Two businesses sell the same underlying capability. One charges three times what the other does and closes more easily. The difference is frequently not quality, reputation or sales skill. It is that one has packaged the capability into an offer and the other is selling components and letting the buyer assemble them.
Packaging decides what question the buyer is answering. That question determines what they compare against, and the comparison determines the price they are willing to pay.
Why does packaging change what something is worth?
Because value is judged relative to a comparison set, and packaging determines the comparison set.
Sell hours and you are compared to other people selling hours, so the conversation is about hourly rate. Sell a defined outcome and you are compared against the cost of not having that outcome, which is usually a much larger number.
Nothing about the underlying work changed. What changed is what the buyer is measuring you against.
Packaging does not describe the offer. It decides what the offer gets compared to, and the comparison sets the price.
This is why packaging decisions belong before pricing decisions. A price is a claim about a package, and rearranging the package changes what claim is available to make.
Why bundling works: the variation argument
The core economic reason for bundling is that different buyers value the same components differently, and bundling narrows that variation.
Imagine two components and two buyers. Buyer one would pay eighty for component A and twenty for component B. Buyer two would pay twenty for A and eighty for B.
Sold separately, you must pick a price for each. Price A at eighty and only buyer one takes it. Price it at twenty and you capture both but leave sixty on the table with buyer one.
Sold as a bundle, both buyers value the pair at one hundred. Price the bundle at one hundred and both buy, and you have captured the full amount from each.
| Values A | Values B | Values bundle | |
|---|---|---|---|
| Buyer one | 80 | 20 | 100 |
| Buyer two | 20 | 80 | 100 |
| Variation | Wide | Wide | None |
That collapse in variation is the whole mechanism. Bundling works best when buyers value components differently and, ideally, in opposite directions. It works least well when everyone values the components the same way, since there is no variation to smooth.
Bundling also has real buyer-side benefits that are not just extraction. One decision instead of several reduces effort. One price is easier to evaluate and to get approved internally. And components designed to work together often genuinely are worth more combined than separately.
Pure, mixed and unbundled
Three structures, each suited to different conditions.
Pure bundling means the components are only available together. Simple, and it maximises the variation-smoothing effect. The cost is that buyers who want only one component either pay for everything or leave.
Mixed bundling offers both the bundle and the individual components, with the bundle priced attractively relative to buying the parts. This is the most common structure in practice, and research generally finds it outperforms either extreme, because it captures buyers who want everything at a good total price while still serving buyers who genuinely need only one thing.
Unbundled means everything sold separately. This suits buyers with narrow, specific needs and lets them pay only for what they use, but it exposes each component to direct comparison and produces a more complex decision.
If you are unsure, mixed bundling is usually the right starting point. It hedges, it lets you observe what people actually choose, and the data from those choices tells you more about your buyers than any amount of speculation.
How this is taught inside Astra Trainer
Offer construction comes before price setting in the Offers & pricing direction of the Marketing, Sales & Attention world, thirteen courses on shaping an offer people want and setting a price that holds up when a buyer says it costs too much.
That sequencing is the useful part, and it is the opposite of how most people approach it. Pricing problems are usually offer problems, and rearranging what is included resolves more disputes than adjusting the number ever does. Lessons take about five minutes, a guide walks you through anything strange, and the discussion thread under each lesson tends to fill with people posting their tier structure and being asked which buyer each tier is actually for.
How many tiers, and what separates them?
Three is the most common answer and it is common for reasons rather than superstition. It gives a floor, a target and a ceiling, creates a middle option that benefits from both comparisons, and remains simple enough to evaluate quickly. More than four tiers reliably increases confusion and decision abandonment.
The harder question is what separates them, and this is where most tier structures go wrong.
Bad separation withholds arbitrary features. The buyer sees a list of things they are not allowed to have and experiences the structure as a penalty. It invites resentment and workarounds, and it makes the lower tiers feel deliberately crippled, which they are.
Good separation tracks who the buyer is. Tiers differ by scale of use, by number of people involved, by depth of need, or by level of support required. A small buyer genuinely does not need what a large buyer needs, so the lower tier is not a punished version of the higher one, it is a correct fit.
The test is whether someone in the lower tier feels well served or feels they are being squeezed. If your smallest tier exists only to make the middle one look good, buyers notice, and the people who notice most clearly are the ones you most want.
Good separating dimensions include volume or usage, number of users, depth of functionality where the depth genuinely matters to bigger operations, and service level. Poor ones include artificially capping something cheap to provide, removing basic usability, or restricting things that cost you nothing.
What should never be the thing you withhold?
Some things do not belong behind a paywall regardless of how well they would convert.
Safety and security. Charging extra for basic security features means selling a product with a known vulnerability to anyone who does not upgrade. This has been done, it converts, and it is indefensible.
Basic reliability. If the lower tier fails routinely and the fix is to pay more, you have not built a tier structure, you have built a malfunction with a price attached.
Data portability. Making it hard for someone to leave with their own data is retention through obstruction. It holds people briefly and guarantees they warn others.
Anything required to make the thing work at all. If the base tier is unusable without an add-on, the base tier is not an offer, it is a lead magnet with a misleading price.
The test for a defensible tier. Could you explain the difference between two tiers to a buyer, out loud, and have them agree it is reasonable? "The higher tier supports more users and includes priority response" passes easily. "The lower tier has a limit we added purely to push you upward" does not. If the rationale only works unspoken, buyers will eventually infer it anyway.
When is unbundling the right move?
Bundling is not always the answer, and several situations favour breaking things apart.
When buyers only want one component. If most of your market genuinely needs one thing, bundling forces them to pay for irrelevance, and a competitor selling that one thing will take them.
When a component can stand alone. Some components have enough value to be their own product with their own market, and burying them in a bundle hides them from buyers who would have bought that alone.
When the bundle has become incoherent. Bundles accumulate. Things get added over years until the package is a list rather than a proposition, and nobody can say what it is for. At that point unbundling and rebuilding around a clear purpose usually beats continuing to append.
When the price has become a barrier. If the total is beyond what a large part of the market can approve, a smaller entry offer can bring people in who later expand.
Industries tend to cycle between the two, bundling until packages become bloated, then unbundling until buyers tire of assembling things themselves, then bundling again. Neither state is permanent and neither is correct in the abstract.
Practising structures rather than reading about them
Packaging is a design skill, and design skills come from repeated attempts with feedback rather than from a single explanation. The difference between knowing that tiers should track buyer type and actually producing a structure where they do is a gap that reading does not close.
The Marketing, Sales & Attention world is built for that gap. Quizzes follow each topic, each course ends with a ten-question final exam, and the daily Connections round and 10x10 crossword are generated from that world's own lessons, so bundle, premium, anchor, decoy and the rest return as practice with a fresh round every day. A Circle gives you a small group with a shared weekly goal, and packaging in particular benefits from other people, since the flaw in your own structure is almost always invisible to you and obvious to someone else.
How do you package a service rather than a product?
Services are harder because the natural unit, time, is the worst possible thing to sell.
Selling hours creates three problems at once. It caps your income at hours available. It makes efficiency a penalty, since getting faster means earning less for the same result. And it invites comparison on rate with everyone else selling hours.
Four alternatives work better.
Outcome packages. Define a result, a scope and a price. The buyer purchases the outcome and stops caring how long it takes, which realigns your incentives with theirs.
Productised services. A fixed scope delivered the same way every time at a fixed price. Easy to buy, easy to deliver, and easy to improve because it is repeatable.
Retainers. Ongoing access for a recurring fee. Predictable for both sides, and it works best when the scope of what the retainer covers is defined clearly enough to prevent slow expansion.
Tiered access. Different levels of involvement, from self-directed with materials, through group support, to one-to-one. This lets one body of expertise serve several segments at different prices without any of them being a degraded version of another.
In each case the shift is the same: from selling input to selling output. That single change usually does more for a service business's pricing power than any amount of work on the number.
What to take from this
Packaging determines the comparison set, and the comparison set determines the price. That is why packaging comes before pricing.
Bundling works by smoothing variation in how different buyers value components, so it works best where that variation is wide and pointing in different directions.
Mixed bundling usually beats the extremes, and it produces information about your buyers as a side effect.
Tiers should separate buyers, not withhold features. If the lowest tier feels punished, the structure is working against you among exactly the people whose opinion travels.
And for services, sell the output rather than the input. Hours are the unit that caps your income and penalises your efficiency simultaneously.
How much cheaper should a bundle be than buying the parts?
Enough that the saving is obvious at a glance, without being so steep that the individual components look pointless. The right number depends on how much you want to steer buyers toward the bundle, which depends on whether the bundle is genuinely better for them.
Should the middle tier be the one most people choose?
Usually, and it is worth designing deliberately for that rather than hoping. The middle tier benefits from comparison in both directions, so it should be the option that genuinely fits the largest segment rather than the option you most want to sell.
Is it wrong to have a tier almost nobody buys?
Not if it is a real offer that genuinely serves someone. A premium tier that a small number of buyers legitimately need is fine, and its presence helps frame the others. A tier nobody could sensibly buy is a decoy, and if someone ever does buy it you have a delivery problem.
How do I stop a retainer expanding beyond what was agreed?
Define the scope in terms of what is included rather than hours available, review it on a stated cadence, and treat expansion as a conversation about a different package rather than something absorbed quietly. Scope creep in retainers is almost always a definition problem at the outset.
Where can I learn this systematically?
The Offers & pricing direction inside the Marketing, Sales & Attention world of Astra Trainer runs to thirteen courses on shaping offers and holding prices. Lessons take about five minutes and the first needs no card. You can see what is inside the world here.