Proof

Every claim you make is discounted by a stranger, and the discount is steepest at the start when you have no reputation to lend it weight. The instinct is to make the claims bigger to compensate, which makes the problem worse — a larger unbelievable claim is still unbelievable and now also suspicious.

What actually converts, roughly in order of strength: a specific result with a number and a name attached, used with permission. A named client. A worked example a reader can inspect — the actual document, the actual before and after. Something you made that demonstrates the skill directly rather than describing it. A relevant qualification. Years of experience, which is the weakest of the lot and the one most often led with.

When you have none of it, which is where everyone starts, there are two honest routes and one dishonest one. You can do the work at cost or free for two people in exchange for permission to describe the result in detail. You can build a public worked example on a hypothetical case, clearly labelled as one, which proves the method even without a client. What you cannot do is invent testimonials, borrow a result you did not produce, or imply a client relationship that does not exist — and beyond the ethics, it is brittle: those claims fail the moment one buyer asks a follow-up question, and the failure is the thing they remember and repeat.