In our last post, we talked about what it takes to find a great data scientist for your team. Besides the basics of a strong background in math, business concepts and communication, a great data scientist possesses traits like empathy for the customer, persistence, and the ability to think in a systems-oriented way.

Finding a data scientist with all of these qualities can be tough. In this post, we’ll explore what it takes to build a winning data team, and whether it makes more sense for your business to build or outsource one.

Why Hire A Data Team In The First Place?

Putting together a data team for your business must obviously start with the need for one. A simple question to ask is whether data is a core competency that your organization needs to build up – is it a competitive advantage? If so, perhaps it makes sense to hire your own team. But if your business is more focussed on content, services, products and so on, then it might be better to look elsewhere for your data needs.

But it’s not as simple as that. It’s not just hard to put together the team, it can be hard to maintain everything that comes with. So what are the pros and cons to hiring your own data team vs. outsourcing? An important way to think about it is in the context of:

  1. Building a quality data infrastructure, and
  2. Being able to do meaningful analysis on that data.

Good data infrastructure is only the starting point.  It enables you to build other solutions on top of it that provide ways to view, export and analyze the data. Common NoSQL solutions for data include Cassandra, MongoDB, HBase, and Redis. Meanwhile, solutions like Google’s BigQuery and AWS Redshift are common places to outsource data hosting and analysis. But while having a solid data infrastructure or hosted solution is important, leveraging your data effectively usually means setting up a distributed computing solution to run meaningful calculations and get good insights from that data. This helps get a good count on first order data (such as page views and CTR), but its real value comes in being able to answer tougher questions like “are iPhone or Android users more loyal?” Solutions here can range from slow running batch jobs like Hadoop to faster in-memory solutions like Spark.

Pros & Cons: Building Your Own Data Team

So given this, what are the pros and cons of building your own data team?


  1. Having your team in-house makes it more efficient to add new new features, backfill older data and input new information to your data storage system.
  2. Building your own data infrastructure gives you a solution that’s more tailored to your business, and allows you to optimize how data is exported and viewed, making it easier to iterate on solutions.
  3. Having a team in-house gives your business the ability to communicate efficiently, grow expertise from within and dedicate resources to your top priorities.


  1. It takes thorough research and prototyping just to decide which solution is going to best meet your needs. Design is crucial here, including ingesting data from your various services – and all of this is very resource-intensive. It can be easier have external teams answer questions like: What kind of data will you be storing? Does it need to be ingested in real-time or intervals? Do you want super fast reads? Super fast writes?  Both?
  2. Depending on the amount of data you’re dealing with, your team will likely need a cluster of machines for storage. This usually requires setting up master-slave replication for redundancy to avoid potential outages. It also requires maintaining these clusters is both very expensive and time-consuming.
  3. Setting up internal data infrastructures and distributed computing solutions brings with it the constant challenges of ensuring there is enough documentation to find quick solutions to problems, maintenance, and keeping up to date with changes in technology.  
  4. Hiring and maintaining a data team is difficult and expensive – the starting rate to hire a team of data engineers, machine learning experts, and data modelers would cost around $1 million. This doesn’t account for cross-training, skill redundancy, and transitions.

Look For The Team That Best Fits Your Needs

Putting a good data team together either in-house or by outsourcing can only happen once you  understand why you want a data science team in the first place. Thinking about your business’s core competency, resources and ability to maintain a robust data team is crucial to deciding how to bring one to your organization.

At Vidora, our AI, machine learning and data science experts are solely focussed on the task of maximizing ROI for your business by developing solutions to questions like:

  • How can I maximize the value of my customers?
  • Can I build a flexible, long-term customer strategy for my business?
  • How can I predictively measure the impact of changes on the long-term loyalty of my customers?

To learn more about how Vidora’s experts can help grow your business and boost customer loyalty, email us at

Want to Learn More?

Schedule a demo and talk to a product specialist about how Vidora’s machine learning pipelines can speed up your ML deployment and ultimately save you money.