greater owned reach between purchases; stronger issuer recall; richer declared and behavioural signals from predictions, topic choices and recall; and, potentially, more mental availability when the next buying situation arrives. None of this is claimed to create revenue automatically. Whether it does is what later experiments must establish.
Nothing in a Set is ever withdrawn. Urgency does not need expiry. Where a brand wants scarcity, it lives in the particular copy: where a Card was minted, the level of recall at which it was earned, and whether it came as a gift. Those are renewable. A closing date is not. Expiry recreates the worst instinct of old loyalty design; an attention system should make the past more valuable, not periodically erase it.
7
The Test
The claim is testable in one quarter, and it names its own failure.
The experiment needs three concurrent, randomised arms, never a comparison with a prior period. Arm A receives the brand’s current best effort. Arm B receives the Digest without Sets. Arm C receives the same Digest with the Commit gate, Cards, one active Set and persistent Album state. It should run long enough for novelty to fade and progress to become visible.
Figure 5. The three-arm test separates the value of publishing from the additional value of collecting.
The first measure is Habit Rate: the share of readers who open at least 60% of the editions they chose over a four-week window. Then come CRR and Real Reach lift, Set Progress, next-open rate by Cards held, and the performance of Living Card editions against ordinary editions. If the Set is doing real work, the probability of the next open should rise as ownership and Know rise.
One measure matters most for the B2B case: Issuer Recall. Every few weeks, ask an unaided question such as “Who publishes your Colours of India Set?” It guards against an elegant failure, in which customers love the Cards, remember the album and forget the brand that funded the experience. The Set must build customer value and issuer memory at the same time.
The falsification conditions are written down before launch:
- The core claim fails if Digest with Sets produces no higher Habit Rate than Digest alone.
- The B2B case fails if Issuer Recall does not rise with Set engagement, even if the consumer product succeeds.
- Wow is decoration if “a Card you own changed” editions do not outperform ordinary editions.
- The network upside fails if brands refuse the shared Album, and the single-brand economics must carry the proposition alone.
This discipline matters because “engagement” is usually a word that survives every result. A collectible attention system should not. It makes specific claims about repeat behaviour, memory, persistence and spread. Each can be measured. Each can disappoint.
The upside, if the tests work, is bigger than a new email format. The brand acquires a second relationship with the customer: not only a ledger of transactions, but a growing set of things the customer has chosen to know and keep. The Digest earns the periodic visit. The Set carries memory across the gap. The Living Card creates reasons to return without inventing a campaign. The Album turns those fragments into something the customer can see.
Points reward purchases. Cards reward attention. One prices the transaction; the other can make the silence between transactions productive.
The loyalty programme of the future may look less like a wallet and more like an Album.
Key points
- The argument is with the points mechanic, not with loyalty. Points reward purchases; Cards reward attention. They are two rails, and Cards cover the far larger share of time, when the customer is not buying.
- Forgotten and unspent balances can signal a quiet relationship. Breakage profits from forgetting; an Album is designed to be remembered.
- CRM is company memory about the customer. The Album is customer-facing memory of the relationship.
- Brands issued collectible Sets for over a century, with a Front, a Flip and a named issuer. Physicality constrained them; digital changes both the economics and the physics.
- The Commit gate creates Aha. The Living Card creates Wow, which no printed card or points balance can offer.
- A Set must be worth collecting without the logo and must come from the brand’s world. An independent editorial team authors; the brand issues and sponsors.
- One brand, one Digest, one Set is a complete proposition. The shared Album is upside.
- Three concurrent arms, one quarter, and an Issuer Recall measure decide whether any of this is true.
