9 Best Practices for Prioritization with Analytic Hierarchy Process
If there is something every organization has, it is a long list of "must-have" project requests. The process of prioritizing these projects is usually very political and ineffective.
We’ve seen cases where 30% of the selected projects were obsolete within a few months – they were the wrong projects.
So how do you fix it? One of ways to address this issue is Analytic Hierarchy Process (AHP). In this article we hope to give you a picture of what we see as best practice.
This is just a sub-set of the information contained in our Ultimate Guide to Project Prioritization.
1. Make it easy to request projects
Collect project proposals in a system that can be easily accessed across your organization. The result of project prioritization can only be as good as the best projects that have been requested.
2. Use current projects to discover business drivers
Use your current portfolio (or new projects coming in) to learn about what the key business drivers are for your business stakeholders. Ask project sponsors what they are trying to achieve and look for patterns across the business. Perhaps it's about operational efficiency or winning market share.
AHP is a multi-criteria decision making method, so the main factor that determines its effectiveness is whether you identified the important project prioritization criteria. The business drivers you identify ARE your main criteria for selecting projects.
3. Don’t prioritize every project request
Eliminate proposals that are not ready to be prioritized or shouldn’t be prioritized at all.
For example, you may block projects that are too risky, not aligned with strategy, don’t comply with legal requirements, or are not clearly defined. Similarly, you may need a project because of new legal regulations. You don’t need prioritize it, you have no choice.
4. Agree on criteria!
Get an agreement on the criteria you are going to use to prioritize your projects. If people see that important criteria are missing, they won’t feel comfortable with the whole process. Step 2 is not, on its own, enough.
If you don’t agree on criteria, you won’t agree on the results!
5. Exclude cost from your criteria hierarchy
Once AHP has helped you identify the value of each project, you can then introduce the cost element. Value divided by cost is a really useful value to use when picking projects, for example, as it shows you which projects deliver the biggest bang for the buck.
6. Use pairwise comparisons to prioritize criteria
When you have your criteria agreed it is time to prioritize them. AHP gives you a powerful way of doing this prioritization – pairwise comparisons.
7. Don't use pairwise comparisons to evaluate projects
Avoid pairwise comparisons when you evaluate projects against the criteria. In most cases, you are judging a very specific aspect of each project. For example, your criterion might be "Does it support our Strategy?" and the answer to that is “Yes, lots”, “Yes, a bit” or “No, not really”.
Learn more: When (Not) to Use Pairwise Comparisons in Analytic Hierarchy Process
8. Build alignment
Once you have everyone’s judgments, identify where there is real disagreement.
Once the issue is on the table, it forces everyone to listen to different viewpoints and then explore what the “right” answer is. The net result of this is that executives end up making much more “fact based” decisions.
Because everyone was part of the prioritization process, and because everyone ends up buying in to the priorities that you agree, the projects that are selected get much better support from the whole organization. This is more than consensus – it’s driving alignment across the whole organization.
In fact, many customers tell us that this is the part of the process that delivers the most value. Getting the whole team aligned is worth the price of entry on its own.
9. Choose the best portfolio
The best portfolio is not a set of projects with highest scores as this doesn’t include the importance of synergy. Priorities of projects are important input to create alternative portfolios based on resources and dependencies. The point is to choose the best portfolio, and this is another decision.