Predictive Project Analytics

I found this two interesting articles that were written by Deloitte Consulting firm about Predictive Project Analytics.   Predictive Project Analytics  (PPA) refers to the quantitative tools and techniques organizations can use to help properly manage project risk and realize the highest value return from large and complex projects.  It uses a qualitative, data driven approach to assess projects instead of an intuitive or benchmarking approach. PPA predicts the likely success of a project through predictive analysis of key project and organizational attributes. PPA can help to identify a floundering project and to avoid the costs of a late or poorly delivered project or an outright failure.


To break it down in simple terms PPA is an analytical, data driven way an organization can assess a project, weigh all the risk and determine if the project should continue, if the project needs to be modified or if the project should be terminated.  Having a process to determine risk of a project is critical in today’s business environment because recent research shows that 63% of companies have experienced a recent project failure.  Research also shows that projects that do reach completion, nearly half of the projects fail to meet time, budget and scheduling goals.


After reading both articles, PPA seems like it is designed for and it is very beneficial for large scale, multimillion-dollar projects.  This makes sense because PPA was developed by Deloitte and Deloitte is a consulting group that works on large-scale projects.  When I was reading I was thinking how PPA could be applied to small-scale projects that most of us work on?  The backbone of PPA is data collections and analysis to determine how to better execute a project or if a project should be terminated.  But with small projects that are a one-time thing, how do you collect data and how can you justify the cost of collecting the data that might never be used again.  With my experience working for a small company I think data can be used with limited value on a project but people’s experiences is the best resource for projects and project management.  Use what has been learned and has been successful in the past and modify and apply to future projects.

4 thoughts on “Predictive Project Analytics

  1. A hybrid approach may be worthwhile in a small company. Develop a set of targeted metrics that you can report on as part of your PMM (project management methodology) and hold projects to report on these metrics to create a score card approach. Several key metrics (in addition to the obvious ones like budget, schedules etc.) can be tracked to give good insight to managers to determine what projects are at risk. Some of these include tracking to baselines, number of changes to the project, completion of key documents (project charter, scope statement), resource utilization, and several others.

    The upfront investment in building this out, even in a small company, makes a lot of sense – and you can probably do it on your own a heck of a lot cheaper than a big four consulting firm could

  2. I also think that a hybrid method should be used to improve projects at a small company. While the collected years of experience amongst company employees is absolutely valuable, at the same time it can cause a company to fall into old habits and lose out on innovative new ways to complete project. I definitely believe that data should be kept in some shape or form of each small project, but not so extensively that it excessively cuts into budget or time. The data can then be reviewed at a later date to see trends that might have been missed. A lot of times numbers look completely different when charted. I think all the metrics mentioned by jessew599 are great benchmarks.

  3. In looking at this article, it’s clear that in order to run predictive project analytics you have to collect alot of historical data (article stated 2,000 projects). Companies that don’t have a project management service offering will likely not spend the time, money, and resources on collecting and investing in a database for this information.

    Absent the funding and technology of a PPA system in an organization, risk analysis and planning is key. Identifying the risks and having a plan in place to address those risks if they happen, is the low-tech version of a PPA system.

  4. I agree that PPA doesn’t seem to make a ton of sense unless a firm has incredible amounts of proceess / project expertise already established. What I do wonder, and what the article doesn’t explicitly call out, is whether any of the standards learned from other companies is appropriate. So for example, is the base PPA system adaptable to others purely based on stastistical analyses? A company wouldn’t necessarily have to perform all functions alone – they would just need to perform the base transactions.

Leave a Reply

Your email address will not be published. Required fields are marked *