Data Supporting Portfolio Analysis

Valuable Portfolio Analyses are created with accurate data – current project data on resource, progress, risk, value, etc, and historical project data on success and failure (attrition).   Bold portfolio analyses will be challenged by those who defend the status quo.   Such analyses will take the heat if the data is real.

  1. Data Integrity (How to be fooled by Portfolio Data)

Validating and ensuring data accuracy is a large task that needs senior management support.   Without accurate portfolio data, managers are better off relying on their experience and instinct.  There are a number of challenges to Data Integrity.  The portfolio manager needs to watch out for the following.

  • Lack of conventions for data types or confusing data types
  • Changes to data type names will hinder historical comparisons
  • Open text fields for important data that could be standardized
  • Low data entry compliance (= lost data) – tension: data entry by all project leaders vs few data entry experts
  • Version control – make sure all are using the same data set

It is best to have a portfolio manager set up protocols to check data integrity that an administrative assistant could perform on a regular basis.

Data storage is a critical component of data integrity.  The primary data set needs to be held on a secure server, with appropriate backup mechanisms.

Copies of the Portfolio need to made periodically (at least monthly) for the following reasons.

  • If something goes wrong with data storage, then at least you have a copy to work from.
  • Portfolio tracking systems are designed to capture things as they are today, not the way they were yesterday or last year.
  • If you want to compare where you are now to where you were in the past you will have a copy of the portfolio at that time in the past (See Project and Portfolio History). Caution – choose a point in the past where the operating definitions and data types are the same as they are now.  Otherwise, very clearly label the differences between the two data sets.
  • Copies of the portfolio need to be made at regular intervals, e.g. monthly, and stored separately from the primary data set.
  1. Proactive Data Analysis – Having the Data Before its Needed

For an analysis to impact decision making, it needs to be ready when the decision makers will be contemplating options.  That typically means that the relevant data needs to be available before the need for the decision occurs.  If the needed data has never been collected before, it could take weeks or even months to acquire, often by surveying staff across multiple sites perhaps to gather years old information – a scenario fraught with the potential for disaster.

The portfolio manager needs to know what types of data are typically needed for not only routine decision making but also for the one-off decision.

  1. Store Your Analyses for Future Use

All data analyses need to be stored in a logical manner in a secure location that is accessible to all portfolio managers.  When a decision maker has gotten value from a particular data analysis, that decision maker is likely to ask for it again at a later point in time (e.g. next year) to see how things have changed.  Imagine if you can’t figure out how to do the analysis the next time it’s asked for.   A fellow portfolio manager may get asked to perform the same analysis by a different decision maker – it’s critical that both analyses provide the same results.  The portfolio manager that did an analysis may leave the department – it’s critical that others can find the work.  A useful tip is to note the specific data file (version control!) that gave rise to the view in the view itself and store both in the same database folder.  Be sure to save the views that were not used in the final presentation – they may serve a future purpose.

4. External Benchmarking Data

Comparative average performance data across a set of companies can be useful to either support the status quo or provide further evidence for the need for change (if your company is near the bottom in some sort of performance criterion).  Several consulting groups maintain consortiums that share anonymized portfolio performance data (e.g. KMR, CMR, BCG, IMS health).  Portfolio managers are well placed to serve as representatives to such groups.

We have prepared a whitepaper that delves into these subjects in greater depth. To learn more about this whitepaper please go to the section “Whitepapers and Webinars on Portfolio Management from James Samanen Consulting” and look for the whitepaper entitled “Data Supporting Portfolio Analysis in Bio/pharmaceutical R&D”.