Scoring scheme for more complex vendor RFP responses and comments
Despite requesting a particular format of response to your RFP, you may well receive a wide variety of responses and comments from vendors. Scoring these can be complicated, especially if there is no standard format. The ideal solution would be to specify and receive standard format responses. The next best - is to create a scoring scheme that categorizes the responses / comments into various categories that are useful to you. Then convert the vendor’s RFP responses into categories and associate each with a numerical score. The example below illustrates the potential solution.
The problem - a typically wide variety of vendor RFP response comments.
The functionality requirement X is achieved by:
- Standard software AA version 1103
- Standard software AA v1103, via tailoring screen configuration and report configuration
- Using windows capabilities
- Using / integrating with software BB
- This could be achieved subject to a full specification, however, it is believed that CC tools may be utilized / interfaced to fulfill the requirement
- Standard software AA version 1104, when released
- Standard software AA will be offering this feature in a future release
- A future release, further discussions are required
- Standard software AA could offer this functionality, subject to a full specification, if customer is prepared to part sponsor this
- Possible modifications required, pending further discussions
- Modifications required subject to a full specification
The solution – is to either separately categorize the responses/comments and then score these, or combine the categorization and scoring eg by creating a table of RFP Response Categories and Associated Scores (with categories / scores that are useful to you), and then convert the vendor RFP responses/comments into a score.
On previous pages we have suggested a simple scoring range from 0 to 3 eg 0 = not met, 1 = partly met, 2 = fully met, 3 = exceeded expectations. But as the responses / comments are more complicated, you could use a wider scoring range say from 0 to 10 (with 10 the best, 0 the worst)