0.2 What the buyer is actually paying for

A product is not what your organisation produces. It is what changes for the buyer once they have it, and those two descriptions coincide far less often than the people producing it assume.

A hotel sells rooms. Some of its guests are buying somewhere to sleep. Others are buying a predictable start to a difficult morning, or the ability to say to a client that they stayed somewhere reputable, or four hours of not being at home. The room is identical in every case. The offer is not, and the marketing that works for one of those buyers is invisible to the others.

The mechanical version of this is the distinction between a feature and an outcome, which is taught everywhere and applied almost nowhere, because the translation is genuinely difficult when you know the product well.

Feature	Outcome the buyer is paying for
Automated invoice reminders	You stop being the person who chases friends and clients for money
Twenty four hour response time	You can promise your own customer a date
Two hundred integrations	The thing you already use does not have to change
Experienced trainers	You will not injure yourself in the first fortnight

Notice what the right column does that the left does not. It contains a person, in a situation, with something at stake. The left column contains the organisation's inventory.

The word that hides the outcome

The most common failure in this exercise is the word "quality", followed closely by "efficiency", "solutions" and "excellence". They feel like outcomes because they are positive and abstract. They are the internal vocabulary of the producer, they name no situation, and no buyer has ever complained about a lack of them in those words. If your outcome column contains any of them, you have translated the feature into a different feature.

There is a framework in wide circulation for this, usually called jobs to be done, which says that people hire products to make progress in a situation. It is a useful framing device and you should know what it is worth: there is no controlled study, no peer-reviewed empirical test and no systematic review establishing that it produces better products, better forecasts or better outcomes than the alternatives, and its most famous illustration is an unpublished consulting anecdote with no data, sample or control. Use it as a way of asking better questions. Do not attach a number to it or cite it as evidence for anything.

What there is measurement on is adjacent and more useful. Across 52 product categories in a survey of about 20,000 adults a year, with a median of 18 brands per category, demographic variables improved the model's fit for whether someone used the product category in 44 of the 52 categories. For which brand they preferred, a predictively useful association was found in none of the 52. Demographics tell you who is in the market. They do not tell you why anyone picks you, and the reason they cannot is that the thing which decides that is the outcome the buyer is chasing, which does not appear in any demographic table.

The practical move for the week: take each feature you currently market and finish the sentence "so that you can", twice. The first completion will be another feature. The second one is usually the offer.