You have more data about your customers, operations, marketing efforts, sales process, and other business insights than you ever have. But how do you effectively gather and use all of that data?
One method is to use a data warehouse. A data warehouse is the entirety of a company's information, pooled together from a large breadth of sources and organized to maximize efficiency. For some businesses—often enterprises—with enormous amounts of historical data, a data warehouse is the perfect place to store all past and incoming data.
But a data warehouse is a massive endeavor, and getting it up and running is a serious commitment that will require months of work before you can even start seeing insights.
Before you commit to building a data warehouse, make sure that it’s what you really need.
Many companies could reasonably benefit from adopting a data warehouse.
That is, if it were free.
Historically, a data warehouse was the only option for businesses looking to implement any business intelligence—which is why only enterprise businesses could really accomplish it. Affordability has always been the main barrier to entry.
The building process alone costs an average of $500,000 in startup costs. And once that’s done, you still have to set up a team of dedicated analysts and IT experts to manage it perpetually.
Here are Cooldata’s cost estimates for even a mid-size warehouse:
First off: you don’t always need a data warehouse.
Unless your business is making more than $50 million annually in revenue, it’s unlikely you’re generating enough valuable data that you could justify the cost of a data warehouse.
Here’s what you can do without a data warehouse.
Even without a data warehouse, you can still pull all your data into one place. Instead of connecting your data sources to a data warehouse, you can connect them directly to your BI solution (where your dashboards live). Data is piped in to your BI solution, where your analyst works to build tables, manage transforms, and create datasets.
Data is stored in the data sources themselves, or cached (saved for a short time for quick loading) in the BI solution. You can gather and organize your data with many of the same benefits of a data warehouse, but without the same overhead.
A BI solution is more user-friendly than navigating a data warehouse. Any day.
The benefit of building your own data warehouse is that you can control every aspect of your system, from data processing to storage. The downside is that you end up excluding non-technical users from your system (the people who most often have to work in that system everyday).
Most data warehouses require a substantial amount of training before your team is ready to use them—this is another part of the reason why they take so long to get up and running. Keep in mind that individual members of your team are used to managing data in their own reporting systems, and they’ll be approaching your BI solution with different perspectives.
Instead of a data warehouse, you can use a BI solution that’s built with the non-technical business user and data analyst in mind. These will allow you to build dashboards and manage transforms using simple tools that don’t require any coding.
Pardon Our Shameless Plug: Grow is simple enough that non-technical team members can feel comfortable operating it, and flexible enough that code-savvy team members can apply their own strategies. If you’re super technical, we give you as much power as SQL will give you. And if you’ve never touched a query in your life, there are tutorials and templates to help guide you through.
You can store and query your historical data without a data warehouse. One of the concerns about not using a data warehouse is that you’re going to lose data, or reach a limit on how much data you can query.
Instead of hosting all your data together in a data warehouse (which you have to host on your own server or in a very expensive cloud) you can continue to store your data in their individual sources.
A BI solution that doesn’t use a data warehouse will manage this in a few different ways:
Any of these options are good alternatives for accessing your historical data.
We’re not saying there are not cases where a data warehouse makes sense—there are. It’s just in most cases, you could do better than a data warehouse. In fact, in most cases, BI is a lot easier without a data warehouse getting in your way.
By eliminating the need to build a warehouse, you can instead focus on the crux of your strategy: the actual data.