Hi, thereader.
You mention;
...based on data on the regional economy, the level of product/service penetration and team experience.
The automation of of anything is always based on data acquisition and analysis, so allow me to comment on that and the Devs in the community can comment on the 'how'. Trusting I've understood your question correctly;
At CBF we have a (manually executed) model for evaluating whether a Cooperative organisation is one we think has a good founding or operational 'basis' and so is worth providing funding for. That may seem a little harsh for a social/community enterprise, but there's only so much money to go around so we have to be choosy.
The key to any evaluation model is defining meaningful, enumerable characteristics that can be measured in a way that is meaningful and can be weighted effectively against others, by context if required. Did I repeat measurable and meaningful enough yet? This is critical as many measures and metrics could be recorded but they
mean very little. Meaning being defined as whether the thing being measured shows the project will be a success and achieve it's goals - that are important to you. For example, we can readily measure things such as; Years of experience of each project founding member, Total number of market competitors already in place, Required funding level to achieve stated project goal.
You could come up with hundreds of measures that distill down to a hard figure, group related measures together and perhaps weight them by importance. It may be that Team related scores are more important than say market competitors because you think the market they are targeting is immature and so ready for another competent player. In our case, the
need the project seeks to address will have greater significance than revenue levels or market penetration. There's nothing in this approach that most people don't take already, for say things like buying a new car or an insurance policy. You look at what's on offer, contextualised to your circumstances and conclude that a given policy or vehicle, a home or even a holiday is the best choice on balance for your needs, budget, timescales, etc.
The other aspect of this is sentiment. Sentiment is harder to fix a number on. This is trying to measure other people's feelings, perspectives, views, interactions, responses to something you're evaluating an investment for. Perhaps your hard-number evaluation, weighting and contextualisation gives a few options - which to choose? Now you'll look at what others think about it. This is like whittling down a product on Amazon to a 2 or 3 choices then looking at customer ratings and reviews to make your final decision. You can still enumerate the characteristics of these but they're more subjective.
With all this modelled out, you have a consistent way of gathering 'hard' meaningful data, combine with 'soft' data around sentiment, that can be used to evaluate a project. Then you can hand over this model as a specification of the system you would like coding into at least an automated evaluation, if not an automated data gathering solution.
Hope that helps.
Regards,
CBF
