Google does not fetch every URL on a company site at the same rate. Crawl budget is the set of fetches the bot will spend in a period. Faceted filters, session IDs, empty tags and copies of the same list spend that set. The service URL where a buyer would enquire ends up at the back of the queue.
Budget is not a page count
A site with a thousand URLs is not more visible than a site with a hundred. The bot follows status codes, duplicates and the addresses your internal links point at. When a board asks for more pages, the queue often gets longer. The record a buyer asked for does not move forward.
Which URL class goes first
The class that should be fetched first is narrow: a service page that answers one commercial question, a product record that can take an enquiry, and the office and contact pages. The class that should wait is also narrow: sort parameters, internal search results, session URLs, tag pages that repeat the same text. In discovery, HED Creative writes those classes. There is no line that says to crawl the whole site.
What the console and the log show
Search Console crawl stats and, where you grant access, the server log show which class consumed the fetches. The figures are not invented. If the property is yours, the data is yours. Without access, the decision starts from the quality of the URL list. A made-up crawl percentage does not enter the report.
The decision is not a position
Moving fetches onto enquiry URLs does not buy a ranking. Google fetching a URL and a buyer choosing it are separate jobs. The second depends on the answer on the page and on the intent map. It cannot be read until the first has happened.
Discovery needs a URL list or a sitemap, the names of pages that take enquiries, and crawl stats if you have them. Without access, the list is enough to start. No ratio is invented.
A written view of which class comes first sits in the SEO agency scope. If the URL structure itself has to be rebuilt, website development belongs in the same brief. Related decision: index coverage.