Decision weighting guide

Using AHP-generated weights with option scores

Understand the difference between what matters to your organization and how each candidate performs—and how DecisioQ combines both to produce an explainable ranking.

The essential distinction

AHP-generated weights and user-entered scores are different inputs to the same decision.

InputQuestion answeredApplies toExample
AHP-generated weightHow important is this factor?The criterion across every optionVehicle quality = 30%
User-entered score or valueHow does this option perform?One candidateAuction A quality = 88

AHP weights answer “What matters most?”

Option values answer “How does each candidate perform?”

TOPSIS combines them to answer “Which option offers the best overall balance?”

AHP does not score auction houses, repair orders, or suppliers. It establishes the relative importance of the criteria. Likewise, an option’s high score does not make that criterion more important.

How the calculation works

Business-priority comparisons
AHP criterion weights
Option values and TOPSIS ranking

1. Establish what matters

The client compares every criterion pair using business judgments such as “quality is somewhat more important than cost.” AHP converts the complete comparison set into normalized weights totaling 100%.

2. Describe candidate performance

The client enters comparable evidence for every option. The same criterion weight applies across all options; each option has its own performance value.

3. Normalize values and respect direction

DecisioQ makes unlike units—such as dollars, kilometres, days, percentages, and scores—comparable. It also follows the catalog direction: higher is preferred for a maximize criterion; lower is preferred for a minimize criterion.

4. Apply weights and rank

The showcases use TOPSIS. It considers each option’s closeness to the best attainable combination and distance from the least-desirable combination. A heavily weighted criterion has greater influence, but it does not decide the result by itself.

Worked auction example

Suppose the AHP exercise assigns vehicle quality a weight of 30%, availability 25%, buyer fee 18%, reputation 15%, distance 8%, and title speed 4%.

CriterionAHP weightAuction AAuction BAuction C
Vehicle quality30%888392
Availability25%969076
Buyer fee18%$850$475$325
Reputation15%948791
Distance8%235 km0 km68 km
Title speed4%899674

Auction C leads on quality and fee; Auction A leads on availability and reputation; Auction B leads on distance and title speed. TOPSIS evaluates the complete weighted pattern. It does not merely choose the option with the most first-place values.

Two ways the result can change

Change the weights

Changing AHP judgments changes organizational priorities. If buyer fee becomes much more important, lower-fee auctions gain influence even though their fee values have not changed.

Change an option value

Changing Auction A’s fee from $850 to $500 improves Auction A’s performance. It does not change how important buyer fee is to the organization.

Using AHP in a showcase

  1. Open the auction-house, repair-order, or parts-supplier showcase.
  2. Find How should the factors be weighted?
  3. Select Help me generate weights (AHP).
  4. Complete all 15 pairwise comparisons.
  5. Optionally load the balanced starter judgments and adjust them.
  6. Select Generate AHP weights.
  7. Review the percentages and consistency ratio.
  8. Enter or review each candidate’s performance values.
  9. Select Get Recommendation.

If any priority judgment changes, generate the AHP weights again before updating the recommendation.

Consistency ratio

Pairwise judgments can contradict one another. A client might say quality matters more than cost, cost matters more than delivery, but delivery matters more than quality. AHP measures this with a consistency ratio.

0.100 or lower: accepted by the showcase.

Above 0.100: review and revise the comparisons.

Consistency does not prove that the priorities are correct. It indicates that the judgments are reasonably coherent. The showcase will not use an inconsistent comparison set for the final recommendation.

Request structure

When AHP is selected, the final decision request includes both the priority comparisons and option-performance data:

{
  "decisionId": "AUTO-AUCT-006",
  "weightStrategy": "AHP",
  "ahp": {
    "criteria": [
      "vehicle_availability_score",
      "buyer_fee_level",
      "average_vehicle_quality"
    ],
    "comparisons": [
      {
        "criterionA": "average_vehicle_quality",
        "criterionB": "buyer_fee_level",
        "importance": 3
      }
    ]
  },
  "options": [
    {
      "optionId": "AUCT-NATIONAL",
      "name": "Northline Vehicle Exchange",
      "values": {
        "vehicle_availability_score": 96,
        "buyer_fee_level": 850,
        "average_vehicle_quality": 88
      }
    }
  ]
}

The shortened example shows the relationship between the two inputs. An executable request includes every authoritative criterion, every required pairwise comparison, and at least two options. DecisioQ derives the weights during execution; the browser does not substitute a hard-coded weight vector.

Common mistakes

Using option performance to answer priority questionsAn auction’s superior quality is evidence about that auction. It does not establish that quality is the organization’s most important criterion.
Entering weights as option valuesA 30% quality weight does not mean entering 30 as every option’s quality score. Enter each option’s actual performance.
Reversing minimized valuesEnter the actual cost, distance, or delivery time. DecisioQ already knows that a lower value is preferred.
Mixing unitsUse the same unit and measurement method for every option within a criterion.
Treating AHP as the ranking algorithmAHP generates weights. TOPSIS consumes the weights and option values to produce the ranking.
Assuming one criterion decides the winnerThe highest-weighted factor has greater influence, but the complete tradeoff pattern determines the result.

Recommended client workshop process

  1. Agree on each criterion’s business meaning and measurement method.
  2. Discuss priorities separately from candidate performance.
  3. Have accountable stakeholders complete the AHP comparisons.
  4. Resolve consistency warnings.
  5. Record the source and date of option values.
  6. Review the full ranking and tradeoffs.
  7. Test plausible changes to important values and priorities.
  8. Retain judgments, evidence, assumptions, and approval decisions.

Continue with DecisioQ

Apply this model in an interactive showcase or study the full weighting contract.

Auction showcaseRepair showcaseSupplier showcaseAHP Weight DiscoveryAPI contract