Your first digital product: choose by constraint, not by trend.

The first product’s job is learning, not scale.

Almost every bad first-product decision comes from optimising for the wrong thing. People pick the shape that would be best if it worked, rather than the shape that finds out fastest whether it will.

A first digital product is an instrument. It exists to convert an assumption — that somebody will pay you for this specific outcome — into evidence. Everything you would optimise for later (margin, leverage, passive delivery, a catalogue) is downstream of an answer you do not have yet, and each one of those optimisations makes the instrument slower and more expensive to read.

The four things it buys you

1. 01The objection language. What people say when they do not buy, in their words. This is the raw material for every page, email and offer you write afterwards, and there is no way to get it except by asking for money.
2. 02Whether they finish. Buying and using are different behaviours with different causes. A product people buy and abandon is a marketing success and a business problem, and you cannot see the difference from the revenue line.
3. 03What support actually costs. Every shape has a delivery cost that only shows up after real buyers arrive. Ten of them will tell you whether the thing you designed is the thing you can run.
4. 04What the next offer should be. The people who buy the first thing tell you — usually directly — what they wanted next. That sentence is worth more than any market research you could have done instead.

A product designed to scale answers none of those four. It defers all of them behind weeks of production, and then presents the answers all at once, at the point where changing course is most expensive.

Nearly every first digital product is one of four things.

The catalogue looks endless and is not. Memberships, cohorts, bundles, challenges and productised services are all variations on four underlying shapes, and the four differ from each other in ways that matter before you have any customers.

The paid workshop — you are selling a date

One live session, one outcome, one date, a price on the door. It is the only shape you can sell before it exists, which is the entire point: the deadline that forces you to build it is the same deadline that tells the buyer why to act now. You find out whether people will pay for the outcome roughly two weeks after deciding to ask.

The trap. Treating it as a lesser version of a course, and so under-pricing it and over-stuffing it. A workshop that tries to cover a curriculum in two hours delivers nothing; a workshop that gets one thing built and working during the session is worth considerably more than its usual price.

The template or toolkit — you are selling the hours you already spent

A spreadsheet, a set of documents, a swipe file, a Notion system, a design kit. The honest version of this shape is a thing you already made for your own work, cleaned up and explained. It is the lightest to deliver and the fastest to buy, because the buyer can see exactly what they are getting.

The trap. Inventing an artifact specifically to have something to sell. If the template did not survive contact with real work, it will not survive contact with a buyer either, and the shape gives you nowhere to hide — a file is judged in about eight seconds.

The coaching offer — you are selling access

One-to-one or a small group, on a stated cadence, for a stated period. Nothing to build, the highest price of the four, and the fastest feedback loop that exists: you watch the buyer try to use what you know, in real time, and every place they get stuck is a line item for whatever you productise later.

The trap. Selling more of it than your calendar can hold, because it is the shape where success and workload are the same variable. Decide the maximum number of clients before you decide the price, not after.

The self-paced course — you are selling a finished system

Recorded modules, a portal, a path from where the buyer is to where they want to be. It is the shape most people picture when they think “digital product”, and it is the worst first move for exactly that reason: the picture is of a finished thing, so people try to build the finished thing first.

The trap. Spending eight weeks recording a curriculum you invented rather than one you have taught. The material is not the hard part of a course; knowing which order it goes in and which parts people get stuck on is, and neither is knowable before you have taught it live at least once.

Score the four shapes on the four things that constrain you.

Not on how attractive they sound. Effort to build, price ceiling, delivery weight and proof required are the four axes on which the shapes genuinely differ, and between them they explain most first-product failures.

Read the levels as comparisons between the four shapes, not as absolute claims. “Low effort” means low relative to the others; a workshop is still real work. Your own situation moves these — a template is low effort if you already own the artifact and high effort if you do not.

How to read a column

• →Effort to build tells you how long you are exposed. It is the time between deciding and finding out. Every week in that gap is a week of conviction with no evidence behind it, which is where most first products quietly die.
• →Price ceiling tells you how many buyers a real month needs. A high ceiling means a small number of buyers is a result. A low one means you need volume, and volume needs distribution you may not have yet.
• →Delivery weight tells you what success costs. The question is not what one sale costs to deliver but what fifty do. A shape where the answer is fifty times the first is a shape with a ceiling made of your calendar.
• →Proof required tells you whether you can sell it yet. Everything is easier to sell with evidence. The shapes differ in how much they need: a two-hour session on a date needs far less than a system that claims to produce a result.

There is no winning row. The scorecard is useful precisely because the shapes trade against each other: the cheapest to build has the lowest ceiling, the highest-priced has the heaviest delivery, and the one that runs itself takes the longest to make. Which trade you should take is decided in the next chapter, and it is decided by what you already have.

Start from what you already have, because that is the only thing that is not a guess.

The scorecard narrows the field. This rule picks. It has one input: which of three assets you currently hold — an audience, expertise without an audience, or results you produced for clients.

The reason this works is that each asset is strong on a different axis of the scorecard, and a first product should be chosen to spend the strength you have rather than to build the one you lack. An audience is distribution — spend it on speed. Expertise without an audience is a product with no route to market — pick something small enough that borrowed routes are sufficient. Client results are proof — spend them on price.

If you hold more than one

Most people hold two, weakly. Rank them by which one you could demonstrate to a stranger this afternoon. An audience you can point at beats client results you would have to reconstruct from memory; results you can describe in a sentence beat a list that never opens anything. Pick the shape that spends the one you can actually show.

If you hold none of the three

Then the first product is not the next move. Do the work for two or three people — cheaply, or free with an explicit agreement that you can write about the outcome — and you will have the third asset within a month. That is a shorter path than building a product and then trying to manufacture credibility for it afterwards, and it is honest, which the alternatives usually are not.

A shape is not an offer. A promise is.

“A workshop about email marketing” is a shape with a topic attached. Nobody buys topics. The move from shape to offer is one step and it is the step that decides whether anything sells.

Write the promise as a thing that will be true afterwards and was not true before, for a specific person, by a specific point. “A two-hour workshop that leaves you with a welcome sequence written, loaded and sending by Friday” is a promise. It is also considerably easier to build than the topic version, because the scope is now decided by the outcome rather than by how much you know.

Narrow the person, too

A promise made to everyone converts fewer people than the same promise made to someone specific — not because the audience is smaller but because the reader can tell whether it is about them. For a first product, narrow enough to feel uncomfortable is about right. You can widen later on evidence; you cannot recover a launch that nobody felt addressed by.

The smallest version that still delivers it

Once the promise is written, ask what the least you could build and still keep it is. Usually the answer removes a module, a bonus, a second format and a community — none of which the promise required, all of which were added because the product felt thin. Thinness is not the problem it appears to be. An unfinished product is.

Price is set by delivery weight and carried by proof.

Two forces decide what a first product can charge. Delivery weight sets the floor — below it you are paying to work. Proof sets the ceiling — above it the buyer is being asked to take a risk you have not helped them with.

This is why the same promise, in two different shapes, carries two different prices. Nothing about the value to the buyer changed; what changed is how much of your time is inside it and how visible the evidence is. It is also why the standard advice to “charge what it is worth” is useless as a first-product instruction — the worth is exactly what you are trying to find out.

When the proof is thin, carry the risk yourself

Proof and risk are two ends of the same lever. If you cannot yet show that this works, you can offer to absorb the cost of it not working — a stated, plainly written guarantee does the job that testimonials do later. It is not a trick and it is not free: it is a decision to take the risk onto your side of the table because you are better placed to carry it than a first buyer is. Refund policies and guarantees covers how to design one you can afford.

Not zero

Free is a different product with a different audience, and it answers a different question. People will take almost anything that costs nothing, which means downloads tell you about interest and nothing about willingness to pay — the single fact a first product exists to establish. Charge something, even if the something is small, and treat the first ten payments as the result rather than the revenue.

Sell it once before you make it properly.

Every shape has a version that can be sold before it is finished. Taking that version to market first is not a shortcut — it is the difference between building for a buyer and building for a hypothesis.

1. 01Write the page as though it exists. The promise, who it is for, what happens after purchase, the price, and what they get if it is not for them. Writing this is also the fastest way to discover that the promise is vague, which is much cheaper to learn now.
2. 02Put it in front of people who have the problem, individually. Not a broadcast. Twenty personal messages to people you could name outperform any announcement for a first product, and every reply is usable copy whether or not they buy.
3. 03Ask for money, on a date. A presell with a delivery date is the only test that separates polite enthusiasm from demand. If nobody pays, you have saved the eight weeks; if a handful do, you now have a deadline and an audience of buyers to build with.
4. 04Deliver the first cohort by hand. Manually, ungracefully, over-servicing the first ten. What you would have automated is now visible, and half of it turns out to be unnecessary.

The workshop shape does this natively — the product is a date, so selling it before it exists is simply how it works. For a course, the equivalent is a presell against a stated start; preselling an online course covers how to do that without leaving buyers holding a promise you cannot keep. For a coaching offer it is two calls booked against a written scope. For a template, it is showing the actual artifact to five people and asking what they would pay.

If you have nobody to ask

That is a distribution problem and it is solvable without a list — borrowed audiences, partner sends, communities you are already in, and direct invitations all work, and all of them need lead time. Launching without a list is the sequence. Start it before the product is ready rather than after, because the asking is the part with a calendar dependency on other people.

Every shape is a draft of the next one.

The first product is not a smaller version of the business you want. It is the input to it. These are the four transitions that show up most often, and what each one actually consists of.

Read down the last column. In every row, the thing that carries over is not the material — it is the knowledge of where people get stuck. That is the asset the first product exists to produce, and it is the reason the order matters: a course built after three workshops is a different course from one built before them, even if the slides are the same.

Graduate on evidence, not on ambition

The common mistake is graduating because the next shape sounds more like a real business — moving to a course because courses scale, or to a membership because recurring revenue sounds calmer. Both of those add obligation before they add anything else. Move when the current shape has stopped teaching you something new, and not before; subscription versus one-time is worth reading before the membership version of that move in particular.

You can also graduate sideways

A template that sells well does not have to become a course; it can become a better template with a higher price and a small installation service attached. A workshop that fills does not have to be recorded; it can simply run monthly. Vertical movement is the default advice because it is the most legible, not because it is always right.

Read the signal before you change the product.

Most first products get changed for the wrong reason, because the symptom people notice — “it did not sell” — has at least four different causes with four different fixes. These five signals separate them.

1. 01Nobody buys, and the objections are all about price. This is usually not a shape problem. Price objections on a first product are proof objections wearing a disguise: the buyer cannot tell whether it will work, so the number feels like a risk rather than an exchange. Strengthen what they can verify before you cut the price, and consider carrying the risk yourself with a stated guarantee instead.
2. 02Nobody buys, and there are no objections at all. Silence is a distribution reading, not a product reading. Before you change anything about the product, count how many people who genuinely have this problem actually saw the offer with enough context to act on it. If the honest number is under a hundred, you have not tested anything yet.
3. 03People buy and then never start. The promise was bought and the delivery was not designed. Almost always the first step is unclear or the thing is bigger than what they thought they were buying. Fix the entry — one obvious first action inside a day of purchase — before you add material to the middle.
4. 04It sells, and you quietly dread delivering it. You picked a shape whose delivery weight rises with every sale. That is survivable for ten buyers and not for fifty. Cap the number you take, raise the price so fewer sales cover the same month, or move the repeatable part of the delivery into something that does not need you in the room.
5. 05Eight weeks in and there is still nothing to sell. Effort-to-build exceeded what you actually had. This is the failure mode that costs the most and teaches the least, because you never got to the part where the market answers. Ship the workshop version of the same promise — one date, one outcome — and let the finished thing be assembled out of what you learn running it.