# Decision Records Introduction
This section captures our architectural decision records. Michael Nygard explains this well in his original article of why we should do this:
Large documents are never kept up to date. Small, modular documents have at least a chance at being updated [...] Nobody ever reads large documents, either. [...] Those documents are too large to open, read, or update. Bite sized pieces are easier for for all stakeholders to consume.
As we're an early-stage startup and we should always be conscious about how we spend our time, therefore you shouldn't capture every single decisions here. Capture what you think was non-obvious; time and energy spent is a good indicator for non-obvious decisions. We want our decision records to save our time, therefore do this for the future you. We do encourage you to write to think too, so that the decision records you write feel less of a chore, but a result of your thinking.
You'll notice that some numbering starts is set to 0, that's because those decisions are made before we decided to record ADRs. We should capture them when recurring questions are coming up (think F.A.Q.).
See the sidebar for the list of decisions.