Modern Data Project Priorities: Team or Architecture?

What is the makeup of a typical modern data analytics team?

How to self-assess what type of team you have?

What platform/tools work best for a less technical team?  For a more technical team?

You’re starting a data analytics project.  You’ve been given a scope and deadline and told to go at it. You’re looking at your team and wondering how it’s going to happen.

Data analytics projects are usually developed to address an internal business need.  The tools and architecture are often dictated by what’s already in house or whatever new and snazzy tool some hotshot has most recently heard about. But is this the most effective way to develop a project of this type?  The problem is that each component of the system used to manage your data solution comes with varying level of difficulty to use and operate. Is it time for you to consider a contrarian approach to building your data analytics program by taking into account the strengths of your most valuable resources first?  Your data analytics team is your most valuable resource in developing any project.  Using their strongest skills and deepest talents will ensure your project is successful.   In many cases it may be easier and more effective to pair the best systems to the team instead of the other way around.

The makeup of a typical data team can range quite a bit based upon different aspects of the team’s organization. Is this team departmental or enterprise?  Is the team’s program heavily supported by the C-suite or not at all?  What roles does the team have filled?  What experience level?  There isn’t an organization on the planet that has every desired role filled with an expert in that area with strong knowledge of every tool available to them and working knowledge of every other team member’s wheelhouse.  Your project may get done faster and better if it includes tools and processes that your team is already familiar with and knowledgeable about.

Another consideration is where the organization falls within its data lifecycle – are they early or late adopters?  How do you know where you are within a data lifecycle?  In 2022, if your organization hasn’t undergone a digital transformation yet (that is, to a cloud platform) whether for enterprise applications or data analytics systems, then it’s safe to say you are a late adopter.

Do not panic!  This just means the constituency of your data analytics team is probably much different than an organization that is further along in the data journey, data lifecycle, or really, whatever you want to call the timeline for updating and modernizing your data analytics program!  Different industry sectors see different rates of adoption.  It’s probably the organizational DNA of an internet startup to be ahead of the curve for adopting the latest and greatest technologies.

Steady state businesses that do not have reason to update technologies as frequently might have a data analytics team which has been in place since Windows 95!  These types of businesses may place greater value on tried and true technology and be hesitant to change.  They may also be intensely sensitive to data security and value consistency over modernity. The data team at this type of organization will look very different than the one at a technology start up.  It will have different strengths and capabilities.  Even if both organizations were attempting to solve the same problem, they could not possibly address it in the same way.

So, this is a time to self-assess.  Take stock of who is on your team.  What are their current skillsets?  What is their and the organizations level of enthusiasm towards change?  Is the team growing or shrinking by attrition or otherwise? What are their current roles? Do you have an enterprise team with full cast of IT roles and support? Or are you a departmental team full of data analysts and report writers?

It matters, but not for reasons you might think.  It matters because a team full of data analysts these days can accomplish just as much as a team of DBAs, network administrators, and the ilk.  Why?  Because cloud platforms and services have abstracted away much of this work.  The key for such a team is choosing the appropriate (mostly) visual tools and services for the job.

If you have a highly skilled but perhaps slightly disenfranchised band of experts then you can afford to get ambitious.  I say disenfranchised because most DBAs these days are not maintaining and tuning databases, only watching hosted, managed services (or even AI) doing their job for them.  Oftentimes, DBAs have skills that lend themselves nicely to implementing a tool that is partly open source or requires more system integration than meets the comfort level for a data analyst.  This will allow you a more expansive selection of ways to solve your business need.

What if there is a mix?  That is, maybe on loan you have part FTE commitment from central IT, and you also have highly skilled local data experts and maybe even a real live data scientist or two?  This is the best of both worlds.  You are not a shadow IT operation, however at the same time in a pinch you can lean on your local talent to find a workaround and/or to pivot fast to get to that solution needed yesterday.   This gives you both breadth and depth in your implementation, allowing a more comprehensive solution or possibly a faster solution.  This team may also have ideas that the business side hasn’t even thought of yet and can drive business performance through innovation.

How enthusiastic is your organization about change? This is very, very important.  Workers often like what is familiar and least risk.  Any big changes, especially all at once, for longtime FTEs can be a recipe for disaster.  Human nature is to sabotage (yes, sabotage) and drag their heels to resist change.  I’ve seen this happen at so many organizations. The fear is job loss, and this is oftentimes a misconception, but not always.  Valued employees need to know both what their role is for a project and what their role will be after the project, especially if the project will change that role.  Their input should be solicited as much as is reasonable, with sufficient reassurance that the organization values them and their contribution.  A team member’s willingness to change can be impacted by retooling, retraining and ownership of the project process.  If your FTEs are accountable and own projects to modernize then you can reduce the fear factor.

Right now when this project is about to be implemented, is your organization scaling up or down?  If you are likely to gain or lose employees during this implementation, it will affect the decisions you make up front.  Any business that is shrinking, and losing employees over time due to cost cutting, might want to consider pay-as-you-go services instead of buying licenses or a chunk of services up front.  Many times I’ve run into small organizations who plunked down $20k for an annual first year contract, but have barely used $1k after several months, during ramp-up.  These organizations could have used their budgets more effectively with a little more foresight.

But scaling up?  Then you want to make sure to accommodate what comes next in your data life cycle.  For example, if you know that your data analytics team will have 6x more people over the course of the next year then why not prepare ahead for business science.  Business science means that you can accomplish some of what a data scientist can do, only with some type of data science tool or platform of services to lean on.  But, you must do some due diligence and allow lead time to get this aspect of your program started.  This would involve training and built out commodity storage (with data) and perhaps even a cloud data warehouse with data as well.  Although these steps require lead time and up front planning, in a growing organization, this will pay off quickly.

After more than twenty years experience implementing data projects of all shapes and sizes across many industries and geographies, what are the top 3 lessons I’ve learned?

  • Teams comprised largely of no-code data analysts should be equipped with visual platforms and tools
  • Teams comprised of highly technical staff will get bored and leave unless challenged with new technology – try open source, try something innovative
  • DO NOT code for code’s sake

Does your organization need some outside help with all of this?  If so, then you are in the right place – that is what we do.  Our goal is to make ourselves obsolete as your FTE staff ramps up to run the show!  We’ve helped to recruit, hire, and train Data & Analytics staffs at small, medium and large organizations.  If you have decided to adopt a modern approach to manage your data and analytics program, then we can help turn your data analyst into your BI Manager; your DBA into your Data & Analytics Engineer; your business analyst into your Business Scientist to manage your data cradle to grave!

Newsletter
Get updates about all latest data (& cognitive) analytics news & information delivered to your email, monthly.