Every tech company, from individual startup to multinational giants is periodically facing this complex decision: choose or prioritize projects. There are usually more opportunities and ideas that available resources and budget.

The so called “product portfolio” or “application portfolio” optimization involves deciding from (or prioritizing) a set of alternatives under multiple criteria. First alternative can present a short time-to-market while the second one can be a very innovative idea with no competition and third one seems to have a great potential. Which one shall you pick?

Analytic Hierarchy Process (AHP) is the perfect technique for this purpose because it uses a rational approach that checks decision consistency. AHP uses computes factors weights and alternative scores from individual pairwise comparisons. Here is a complete example of applying AHP to solve this problem using Welldecided that is freely available on the net. For the impatient here is a 90 seconds screencast:

To try Welldecided on your own projects click here.

We will consider 4 key factors shown in next table. You easily expand or reorganize such factor list for your specific case. An inspiration source can be the book “Will it fly? - How to know if your new business idea has wings ...” by Thomas K. McKnight. Another book about this subject is “Manage Your Project Portfolio” by Johanna Rothman (still to be published).

Our example considers 3 alternative sales (marketing) applications:

1. Small CRM

2. Travel Expenses

3. Prospect evaluator (computes close probability)

A Welldecided model with those predefined factors is available here (requires Java 6 in your browser). If you click on factors tab you see that factors have already been defined and their relative importance assessed (we are going to change it in a few moments).

To start evaluating alternatives choose Decision > Evaluate (F9). Since there are no alternatives Welldecided asks you to enter at least two (you can always add more later on Alternatives tab):

After entering our 3 alternatives the assessment process begins with first factor. You are asked to compare alternatives in pair. AHP uses this information to verify consistency: if A is better than B and B is better than C and C is better or equals to A your assessment is obviously inconsistent!

Welldecided shows undefined comparisons with a yellow background. You assess relative preference of alternatives by dragging slider thumbs. When you do that:

1. yellow background disappear,

2. bottom line displays the textual meaning of comparison,

3. color of alternatives changes to reflect assessment (green is better and red is worse).

Here is the resulting assessment for “Need” factor we possibly obtained by inquiring potential buyers: a small CRM system is more compelling than Travel expenses and Prospect evaluator and Travel expenses is minimally better than Prospect evaluator.

Here is the full set of comparisons:

After entering assessment we get ranking of alternatives: Prospect evaluator gets 1^st place.

To understand what produced such high rating click on “Values” tab and see that Acceptance factor played an important role. Being able to demonstrate in minutes the benefits of application is in fact a key success factor.

The Weights tab shows the weights of each factor, also obtained through AHP algorithm:

To change such weights, choose Factors tab, select root branch (Select best project) and click on dial icon (evaluate this factor) or F5. You will get the usual assessment dialog, but for factors this time:

The same way you can at any moment re-assess any factor by selecting and re-evaluating it. Just remember to look at Inconsistency Ratio (IR) shown on factors tree: it shall be lower than 10% (it is shown in red otherwise). If any inconsistency ratio is higher just re-asses it by fixing pairwise conflicts.

You can obviously add remove and rearrange (possibly in multiple levels) factors and alternatives or use execute Welldecided for any other multi-criteria decision problem.