With the modern data stack, data warehouses for early-stage startups are now easy to set up and cheap to run. And given all the benefits that they bring, there’s no reason your startup shouldn’t have one.

But the term data warehouse can be a daunting one, and we’re not talking about building something expensive and big. We’re talking about building something quickly that maximizes the benefits for startups and minimizes the costs. That’s why we’re calling this early version of a data warehouse for startups the “data shelf.”

And the good news is that you can build your first data shelf in three hours or less!

Why You Need It

Data Warehouses sound scary and expensive. Many startups assume they shouldn’t even think about building one for several years. But you can and should start building a mini data warehouse (aka a data shelf) soon, like probably now.

But is it really for me, you ask? Any company that uses data to make decisions should have a data shelf. That’s our statement, and we’re sticking by it. With the modern data stack, building one is just too easy. And data shelves bring clear benefits, so there’s no reason not to have one.

What benefits, you ask? With a data shelf, startups can:

1. Do all of their reporting from a single app

   • No more opening five different apps to get a wholistic picture of your business

   • There’s no more logging into Salesforce, Hubspot, Google Analytics, etc., to find the data you need, and no more debating who really needs a license to these tools

   • You can use a single tool like Daydream to view and report on all of your data

   
   

2. Break down the data silos

   • Combining disparate groups of data is easy within a data shelf

   • For example, you can combine sales pipeline data with data from your marketing campaigns

   
   

3. Define metrics in a single place

   • Key metrics can be defined clearly for everyone

   • No more frustrating questions about why different reports don’t match

   
   

4. Send data back to important tools

   • If you need important data to be fed back into apps like Salesforce, you can use a reverse ETL tool like Hightouch to send data from your data shelf back to your apps

In the past, executives have had to weigh these benefits against the cost of a lengthy and expensive data warehouse implementation. Enter the modern “data shelf,” which makes unlocking these benefits cheaper and easier than ever.

It’s Really This Easy

Building your first “complete” data shelf will take about 3 hours. That’s right, not three months, not three weeks, or even three days, just 3 hours.

To build your first data shelf, all you have to do is suck the data out of one of your apps and move it to the cloud, where you can access it and manipulate it however you like.

Here’s what you need:

1. Some existing data (or apps you use that have data)

   • For example, data in an application, like Salesforce, Hubspot, Stripe, Quickbooks, etc.

   • Or data in your own database like MySQL or Postgres

   
   

2. An “ELT” tool

   • This is the tool that “sucks” the data out of your apps and moves it to a cloud data store under your control

   • Popular tools that do this are Fivetran and Stitch

   
   

3. A cloud storage tool

   • Popular options include BigQuery, Snowflake, Redshift and Postgres

   
   

4. A data visualization tool

   • We're biased towards Daydream so you can action data with your team ;) — but Looker Data Studio, Tableau, Power BI, etc. all work too

Those are the basic pieces. And with those basic building blocks, you unlock many of the benefits of a data warehouse with minimal effort.

As you need more from your data, it’s easy to add additional tools to get what you need. For example, if you want to send data back to applications like Salesforce, you can add a reverse ETL tool like Hightouch to this stack. Or, if you want to schedule the creation of new data tables, you can add a tool like DBT. These additional tools are easy to add to this stack, and almost all of them are competitively priced. But how easy is this really to do in practice?

A Concrete Example: Salesforce to BigQuery

Let’s walk through with a real-world example. Say you’re working at a startup, and your sales team uses Salesforce. Your CEO is tired of having to log into Salesforce, and you’re tired of grabbing screenshots out of that tool each week/month for regular reporting. Instead of just living with this situation, let’s build a simple data shelf to deliver a great result for you and your CEO.

We’re going to walk through things in a little more detail here, and, as such, we’re going to reference specific tools, but please note that you don’t have to use the specific tools mentioned here. For example, you can use Snowflake instead of BigQuery or Stitch instead of Fivetran. However, the basic steps of what we’re doing and the benefits you’ll gain will remain the same regardless of which tools you use.

You can find more detailed tutorials on the web, but hopefully, this level of detail is easy enough to follow and makes the idea of building a data shelf feel well within your reach.

1. Create 2 Accounts: BigQuery and Fivetran

   1. We’re going to use two tools, Fivetran and BigQuery. Fivetran will let us move data between applications, and BigQuery will give us a place in the cloud to store that data.

   2. Create a Fivetran Account

      1. Creating a Fivetran Account should take about 20 minutes at most.

      2. Signup for a free 14-day trial of Fivetran; details here

   3. Enable BigQuery

      1. BigQuery is accessed through the Google Cloud Console, so we’ll create a project in the Google Cloud Console and then enable BigQuery from within that project. It should take about 10 minutes.

      2. Go to the Google Cloud Console and create a new project

      3. Search for BigQuery within your new project and enable BigQuery

      4. Billing & Pricing for BigQuery

         1. You might have to put in some billing information, but in most cases, Google will offer you some free credits

         2. For more information on BigQuery pricing, you can go here

   
   

2. Connect Fivetran to BigQuery

   1. Now that we have Fivetran and BigQuery, we need to connect them. More specifically, we’ll add BigQuery as a destination within Fivetran. This should take about 30 minutes at most.

   2. You can follow these instructions to add BigQuery as a destination within Fivetran

   
   

3. Connect Salesforce to BigQuery

   1. We now have a tool to move the data (Fivetran) and a destination for that data (BigQuery); now, all we need is some data to move. So, in this step, we’ll use Fivetran to move Salesforce data into BigQuery.

   2. Connect Salesforce to Fivetran

      1. Depending on your Salesforce setup, this will likely be very straightforward, only require a few button clicks, and take about 5 minutes. Note, though, that the sync time (i.e., the time to actually move the data from Salesforce to BigQuery) could be much longer, like 1-2 hours.

      2. You can follow these instructions to add a Salesforce connector to Fivetran.

      3. We also recommend enabling the “quickstart transformations” within Fivetran. That will make the Salesforce data even easier to analyze once it’s in BigQuery.

   3. If you’ve made it this far, you should be “syncing” data from Salesforce into BigQuery. And if you want any other data in your data shelf, all you need is to add another connector to Fivetran. So the data from Google Analytics, Hubspot, Quickbooks, Stripe, etc. can all be moved in BigQuery just by adding additional connectors in Fivetran.

   
   

4. Visualize the Data

   1. Fivetran has helped us move the data from Salesforce to BigQuery. So now we can do whatever we’d like with the Salesforce data, including visualizing it, downloading it, or bringing in additional data sources to join with it. For right now, we’re going to focus on the common use case of visualizing the data.

   2. In about 10 minutes you can connect BigQuery to a data visualization tool like Daydream. Daydream makes it easy to build visuals, do ad hoc analysis, and create regular reports all in one place.

   3. To book a demo of Daydream, click here.

And that’s it. We now have a single tool, Daydream, that visualizes our Salesforce data. We can share a link to any of those visuals with our CEO. And, as we report on more data, say marketing data, we can add that to our Data Shelf and visualize it within Daydream as well.

Still Don’t Wanna Go Through The Trouble?

If this all still seems like too much, don’t worry, Daydream has you covered. In addition to letting you connect to a data warehouse or data shelf, Daydream can support hundreds of direct connections to applications like Salesforce, Hubspot, Quickbooks, Stripe, etc., so you can connect any tool with a single click.

So, if creating a data shelf isn’t for you, try logging in and connecting your app directly to Daydream instead. To book a demo, click here.