Comparison
Spreadsheet invoice audit vs. a matching engine
A VLOOKUP audit finds your worst supplier once. Here is exactly where it stops working, and what has to exist for the audit to survive month two.
The spreadsheet audit is a good first move
Export a month of invoices, export the price list, VLOOKUP the SKU, subtract the two prices, sort descending. That takes an afternoon and it will find your worst offender. If you have never audited anything, do this before you buy software. The question is what happens in month two.
Where it breaks
- Dated prices: a VLOOKUP returns one price per SKU. Contracts change mid-period, so half your comparisons are against the wrong version and the resulting 'variances' are noise.
- Unit of measure: per-square against per-bundle produces a variance in the hundreds of percent. You either hand-fix them or learn to ignore the top of the list, which is where the real errors are.
- SKU drift: distributor codes, pack suffixes and renumbered parts all return #N/A. Unmatched lines pile up and nobody has time to triage them, so they get deleted.
- Credits and rebills: a correction pair reads as one large overcharge and one large undercharge. Both are wrong.
- No memory: last month's judgment calls are not carried forward, so you re-decide the same lines every cycle.
- No lifecycle: the spreadsheet tells you what you found, never what you recovered.
Side by side
- Price lookup — spreadsheet: current price only. GuardianPrice: version in effect on the invoice date.
- Unmatched lines — spreadsheet: #N/A. GuardianPrice: a reason code per line, ranked by cost.
- Units — spreadsheet: manual. GuardianPrice: per-supplier alias table with explicit conversion factors.
- Credits — spreadsheet: counted twice. GuardianPrice: paired with their rebills before classification.
- Setup per file — spreadsheet: every time. GuardianPrice: column mapping saved per supplier.
- Recovery — spreadsheet: none. GuardianPrice: claim lifecycle from open to credited.
When the spreadsheet is still the right answer
One supplier, a price list that changes once a year, consistent units, and under a few hundred lines a month. Below that threshold the effort of loading data outweighs what an engine adds. Above it, the manual audit stops being run at all — which is the real failure mode, not inaccuracy.
Common questions
- Can we keep using our spreadsheet alongside it?
- Yes. Variance results export to CSV and Excel, so anything you build downstream on the numbers keeps working.
- How long does the first import take compared to a spreadsheet audit?
- The first supplier takes roughly the same afternoon, because you confirm the column mapping once. Every file after that loads against the saved profile.
- Do we need clean data first?
- No. Messy data is the reason the reason codes exist — they tell you which data problems are costing money so you can fix those first and ignore the rest.
See your own numbers
Audit your first 500 invoice lines free. No card, no sales call.