Comparing bids

When you have multiple bids analyzed for the same procurement, you can compare them side by side.

The Compare & Export view

Open Compare bids in the sidebar. The Compare & Export page shows all analyzed bids, with a quick overview of which vendors have the fewest issues and the highest scores:

• Vendor name and overall recommendation
• Finding counts (Critical/Major/Minor per vendor)
• Criterion scores (if you configured evaluation criteria)

The view has three tabs, each showing a count:

• Criteria — each vendor's score against your evaluation criteria
• Requirements — every approved RFP requirement against each vendor (see below)
• Lots — per-lot offered terms and compliance (see lots)

Bid scores

At the top of the Criteria tab, Bid scores turns the tender's own scoring rules into points for each bid, then shows a Total, a Rank, and whether each bid passes the Gates. Click any cell to see how it was scored.

Bid scores appear in procurements you run as the buyer. They don't appear in a bid workspace where your company prepares its own bid.

Where the rules come from

The scores use the scored award criteria of your approved evaluation scheme (see evaluation criteria). The AI reads each criterion's scoring method from the tender text and proposes a rule, for example "Lowest value gets the most points", "Points by range", or "Points per element". Reading the rules uses credits.

The rules are read automatically once the procurement has scored criteria, at least one bid, and credits. Click Read scoring rules (or Read scoring rules again) to start it yourself. It takes about a minute. If the AI couldn't read a rule, you see Couldn't read the scoring rules: enter the rule yourself, or click Try again.

Confirm or edit a rule

Click a criterion name to open its scoring rule. You see the method, the tender text it came from, and its status: rule proposed, not confirmed, rule confirmed, or no rule yet.

• Confirm rule accepts the proposed rule as it is.
• Edit rule lets you set the method and its values yourself: the figure to read from each bid, its unit, currency and VAT basis, ranges or elements, rounding, and Minimum points when the tender excludes an offer below a minimum in that criterion.

Saving a rule recalculates every bid. If the tender text changes later, the rule shows tender text changed, confirm the rule again.

Figures and points for each bid

Click a bid's cell to see the figure found in the bid, the calculation, and any problem, such as a missing currency or a figure for a lot the bid doesn't take part in.

• Enter the figure — when the bid states the figure but it wasn't read correctly, enter it yourself. It must be a figure that appears in the bid text shown. If the bid states no figure at all, ask the bidder for a clarification.
• Accept figure — accept a figure the check couldn't verify.
• Confirm N points — accept proposed points, for example for a criterion the evaluators judge.
• Set different points — override the calculated points. You must give a reason. Reset to calculated points removes your override.

Every decision is listed under Decisions in the cell and recorded in the organization's audit log under Bid scores.

Gates, minimums, and lots

• Qualification gates are pass/fail criteria marked Must pass. A bid that fails one is excluded from the ranking.
• If a criterion has minimum points, a bid below the minimum is excluded too.
• In a procurement with lots, scores are calculated per lot. A bid not taking part in a lot shows Not bidding.

Preview and final ranking

The ranking is Final ranking only when every score in the lot is final. Until then it reads Preview ranking: some points are not final yet, and each missing or provisional score says what is needed. If the scores can't be calculated automatically, for example because weights don't match the maximum points, the lot says why.

Comparing criterion values

This is where evaluation criteria really pay off. If you set up criteria before running analysis, the comparison shows each vendor's extracted value side by side:

At a glance, you can see where each vendor excels and where they fall short.

Requirements matrix

The Requirements tab lays out every approved RFP requirement as a matrix:

• Rows: each approved requirement
• Columns: each vendor/bid
• Cells: a compliance status — Compliant, Partial, Non-compliant, Unverified, Not applicable, or Not analyzed

This complements the Criteria tab (which shows scores) and the Lots tab (which shows per-lot terms) — same vendors, viewed through their obligations.

Open any cell to see the evidence behind the status: the offered value, the AI's reasoning, and quotes from the proposal. This makes it easy to spot a requirement that several vendors struggle with, or a single vendor that fails an obligation everyone else meets.

Comparing finding counts

Even without criteria, the number and severity of findings tells you something:

• Vendor A: 0 Critical, 2 Major, 5 Minor
• Vendor B: 3 Critical, 8 Major, 12 Minor
• Vendor C: 1 Critical, 1 Major, 3 Minor

Vendor C looks strongest here, but dig into the actual findings before drawing conclusions. One vendor might have a "Critical" finding for a technicality while another is missing fundamental requirements categorized as "Major".

Exporting to Excel

Click the Export XLSX button on the Compare & Export page to download a complete Excel file. The export includes:

• Overview sheet — procurement details, summary statistics
• Findings — all findings across all bids, with severity, description, evidence, and verification status
• Criterion values — extracted scores and pass/fail results for each vendor
• Comparison — side-by-side vendor data
• Scores — the bid scores per lot: each bid's points, the figure they came from and their state, then Total, Rank and Gates, with the time they were calculated (buyer procurements only)
• Entities — captured people and companies with registry matches
• Eligibility and capacity — every register check per screened subject, with its status, how the subject was matched, the source register, and the date the data was current

The Excel file is self-contained — anyone can open it without needing access to the platform.

The Download Excel button on a single analysis run also includes a Scores sheet for that bid. If the run is not the bid's latest check, the sheet explains that scores are left out.

Use cases for export

Sharing with the evaluation committee. Not everyone needs a platform account. Export the results and distribute to committee members for review and discussion.

Archiving. Keep a record of the analysis alongside your procurement documentation. The export captures the state of the analysis at the time of download.

Further analysis. Import into your own spreadsheets, pivot tables, or reporting tools. The structured format makes it easy to slice the data however you need.

Audit trail. The export documents what the AI found and on what basis. Useful if procurement decisions are later questioned.

Best practices for comparison

Analyze all bids before comparing. Running all analyses first ensures each was evaluated under the same conditions. If you configured criteria after analyzing the first bid, re-run that one so all bids are measured the same way.

Use consistent document quality. If Vendor A submitted a detailed 200-page proposal and Vendor B sent a 10-page summary, the analysis will naturally find more in Vendor A's submission. This doesn't necessarily mean Vendor B is worse - they might have just been less verbose.

Export for team review. Click Export to get an Excel file with all comparison data. This makes it easy to share with colleagues and discuss offline.

What comparison doesn't do

The comparison shows you the data. It doesn't pick a winner for you - and it shouldn't. Procurement decisions involve context that the AI doesn't have: vendor relationships, strategic priorities, budget constraints, past performance.

Use the comparison to organize the data and spot patterns. Use your expertise to make the final call.