video1.0<iframe src="https://www.loom.com/embed/0fde4e183c4f46a992f79568eaf436a5" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/0fde4e183c4f46a992f79568eaf436a5-2237fd34f5ca2795.gif523.7This Loom demonstrates Timber, a tool that analyzes a brand’s existing content to generate a tailored brand voice guide. The presenter shows Timber working with Ben and Olson sources, including uploading folders and links, and notes that input size matters with a target minimum of about 35,000 characters, such as 33 sources totaling 77,000 characters. Timber reads different file types, strips out non-representative text like navigation, cookie banners, and social comments, and builds guides through eight separate passes in about two minutes. They also explain how Timber was built by training on over a dozen brand guides, testing with brands like Headspace without using their guide, and iterating based on comparisons.