The lesson
Explain how the customer tested you before giving the result, and show the savings after subtracting your own price.
Under the headline "How freecode reduced agent costs with Context Highlights" and a date, the first section, "Building a free coding agent", explains what freecode is and why founder Andrew Schmitz cares about cost. A terminal screenshot follows, captioned as freecode building a reading-list app from a plain-English request. "Making web research more efficient" names the prior setup (Exa search and direct fetch), explains why extra tokens raise costs across later turns, and describes the benchmark on replays of real sessions. It then gives the results: input tokens fell 9.8%, modeled LLM costs fell 13.8%, total costs including retrieval fees were 8.9% lower, and the savings covered Context's cost approximately 2.8 times over. "Continuing to push harness efficiency" says freecode replaced its tools in production within a day, then a "Start building with live web context." block offers 1,000 free credits.
Don't copy freecode's token and harness details, the Exa comparison or Context's pricing figures. Keep the order: customer and why cost matters, the old setup and how it was tested, the measured result net of your own fee, then how fast they moved to production.
Tell us what you want to make. We’ll talk through how to put these ideas to work for your brand.