FAQ Sections

How should I combine Question and Section Weightings?

Posted in RFP Scoring and Bid Analysis last updated on April 15, 2016

The aim of an RFP process is to reach a decision about which vendor, product or service best meets your requirements. The basic principle of a structured RFP is decomposition - breaking down a big decision into smaller and smaller component pieces by building a hierarchically structured questionnaire. Having done this, and having scored vendor's responses, the challenge is then to re-compose this information into a single answer - i.e. we need to calculate scoring totals.

SupplierSelect's scoring calculations are based on the principles of decomposition:

  1. The RFP questionnaire is structured as a hierarchy of sections, subsections and questions, nested to any depth.
  2. "Siblings" refers to groups of subsections or questions that share a common parent (e.g. questions in section 1.2.3)
  3. Each section and question has been awarded a weight (a raw numerical value)
  4. The relative percentage weight of a section or question is given by (weight / (sum of all sibling weights)).
  5. The maximum contribution of a subsection or question is capped by the weight of its parent section.
  6. The absolute total weight of a question is therefore the product of (relative weight) * (parent absolute weight).
  7. The normalized total score for a vendor is the sum of (question score * absolute total weight).

Normalized scoring aligns with most peoples' intuition. It feels logical that a weight applied to a section caps the contribution of that section (and all subsections) to the overall score.

A longer article describing this approach to calculating total weighted scores is available on the SupplierSelect news site.

Related Articles


Comments powered by Disqus