What to Look for in Tender Comparison Software

Tender comparison software promises to replace the levelling spreadsheet. Whether a tool actually helps depends less on its feature list than on a few fundamentals: can you trust its numbers, can you see where they came from, and does it tell you what it doesn’t know? The checklist below is written so you can apply it to any product during a trial.

1. Every figure should trace to its source

When the software says Supplier B charges $113.64 per tonne, you should be able to click through to the page and line (or sheet and row) the figure came from, and see every adjustment applied on the way. If you can’t audit a number, you can’t put it in front of a client, a project manager or a lender.

Ask: show me the source of this rate, and the steps between the quote and this figure.

2. Units and GST must be handled explicitly

Tenders mix kilograms and tonnes, m³ and tonnes, daily and weekly rates, ex-GST and inc-GST prices. Good software converts exactly where the conversion is exact and asks for an assumption where it isn’t:

Exact: 10,000 kg ÷ 1,000 = 10 t
Exact: $1,250.00 inc GST ÷ 1.1 = $1,136.36 ex GST
Rate: $1,136.36 ÷ 10 t = $113.64/t ex GST
Not exact: 25 m³ → tonnes needs a stated density (e.g. 1.6 t/m³ → 40 t)

Ask: what happens when one quote is in m³ and another in tonnes? Where is the GST treatment of each quote taken from?

3. Unknown costs must stay unknown

This is the test most tools and most spreadsheets fail. When a tender says “freight TBA” or is silent on installation, a blank cell becomes a zero in the total, and the incomplete tender looks cheapest. Software should label the comparison incomplete and, ideally, tell you the break-even:

Supplier A complete price = $1,200.00 + $300.00 freight = $1,500.00 ex GST
Supplier B known price = $1,136.36 ex GST, freight TBA
Break-even freight for B = $1,500.00 − $1,136.36 = $363.64

Ask: what does the total show when one tender has an unpriced line?

4. Matching should be classified, not guessed

Two lines with similar wording are not necessarily the same item: “20mm drainage gravel” and “20mm crushed aggregate” are different products. Look for software that tells you how confident each match is and why, and lets you override a match with a recorded reason.

5. It should compare against your scope, not just against each other

Comparing tenders only with each other finds differences between them; it can’t find scope that all of them missed. If you work from a BOQ or a scope schedule, check whether the tool can use it as a baseline and show coverage per line.

6. Revisions and variations

Tenders get revised. Check whether the software can show what changed between revisions (added and removed items, changed rates, changed exclusions) rather than just a new total.

7. Outputs you can use

Check the exports: a PDF report for approvals, Excel for your own adjustments, CSV for other systems. Make sure assumptions and open questions are included, not just the winning price.

8. Data handling

Tender documents are commercially sensitive. Ask where documents are stored, who can access them, whether files are ever publicly reachable, whether any AI provider receives the text, whether your documents are used to train models, and how deletion works.

How QuantDoc answers this checklist

For transparency: QuantDoc references each extracted figure to its page/line or sheet/row, records every adjustment, uses deterministic decimal arithmetic, requires explicit assumptions for cross-dimension conversions, keeps unknown costs unknown with break-even thresholds, classifies matches, and exports PDF, Excel and CSV. Scope/BOQ baselines and revision comparison are on the Professional plan. Scanned PDFs have limited support. Apply the same checklist to it as to anything else. The example below is real engine output.