Add a Questionnaire
Collect structured answers from every bidding supplier and let AI Scoring factor them in alongside price -- so the best supplier wins, not just the cheapest.
What you'll learn
- Understand how questionnaire answers feed into AI Scoring
- Design questions with the right answer types and weights
- Use Show Stopper questions to disqualify non-compliant bids automatically
- Know the rules around weights summing to 100%
Scale The Questionnaire is available on the Scale plan.
Price is one dimension. A supplier who quotes 10% lower but lacks a required safety certification, cannot deliver in your country, or has never handled your order volume is not actually a good deal. The Questionnaire turns those concerns into structured data that AI Scoring can rank alongside price.
When a supplier submits a bid, they answer every question you set. Their answers combine with their commercial pricing into a single score from 0 to 100. You get a ranked list that reflects both what they offered and whether they meet your criteria -- without reading through every bid manually.
Why this matters
Without a Questionnaire, AI Scoring can only compare price, quantity, and delivery. With one, it can weigh compliance, technical capability, experience, or any other factor you define. The weight you assign each question tells the system how much it should shift the score -- a question worth 40% has twice the scoring influence of one worth 20%.
When to add a Questionnaire
Good use cases
You are sourcing a specialist service where supplier experience matters as much as price. You have a regulatory requirement that every supplier must certify compliance with before you can award them. You want to compare suppliers on lead time or warranty terms in a structured way rather than reading through free-text notes. Any of these calls for a Questionnaire.
| Question type | Best for |
|---|---|
| Yes / No | Compliance checks, certifications, binary eligibility criteria. Yes/No questions are the only type that can be marked as Show Stoppers. |
| Text | Open-ended capability descriptions, methodology explanations, references. Scored by AI, not by a fixed rule. |
| Number | Years of experience, number of past projects, delivery lead time in days. Scored against a rule you define (greater than, less than, between). |
| Single Select / Multiple Select | Preferred payment terms, location coverage, service tiers. You define the options; the system scores against your preferred answer. |
| Date | Earliest start date, warranty expiry, last audit date. Scored against a date rule. |
Before you start
- Your request must be on the Scale plan. The Questionnaire step in the wizard will show an upgrade prompt if it is not.
- Plan your evaluation criteria before you open the wizard. Good questions come from knowing what would change your decision, not from thinking of things to ask.
- If you weight your questions, the weights must add up to 100% — or you can leave them all at 0 for an unscored questionnaire. Decide the relative importance of each area before you start adding questions.
- If you want AI to suggest questions based on your request category, have your request type and Scope details filled in -- the suggestion feature uses them as context.
Adding a Questionnaire to your request
Open the Questionnaire step
The Questionnaire step appears after Bill of Quantities. At the top you will see the AI Question Suggestions panel for your request category. You can load suggestions to get a starting set of questions, or build from scratch.
To load AI suggestions, select Load suggestions. A confirmation dialog shows the credit cost before you confirm. Suggested questions come pre-categorised and can be added to your Questionnaire in one click. Show Stopper suggestions are grouped at the top.
Add and configure each question
Select Add Question to create a question. Each question card expands to show four configuration fields:
- Answer Type: choose the format that matches the answer you need (Yes/No, Text, Number, Date, Single Select, or Multiple Select).
- Category: label the question as Compliance, Technical, Commercial, or General. This organises questions in your review view.
- Weight: the percentage this question contributes to the questionnaire score. Weights must total exactly 100%, unless you leave them all at 0 for an unscored questionnaire. A weight tally at the top of the step shows you the running total.
- Scoring Mode: how the answer is evaluated. Text answers are always scored by AI (the system reads and interprets them). Yes/No answers are always scored by rule (matched against your expected answer). Number, Date, and choice types let you choose between rule-based and AI scoring.
- Question text: the actual question the supplier reads and answers.
Use Quick add at the top of the questions card to insert common compliance and capability questions without configuring them from scratch. Quick-add questions come with sensible defaults you can adjust.
Configure the scoring rule or AI guide
For rule-based questions, you tell the system what the right answer looks like:
- Yes/No: select the expected answer (Yes or No). The system scores the highest for the expected answer.
- Number: set a condition (greater than, less than, between, or equals) and the target value.
- Date: set a date condition (before, after, or between) and the target date.
- Single or Multiple Select: add the answer options first, then select which option(s) represent the preferred answer.
For AI-scored questions (all Text questions, and any other type you set to AI mode), add an AI Scoring Guide: a plain-language description of what a strong answer looks like. For example: "A strong answer describes a documented QC process with at least three checkpoints, references to relevant standards, and evidence of third-party audits in the past two years." The more specific your guide, the more consistently the AI scores answers.
Mark Show Stopper questions
Show Stopper questions apply only to Yes/No answer types. To mark one, toggle the Show Stopper switch on the question card after you have set the expected answer.
When a supplier answers a Show Stopper question against your expected answer -- for example, they answer "No" to "Do you hold the required safety licence?" -- their bid is disqualified. It appears in the Bids tab but is excluded from AI Scoring and the ranking.
Mark a question as a Show Stopper only when a wrong answer is genuinely disqualifying -- not just unfavourable. Use it for hard regulatory requirements, not preferences. A Show Stopper that knocks out most of your suppliers defeats the purpose of running an open request.
Check the weight total and save
The weight tally at the top of the step shows your running total. It turns green at exactly 100% and red if you go over. You can advance to the Review step once the total is exactly 100% — or left at 0 for an unscored questionnaire.
Common weight distributions: give the most weight to the criteria that would genuinely change your shortlist. If price matters most and the Questionnaire is a compliance filter, keep question weights modest (15--30% each). If supplier capability is the primary decision factor, let those questions carry more weight.
When the total reaches 100%, select Next to move to the Review step.
What the supplier sees
When a supplier opens your request to bid, the Questionnaire appears as a required section of the bid form. Each question is shown with its category label and answer type. The supplier does not see the weights you assigned or whether a question is a Show Stopper. They see the question text and the answer format, and they must answer all required questions before they can submit their bid.
Changing the Questionnaire after publishing
While the request is a Draft, you can add, remove, and reweight questions freely. After you publish, changing the Questionnaire is handled as an amendment, and suppliers who have started or submitted a bid are asked to review the change. Once the request is awarded, the Questionnaire is fixed.
It is still worth planning your questions carefully before you publish, so suppliers bid against a stable brief.
How it connects to AI Scoring
When you run AI Scoring on the AI Scoring tab, the system calculates two sub-scores for every eligible bid: a commercial score (based on price, quantities, delivery, and Importance levels) and a questionnaire score (based on the answers suppliers provided, weighted by your question weights). These combine into an overall score from 0 to 100 and a rank across all bids.
What good looks like
- Five to eight focused questions usually produce a more reliable score than fifteen diluted ones. More questions with small weights reduce the signal from any individual criterion.
- Every question should be something you would actually read and compare manually if you had to. If you would not, it does not belong in the Questionnaire.
- Show Stopper questions should map to hard regulatory or contractual requirements, not preferences. One or two is usually right.
- Weights should reflect the actual importance of each criterion to your sourcing decision, not be distributed evenly for tidiness.
- The AI Scoring Guide for text questions should describe a strong answer in concrete terms, not vague quality language.
Next steps
Related guides