Let's say that I have a table of 60,000 patients, that include columns such as Date of Birth, Primary Care Physician, Insurance Type, Collection Status, and Balance.
If I wanted to do a lookup to find the associated invoice for a particular visit, I would want to first try to match on multiple criteria, but if there wasn't a "perfect" match, I would like to try to match on fewer criteria. For example, if the 60,000 records try to match on Date of Birth, Primary Care Physician, Insurance Type, and Collection Status, and 45,000 get matched, then 15,000 would go to the No Match output. So then I would want to try to do the SAME lookup against the SAME reference dataset for the remaining 15,000 on, say, Date of Birth, Insurance Type, and Balance.
The way I have it set up is with multiple lookups, each no-match output leading to another lookup's input, working downward in terms of relevance. The issue is that it takes so long to perform due to the fact the reference source has around 2,000,000 records, and each of the lookups seems to need to cache that table, even though it is the same for all lookups. How can I "share" that reference table for all of the lookups?