By James Gilleard
How we use our Data Warehouse to make decisions
By
Introduction
We have previously talked about our Data Warehouse (https://blog.e-verse.com/build/e-verse-data-ecosystem) in broad terms. There, we exposed some concepts and explained why it’s beneficial to have a Data Warehouse running.
Since then, we’ve kept evolving our internal infrastructure, connected more datasets, built more reports, and developed a framework for goal setting for e-verse. We’re at the final stages of building a data warehouse: using its data to produce reports displaying KPIs, measure goals, and ultimately make decisions.
We invite you to take a glimpse!
We will look at the components that make up our Data Warehouse and how that empowers our goals framework and overall decision making.
Add Your Heading Text Here
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
The Components
The platforms where data is generated
Some data connectors
A big database with multiple datasets
A data visualization platform that updates in real-time
1. The platforms
We use these day-to-day platforms: HubSpot, Google Workspace, ClickUp, GitHub, Slack, BambooHR, Quickbooks, and others.
The approach many companies have is that people do their daily work, and after they do it, they build reports to show progress in a given frequency to their manager, executive team, or whoever it may be. So, to name one of the many possible examples, an accountant may load up invoices, make payments to vendors, pay the staff for their work on different projects, etc. Then, at the end of the month, build reports for the CFO displaying how much they spent paying vendors on a project. Didn’t they already input this data somewhere in the company ecosystem? We must strongly argue that this is redundant, tedious, error-prone, and unnecessary.
We base our approach on a simple concept: eliminate all redundancy and extra work associated with performance measurement. We stop at “people do their daily work”. If something was input somewhere, at least once (and, ideally, it should happen exactly once), then reinserting that same data somewhere else is redundant and should be avoided.
What does this mean? In the previous example, the accountant loads up invoices and makes payments to staff, vendors, etc. Then, in the accounting system (in our case, Quickbooks), this data is picked up and sent to the rest of the ecosystem automatically, saving extra work and potential errors. Another example: a developer is working on a task (ticket) in ClickUp or Jira. They press play in the task timer, do some commits, push them to GitHub, then press stop. They’ve tracked work done on a task and pushed their code to a repository. From here on, that data is too picked up and sent to the rest of the ecosystem