eProcureAI / Resources / Playbook
PlaybookEvery supplier list decays. Duplicates accumulate, documents expire, and nobody owns the cleanup because it is nobody's actual job. Here is a sequence that gets it done in about four weeks.
A supplier list is the least glamorous data set a finance function owns and one of the most expensive to get wrong.
Duplicates cause payment errors and hide consolidation opportunities. Expired documents create exposure nobody knows about. Stale banking details are a fraud route. None of it announces itself, which is why cleanup projects only start after something goes wrong.
This sequence assumes you have a few hours a week and no dedicated resource, which is the normal situation.
Three reasons, all structural rather than anybody's fault.
Suppliers get added under pressure. Somebody needs a purchase today, the record is created quickly, and the details are completed later or never. Do that a hundred times and the list is full of half finished records.
The same supplier gets added twice because a search for their name did not match how somebody else typed it. Acme Corp and ACME Corporation look different to a database and identical to a person.
And documents expire silently. Insurance and certifications have dates nobody watches, so the list is accurate on the day it was built and progressively less accurate thereafter.
Start here because it produces the clearest result and builds credibility for the rest of the work.
Look for three patterns. Similar names, which catches the spelling variants. Shared tax identifiers, which is the definitive test and catches cases the name never would. And shared banking details, which occasionally reveals something more interesting than a duplicate.
For each candidate pair, decide which record survives and why. Usually the one with more complete documentation, not the one with more transactions.
| Pattern | How to find it | What it usually means |
|---|---|---|
| Similar names | Fuzzy match across the supplier list | Spelling variants of the same company |
| Shared tax identifier | Exact match on the identifier | Definitively the same legal entity |
| Shared bank account | Exact match on account details | Same entity, or something worth investigating |
| Same address | Match on address fields | Related entities, or a duplicate with a rename |
Tax identifier is the strongest signal. Name matching alone produces false positives that waste time.
Now that the list is shorter, work out what is missing. For each active supplier, which required documents are absent, expired, or expiring within ninety days.
Do not try to collect these yourself. Send the request to the supplier and let them provide it, because they hold the documents and chasing by email is what made this a problem in the first place.
Prioritise by exposure rather than alphabetically. A contractor working on your sites with lapsed liability insurance matters considerably more than a stationery supplier with an old tax form.
With a clean list you can finally see something useful. Group twelve months of spend by category and look for suppliers doing the same thing.
Three or four suppliers in one category is usually not a strategy, it is an accident. Consolidating creates leverage, reduces administration, and often reveals that you were paying different prices for the same items.
Be selective. The goal is not the fewest suppliers, it is the right number. Single sourcing a critical category to save administration is a bad trade.
The cleanup fails if it is a one off. The last week is about making the clean state maintainable.
Give suppliers a way to maintain their own records, so the data stays current without your team chasing. Put expiry dates on every document with automatic reminders. And name somebody who owns the supplier list, because unowned data decays fastest.
Three habits do most of the work. Check for duplicates at the point a supplier is created rather than annually. Require the documents before a supplier can be paid rather than after. And review the list quarterly against a filter for expiring documents, which takes minutes rather than reopening the whole exercise.
If you do those three things, you will not need to repeat this playbook.
Thresholds that catch real risk without creating delay.
Read it AnalysisThe four places money leaks, with typical patterns.
Read it PlaybookThe most common rollout mistake is a thin catalog.
Read itWe will run duplicate detection on it live and tell you what consolidating would be worth.
Book a demoBack to all resources