Write the veto before the score
A required application, approval or installation condition is a screen. A preference score cannot compensate for a failed mandatory requirement. Name the responsible reviewer and source for every mandatory condition. Keep personal preferences such as color variation or willingness to refinish separate from technical acceptance.
Define your scale before naming brands
Write what 0, 3 and 5 mean for each criterion. For appearance, 0 might mean the owner will not accept the sample and 5 might mean it fits the agreed brief. This is a personal rubric, not measured performance. Use the sensitivity table to see whether the conclusion depends heavily on one priority. Preserve the reason for the chosen weight.
Define the deck's job before scoring materials
A shaded family deck, a full-sun rooftop and a busy hospitality terrace do not present the same selection problem. Describe exposure, use, access, appearance preferences and maintenance capacity in ordinary language before assigning scores.
Include constraints that can eliminate a candidate: permitted applications, required documentation, existing assembly compatibility or practical installation access. Keep these separate from preferences such as color or grain. A high appearance score should not compensate for an unresolved requirement.
Separate requirements, preferences and unknowns
Use three categories in the comparison. Requirements need a documented answer. Preferences can be weighted by the owner. Unknowns need investigation. Combining all three into a single unexplained score makes uncertainty look like a measured difference.
| Decision type | Example | Treatment in the comparison |
|---|---|---|
| Requirement | Product suitable for the proposed application | Resolve before selection |
| Preference | Owner prefers visible natural grain | Compare using labeled samples |
| Operating choice | Owner accepts a defined maintenance routine | Review tasks and access |
| Unknown | No evidence supplied for a claimed property | Keep unresolved; request the evidence |
A simple shortlist is often more useful than a complicated ranking. First remove or hold candidates that cannot yet meet the defined requirements. Then compare the remaining choices against the owner's priorities.
Test whether the answer depends on one assumption
If maintenance effort carries the greatest weight, change that weighting and see whether the preferred product changes. Repeat for appearance, installed cost or repair access where relevant. A result that changes with a small adjustment is not necessarily wrong; it tells you the decision is sensitive to that preference.
Do not manufacture precise scores for evidence that is only qualitative. “Owner prefers Sample A” is an honest observation. “Sample A is 23% more attractive” is not a useful technical claim. Keep measurements, documented limits and personal preferences in their own columns.
Write a short selection explanation that someone else can follow: the requirements met, the tradeoffs accepted and the unresolved items that must close before ordering. The maintenance workload guide can help define one of those tradeoffs without reducing it to a vague “low maintenance” label. The comparison should make the owner's choice clearer, not disguise it behind an unexplained total.
A change in decking priorities
A low-care preference initially dominates, but difficult access for future board replacement changes the discussion. Revise the brief rather than presenting the first score as objective truth.
Hypothetical planning example.
Technical context
Evaluation documents have defined product and use conditions. They do not replace the owner’s selection brief.
ICC-ES: Evaluation report program