Multi-Criteria Site Scoring: Comparing Evidence and Trade-offs

For this illustrative brief, planning fit has a weight of 40%, access 35%, and physical site conditions 25%. The weights sum to 100%.

Weighted total = planning score × 0.40 + access score × 0.35 + site conditions score × 0.25.

Site B ranks first under these assumptions. Its access advantage offsets its lower planning score. The total is a prioritisation aid, not a probability of permission or an investment return.

Now change the weights to planning 70%, access 15% and site conditions 15%, keeping the scores unchanged.

Site C now ranks first. The recommendation depends on which priorities the team accepts. Record that decision rather than presenting either ranking as an objective answer. If Site C’s access evidence becomes unverified, pause its ranking; do not redistribute the access weight across the remaining criteria.

Download the worked comparison CSV or blank scoring register. The worked file contains values, not spreadsheet formulas, and labels every row illustrative. The blank register gives each criterion space for a score, weight, source, date, reviewer and status.

1. Agree criterion definitions and weights; check the weights total 100%.
2. Record evidence and resolve critical exclusions.
3. Give every eligible site a score on the same scale.
4. Calculate the weighted sum and independently check the arithmetic.
5. Test at least one plausible alternative weighting.
6. Save the brief, evidence and recommendation together.

If you map suitability across a site, use the same scale and legend across options, state the resolution, and distinguish areas without data. A detailed heatmap does not establish detailed evidence.

State the use, development stage and alternatives being compared. A study selecting sites for further investigation differs from choosing a detailed design. Use the same boundary rule, evidence date and criteria across alternatives; record when coverage differs. Do not silently treat an unavailable flood or access result as a favourable result.

Some requirements cannot be offset by a high score elsewhere. Establish applicable safety, access and planning conditions separately with the relevant advisers. A sunny site should not gain permission through arithmetic that compensates for an unresolved constraint. Mark each required condition as satisfied, unresolved or not satisfied for the decision in question.

Define what each score means and whether higher is better. A raw distance, noise estimate and percentage need justified transformations before combination. Avoid double-counting related criteria such as station proximity and a transport score derived from that same distance. Agree weights with the people responsible for the decision and document their reasoning; a total of 100 does not itself validate them.

The reviewed interface exposes weights for solar, transit, green space, terrain, quietness, walkability, flood risk and view quality. It validates their total before requesting a result. That interface does not establish that every source is available, current or suitable for a statutory assessment. Inspect the returned evidence and identify gaps before interpreting any grid or summary.

Change uncertain inputs and plausible weights, then check whether rankings reverse. Examine close results in their original units instead of treating a small score difference as meaningful. Grid resolution describes sampling, not positional accuracy. Report the leading option, decisive evidence, assumptions, unresolved issues and what additional information could change the choice.