Practical Applications and Tools
Project data science is a growing field. After all, we have a lot of data available to us, and it’s become even easier to leverage the insights hidden in the data with AI-powered tools.
There are 2 types of analytics that project professionals should be aware of:
- Descriptive analytics: how to present data in the most engaging and effective way.
- Predictive analytics: how to use historical data to predict future performance.
We think it’s becoming more and more important for project managers to integrate data science into their workflows to improve decision-making and project outcomes. But how do you get started?
Descriptive analytics
It’s common practice now to use graphs, charts, infographics and tables to share information about project performance. Burndown (or burnup) charts are essential sources of information for agile teams, and earned value analysis metrics feed into charts to provide forecasting and trend data for predictive projects.
You can go beyond spreadsheet-created charts by using analytics tools that extract data and present it in dashboards to surface the hidden gems of information that will help decision makers take the right next steps.
Set up dashboards that show what key stakeholders need to know so they have data available in a single click.
Predictive analytics
Research carried out by APM shows that predictive performance of analytics tools out-performs project managers – in other words, the collective and aggregated data is better at predicting what will happen than a human.
However, many project teams don’t yet have access to either the data sets or the tools to run the analytics. We think this will change, and firms working in construction and engineering are likely to have the data to tap into in a more reliable format that knowledge-based industries.
Organizations might have a large number of data points, but only run a small number of large projects, so the data can be unbalanced. Many projects only capture a small number of data points, such as EVA metrics or cost performance. And it can be a manual process to log and record data which can then quickly get out of date.
It’s not all about AI
While AI regularly hits the headlines, project data analytics is about far more than that. It covers automation and workflows as well, some of which might already be built into the project management tools you use. Once you can take the human out of the workflow, you can truly make progress with benefitting from automating routine data analysis tasks.
Data analytics also covers the way software interfaces with other software. This is called Application Programming Interface (API) and many tools have open APIs to enable them to be connected. As data is often held in various systems, project teams and data analysts need to know where data is so they can extract it and blend it to get the outputs they need.
Practical applications
Data science can help with resource forecasting, scheduling, risk management (think Monte Carlo at scale) and more. The more you use data, the more you’ll find reasons to keep using it for project decision making, prioritization and every day management.
Getting started
The first step to getting started is to have some data. You might think that you need a large enough dataset, but you can begin with small data sets and go from there. The data does need to be in a format that can be interrogated, so with common, structured formatting. In itself, this can be a challenge, but tools like Primavera P6 capture lots of data points and can be used as your starting point.
Find out what data already exists in the organization and think about how you could apply it. For example, you could start with making the most of data on a single project. Or you could go enterprise wide but on a single topic, such as cost performance. Start with something that feels useful and manageable and build up from there.
Identify team members who would be the champions for data science in the organization, and get buy in from senior leaders.
Consider what automation you could introduce to reduce repetitive tasks related to data processing, such as logging dates or activities, or creating reports. Pick the most resource-intensive data jobs as candidates for automation.
Upskill the team
People will want to make use of data analytics if they understand the value of it and have the tools to use it.
Think about upskilling the team in analytics software such as the powerful dashboarding components of your project management software, or products like PowerBI or Tableau. Talk about the value of the data and help people see that it’s more than just an alternative to creating graphs in spreadsheets.
If you have someone on the team who has a passion for analytics, they can quickly show the rest of the team the art of the possible. Sometime the challenge is working out what you want, because the tech is there to be able to deliver it – you simply have to imagine it first.
Get the team using smaller reports, or data sets for decision making locally, and then move into putting the case forward for greater investment (of time and resource) in setting up tools that serve the organization more broadly.
Build a strategy
Even if you aren’t in a position to move forward with data analytics n any practical sense yet, you can still put together a strategy. A data strategy sets out how you want to use data and the use cases and systems that will help you achieve that.
Your strategy can cover your approach to legacy systems, governance and ethics, scope and culture. And you’ll want to consider the organization’s willingness to invest as well, both in hardware and software and training.
Think about how you will measure progress against your strategy, and what building blocks form the roadmap to maturity.
This is just the start! Data analytics is a powerful, practical tool that will take center-stage in project management over the years to come. You can get ahead now, by making sure you’ve got the right strategy, mindset and tools available to capitalize on what you can learn from data you already have.