What is the Analytic Hierarchy Process?

Rationale

We regularly face decisions involving multiple competing factors that resist easy comparison. Selecting a supplier, prioritising projects, choosing between job offers. These situations share a common challenge: how do you weigh "quality" against "cost" against "risk" in any systematic way?

The Analytic Hierarchy Process (AHP) offers an answer. It's a structured method that helps you break down complex decisions and arrive at a defensible ranking of your options. Rather than relying on gut instinct or getting overwhelmed by competing considerations, AHP provides a framework for working through the problem methodically.

Origins and Development

AHP was developed in the 1970s by Thomas L. Saaty, a mathematician and operations researcher working at the Wharton School of the University of Pennsylvania. His key insight was to combine mathematical structure with how humans naturally think about preferences. Rather than asking people to assign abstract weights to criteria, AHP asks simpler questions: given two factors, which matters more, and by how much?

Since its introduction, AHP has become one of the most widely used methods for multi-criteria decision analysis. It's applied across business, government, healthcare, and engineering. The method is taught in business schools worldwide and features in quality management methodologies including Six Sigma.

The Core Idea: Hierarchy and Decomposition

AHP works by breaking a complex decision into a hierarchy with three levels:

• Goal: What you are trying to decide
• Criteria: The factors that matter for this decision
• Alternatives: Your available options

Instead of comparing everything at once, you compare elements two at a time. This "divide and conquer" approach transforms an overwhelming decision into a series of manageable questions.

Pairwise Comparisons: How It Works

At the heart of AHP is a simple question: "With respect to [this criterion], how much more important is Option A compared to Option B?"

You answer using a scale from 1 to 9, from equal importance to extreme importance. There are mathematical reasons why the scale is 1 to 9 instead of 1 to 10 (stability of eigenvector calculation).

Why pairwise comparisons? Research shows that humans are more accurate when comparing two things directly than when trying to rank many items simultaneously. It's easier to answer "Is safety more important than price, and by how much?" than to assign percentage weights to a list of ten criteria.

You make these comparisons twice: first for criteria (which factors matter most to this decision?) and then for alternatives (which option performs best on each criterion?).

A Simple Example: Choosing a Smartphone

To see how AHP works in practice, consider choosing a smartphone. Your goal is to pick the best phone from three options: Phone X, Phone Y, and Phone Z. You've identified three criteria that matter: Price, Camera Quality, and Battery Life.

Step 1: Weight the criteria

First, you compare the criteria pairwise. Perhaps you decide that Camera is moderately more important than Price (a 3 on the scale), Camera is strongly more important than Battery (a 5), and Price is moderately more important than Battery (a 3). From these comparisons, weights are calculated. In this example: Camera 63%, Price 26%, Battery 11%.

Step 2: Evaluate alternatives on each criterion

Next, for each criterion, you compare how well the three phones perform. On Camera, maybe Phone Y is strongly preferred to Phone X, and so on. This produces a score for each phone on each criterion.

Step 3: Calculate overall scores

Finally, you combine the results. Each phone's score on each criterion is multiplied by that criterion's weight, then summed to produce an overall priority score.

The result is a defensible ranking of your alternatives, with full transparency about how the ranking was derived.

The Consistency Check

One of AHP's distinctive features is its built-in consistency check. Human judgments aren't always logically coherent. You might say A is better than B, and B is better than C, but then accidentally indicate that C is better than A. AHP detects these inconsistencies.

The method calculates a Consistency Ratio from your pairwise comparisons. If the ratio exceeds a threshold (typically 10%), you're prompted to revisit your most inconsistent judgments. This isn't about forcing artificial precision; some inconsistency is natural and acceptable. But flagging significant logical contradictions helps keep your results grounded and defensible.

Where AHP is Used

AHP has found applications across virtually every sector where complex decisions arise:

• Business: Supplier selection, project portfolio prioritisation, strategic planning, resource allocation, vendor evaluation
• Government: Policy evaluation, infrastructure planning, budget allocation, regulatory decisions
• Engineering: Site selection, design trade-offs, technology assessment
• Personal decisions: Major purchases, career choices, relocation

The method scales from individual choices to large group decisions involving multiple stakeholders with different perspectives.

Strengths and Limitations

AHP offers several advantages over informal decision-making:

• It breaks complex problems into manageable pieces
• It combines subjective judgment with structured analysis
• Results are transparent, so you can trace exactly how the final ranking was derived
• It works well for group decisions, synthesising input from multiple stakeholders
• The consistency check keeps judgments logically grounded

However, the method has limitations worth noting:

• The number of pairwise comparisons grows quickly. Ten criteria means 45 comparisons just for weighting, which can become tedious.
• The quality of results depends heavily on thoughtful criteria selection. Poorly chosen criteria produce misleading rankings.
• Best practice suggests keeping criteria groups to around seven items or fewer, reflecting cognitive limits on how many factors humans can reliably compare.

Variants and Extensions

Over the years, several variants have emerged to address specific challenges:

Analytic Network Process (ANP) is Saaty's own generalisation of AHP that handles interdependencies and feedback between criteria. Where AHP assumes criteria are independent, ANP allows for more complex relationships. This is useful when, for example, price affects perceived quality, or when alternatives influence which criteria matter.

Fuzzy AHP combines AHP with fuzzy set theory to handle uncertainty in judgments. Instead of precise numerical comparisons, decision makers can express preferences as ranges (e.g., "between moderately and strongly important"), accommodating the vagueness inherent in human judgment.

AHP-express is a simplified approach that reduces the number of required comparisons, making the method more practical for time-constrained business applications while preserving the core logic.

Ratings mode (sometimes called absolute measurement) offers an alternative to pairwise comparison of alternatives. Rather than comparing alternatives against each other, each is scored against a predefined scale (e.g., Excellent/Good/Average/Poor). This approach is useful when you have many alternatives or when alternatives are evaluated independently over time, such as in ongoing project intake processes.

Getting Started

AHP can be done with pen and paper for simple decisions, but software makes it practical for real-world applications. Tools range from free options like SuperDecisions to purpose-built web applications (such as this one, decisionpoint.io) designed for ease of use and collaboration.

The most important first step isn't choosing software but clearly defining your goal and identifying the criteria that genuinely matter. Time spent on this foundation pays dividends throughout the process.

Conclusion

AHP provides a structured, transparent way to navigate decisions with multiple competing factors. Rather than making decisions for you, it helps you understand and articulate your own priorities, then applies those priorities consistently across your alternatives.

Whether you're an individual weighing a personal choice or a team trying to align on strategy, the same principles apply: decompose the problem, compare systematically, check for consistency, and synthesise the results. The outcome is a defensible ranking that you can explain, justify, and revisit with clarity about what would need to shift.

References