projectsandprograms article

Top 5 Questions When Starting With Analytics



  • Many clients and organizations ask us questions about how to start an analytics program or project. This article aims to provide some answers to common questions.
  • The common challenge is what to start on first and prioritizing work from a demand management perspective.  
  • The next challenge is how execution should work and what parts lead to successful implementation.

The world has an insatiable demand for analytics, and the industry has never been hotter. One needs to look no further than the level of activity on LinkedIn and the number of open positions that many companies are trying to fill to see how much opportunity exists in the data analytics space.

With this increased demand for data skills, it often means that many teams will be very new, or at a minimum, have many members that are still new to analytics. This type of team has its own inherent challenges and opportunities, which we will discuss later in this blog.

If you are on the verge of developing an analytics program, this blog post is for you. Whether you are a newcomer to analytics leadership or a seasoned business or analytics veteran, it always pays to reflect on journeys past to get a sense of how to prepare.

The first step is, of course, to take the plunge and start your data analytics journey as at some point, you need to start as most sponsors will not let you sit in planning mode forever. They usually want results.

Once the work begins on your first analytics project, ideas will flow throughout your organization, and many questions will surface around where to start first, who will do the work, and what to expect once a project gets going. I have worked with many teams that have all the necessary tech skills but cannot deliver value to the hands of business users.

As an analytics consulting company, we have seen many different approaches to building analytics projects and programs. The common challenge is choosing what to start on first and prioritizing from a demand management perspective. The next challenge is around establishing processes for successful execution.

We will focus on how to get your first (organizational) analytics project going and how you then take elements from that project and move them to a program level. You or your organization may have run several independent analytics projects in the past but let us consider this the first project that will run under a newfound level of organization-wide support.  

Analytics Projects vs. Programs

So first off, what is the difference between a project and a program?

  • A project represents a single-focused delivery-based endeavour that typically has a defined scope, start date, end date, team, and budget. For example, your first project may be to assemble data to support Inventory analytics.  
  • A program represents a collection of projects with related execution, from development standards to architecture. Programs have a defined start date and may run for a very long time or forever, depending on the program’s characteristics. For example, once an organization establishes an analytics program, it would likely be expected to last forever. It would be responsible for running many projects over time and common structures such as architectural standards, training, and the Center of Excellence.  

Most organizations will start with a project first and then, over time, will begin to develop a supporting program of activities. Therefore, analytics teams should establish an adaptable, iterative team that is willing to learn, iterate, and collaborate throughout the process.

We are seeing many teams formed around leaders inexperienced in the known pitfalls that will occur. They are left wondering how to manage the demand of work and supplies of skills without bottlenecking the process. This blog aims to relieve some of the questions organizations face when starting the first project within an analytics program. Consider it a guide to what’s normal, what’s not, and how to avoid common challenges. 

While this is not a definitive list, these are the top 5 questions companies ask me as they work to deliver their first project and look to lay the foundation for an analytics program.

1. How do I get started with Analytics?

The first step is to define what analytics means for your organization. This definition will help clarify what data is essential for your organization and provide a roadmap to reach a collective understanding of that data, or a Central Version of the Truth.

A strong foundation goes a long way with analytics projects. You can avoid many common problems in legacy data centers by managing expectations from the beginning and taking an iterative approach. 

Here are a few touchpoints to consider when laying the foundation for your data project: 

  • A grandiose plan is not necessary. Prevent paralysis by analysis in analytics projects by choosing a starting point and iterating to reach the ultimate goal.
  • Demand will always be higher than what an analytics team (if one exists) can deliver on. Manage expectations from the start.
  • Analytics is highly iterative and solving one problem can lead to the discovery of many more.
  • Analytics touches everyone. You cannot make everyone happy, and you cannot queue everyone up.

2. Should I build a data analytics team in-house or hire consultants?

The next step in executing an analytics project is to build your data team. All organizations should have some seeds of an analytics team. Depending on how new you are, you may have an entire department built out, a few individuals, or nobody at all.

Occasionally, there is a need to augment the team. You may need to bring in the necessary skills or relieve capacity issues. Regardless of the size or scope of your current in-house team, hiring outside help is always an option.

We’ve seen many organizations take a DIY approach to build their data team. I like to compare this approach to living in a home under renovation. The DIY approach can cause lots of years of pain and money that could have been saved by hiring a professional contractor to assist.

The same goes for analytics. Hiring a professional consulting company streamlines processes by providing expertise and extra horsepower for projects. Consultants can come in and help get things moving along.

At the end of the day, organizations should have their own internal data team. A ratio that I’ve seen work for most organizations is to have consultants comprise about 20-40% of their team. This percentage should fluctuate based on team size, the number of projects, and the total amount of work.

3. What are common problems or roadblocks to look out for?

Many analytics projects run into similar issues. First, demand will always be higher than what an analytics team can deliver on. Once the work begins, we need to consider the different things that are going to happen that may cause problems for the team. Here are some of the most common roadblocks we see:

  • Lack of data quality: This can occur when there is no strategy to deal with quality issues. Businesses will have problems if they have not taken the time to establish a baseline for quality data and determine how to find it. Data quality is an ongoing process, not a one-time activity.
  • Expectation management: It is essential to manage expectations before a project begins to ensure teams understand goals, timelines, and responsibilities.
  • Misaligned or weak requirements: Your requirements must serve an actual business goal or provide solutions  to business problems. Otherwise, you are just completing an academic exercise in data analytics team building. Testing requirements ensure they serve a purpose in the organization.
  • Skills gap: Assess whether your team possesses the skills necessary to do this work. Survey team members on their understanding of Cloud computing, Power BI, and the intricacies of modelling. Many people get into analytics solely because it is in demand and pays well, but they don’t have the background to produce results.
  • Lack of collaboration and communication: Analytics is intrinsically collaborative. Strong collaboration and communication skills ensure a focused team and that everyone is on the same page.

4. What does “normal” in Analytics look like?

Analytics is highly iterative and collaborative. Be prepared to iterate, fail, reset the course, and learn things as you go. Here are a few tips for getting started and what to expect on your data journey:

  • Work in iterations: Be prepared to toss any old legacy data center mindsets. Legacy data warehouses cost tons of money and lead to frustration with IT. Instead of scoping out a 3-year project and hoping to make the markdown the line, make changes as necessary to continuously deliver value.
  • Collaborate: Data teams must be willing to talk to business users to build what they want. Likewise, business users need to be available to tweak and tune as necessary. A strong collaborative environment between IT and business users enables both teams to work their way to success together. 
  • Leadership buy-in: Executive support is critical to the success of any analytics implementation. Business users, IT, and leadership must work together to handle various demands and expectations, thus solving problems and developing solutions.
  • IT/business strains: While it is normal to experience strains between IT and business users, there are techniques to minimize setbacks in developing your analytics program. Start by breaking the ice between business and IT. Help out when necessary by giving advice and teaching business users how to use new tools. Establishing a relationship early on will lead to more willingness to work together. Creating a Center of Excellence to build up a Central Version of the Truth will help with this. 
  • Establish agile management: Improve the team throughput by keeping expectations in check, assigning tasks based on skills and time, and updating processes with quickly evolving tools.

5. What is a center of excellence and does my organization need one?

Center of Excellence (or competency center) is a team of experts who collaborate and communicate to support activities throughout the organization. Centers of Excellence help drive collaboration and communication by ensuring the right people are working on the right tasks at the right time. 

COEs are a program-level component. However, the seeds of it will exist in your very first project. Establishing a Center of Excellence provides a means to set short-term goals for the analytics team and make sure those goals are delivered on time and align with the long-term strategy. This group will help provide breathing room for the team to focus on critical initiatives and prevent minor requests from consuming valuable resources.

A true analytics team enables the business (through all stakeholders) to make impactful data-driven decisions. The best analytics teams empower stakeholders with the right tools, data, and a good story that aligns with business needs.

Remember, done is better than perfect and allows opportunities for further iterations to improve the analytics process. Need help starting the process? We are here to help. Contact us to start building your dream analytics team today.

For data analytics consulting, contact us.

Share this post

Let's get started with a consultation

get started with training