Six terms get used interchangeably in strategy conversations. They answer four different questions, and mixing them up is how a meeting agrees on a word instead of a plan.
The reason these terms feel interchangeable is that they all involve giving something away. The reason they are not interchangeable is that each one is an answer to a different question. Sort them by the question first and the confusion dissolves.
| Question | Terms that answer it |
|---|---|
| Pricing mechanic. How does someone experience the value before paying? | Free trial, freemium |
| Unit economics. Where does the margin actually come from? | Loss leader, razor and blades |
| Sequencing. What do we sell first, and what does that let us sell next? | Wedge, land and expand |
| Channel. How do people find out we exist? | Engineering as marketing |
Free trial is the full product, time-boxed. The clock is the forcing function, and the user knows from day one that they are a guest. Freemium is a permanently free tier of a limited product. There is no clock. The forcing function has to be built into the limit itself, which is a much harder design problem and the subject of Lesson 4.
The practical difference: a trial converts on evaluation ("have I seen enough?"), a freemium tier converts on outgrowing ("I now need more than this gives me"). They fail differently too. A trial fails when the user never reaches value inside the window. A freemium tier fails when the free version is quietly sufficient forever.
Loss leader, from Lesson 1: a lead item sold below its own cost, with margin recaptured elsewhere. Razor and blades is the special case where the lead item makes the paid item necessary. Every razor-and-blades model is a loss leader; almost no loss leader is razor and blades, because necessity is rare and hope is common.
Wedge: attack a narrow, overlooked problem, win users on something simple, then expand into the bigger product from that position.1 Land and expand: get into an account cheaply on one team or one module, then grow seats and modules inside the same customer. A wedge expands the product. Land-and-expand expands the account. Both can be true at once; they are not the same move.
Engineering as marketing: spend engineering time on a free tool that attracts exactly your target audience, and treat it as a marketing channel with a cost per lead, not as a product line with a roadmap.2 This is the one most often mislabelled. If a thing exists to generate signups you will nurture into a different product, it belongs in the marketing budget and gets judged on cost per qualified lead. Calling it a loss leader implies a recapture path through that product, which is a claim you probably cannot support.
The terms are not rivals. A single product can be all of these at once, and stating each layer separately is what turns a slogan into a plan:
A wedge (narrow: capture what happens in meetings, not the whole CRM), monetised as freemium (permanent free tier with a usage cap), whose unit economics are a loss leader (free-tier inference costs money, recaptured on subscriptions and later on CRM seats), promoted through engineering as marketing (a separate free video tool that draws the same audience), landing in small agencies and later expanding across their seats.
Six sentences, six different things to be wrong about. That is an improvement over one sentence you cannot argue with.
Which of the four questions does each term answer?
Real cases. Pick the term that fits best.
The document under review is titled "Loss leader Product ideas" and opens by asking how to enter France before introducing the CRM. Read the proposal through the four questions and it decomposes cleanly:
That is a coherent structure, and being able to say it in those words is already most of a review. The open question - the one Lesson 3 arms you for - is whether the last step is a real expansion or a hope.
Read Chris Dixon's The "thin edge of the wedge" strategy. It is about 600 words, it is the origin of the term as used in software, and Lesson 3 assumes you have read it.
Confused by an edge case - is open source a loss leader, is a free tier for students land-and-expand? Ask your teaching agent. Edge cases are where these definitions earn their keep.