The Narrator team in New York City in a socially distant group photo.

In my 25 years working on product and design teams, one of the most vexing issues has always been how to get the most accurate and useful data to help make business decisions. Whether you are an early-stage founder seeing product-market fit or an established Fortune 1000 executive optimizing your marketing spend, the difference between success and failure is your ability to understand and believe in the data that your systems provide.

Why is this core aspect of every business so difficult (and expensive) to get right?

  1. Data is in a variety of systems and isn’t standardized.
  2. A massive industry of business intelligence tools has evolved to help structure data. These all rely on a star schema to create relationships between data in these various systems.
  3. Humans are required to model this data. There’s so much complexity that it takes specialized data scientists and engineers to create and maintain data models (based on the star schema) so that data analysts can manage and respond to how the data is constantly changing and evolving.
  4. As the number of data models increases, it becomes impossible to maintain a single source of truth across business concepts like leads, sessions, completed orders, etc.

Modern data teams spend a disproportionate amount of time wrestling with these issues — essentially building a skyscraper on top of quicksand.

Meanwhile, despite a predicted $77B market by 2022, current solutions are focused on showing dashboards that still require interpretation and don’t actually provide answers to the underlying business questions. Consequently, I was excited to meet the Narrator team and learn about their innovative new way of looking at how data analysis can and should be done.

Narrator was founded by Ahmed Elsamadisi, Cedric Dussud, Matt Star and Mike Nason, all of whom were previously at WeWork. Before co-founding Narrator, CEO Ahmed Elsamadisi built WeWork’s data system and experienced the aforementioned problems of data modeling. He and his 45-person team leveraged all the latest tools but still struggled to quickly answer key business questions.

Out of these frustrations came the idea for Narrator: A universal 11-column data model that ensures a company’s data is always up-to-date. The data model is a union of templated queries, each representing one customer action and is made up of the Narrator Engine and the Activity Stream. The Narrator Engine keeps its Activity Stream data model up-to-date in a company’s data warehouse while the Narrator Platform allows data teams to generate tables (for analysis and reporting) using the Activity Stream. This delivers a single source of truth for data analysis combined with simple traceability and minimal maintenance. Learn more about Narrator and its 11-column data model (just some of the questions Narrator can help a company quickly and accurately answer: conversion rate by ad source, support tickets that lead to churn, visitor to sale conversion, time between purchases, and many more).

Image for post
The Narrative format

Narrator also takes this a step further with its Narratives — action-focused data analyses in a story-format. Business and data users can see clear goals and recommendations based on data. Instead of relying on dashboards for interpretation, Narrator’s action-oriented Narratives provide human-understandable reasoning and context behind the data that is needed to make important marketing, sales, product, and operations decisions.

Because of this, Initialized is pleased to lead Narrator’s $6.2M Series A and join continuing investors Flybridge Capital Partners and Y Combinator in this new round. I’ve joined the board and am excited to work with this amazing team of builders to make it faster and easier for companies to understand their data and their customers. The first step in this journey is the release of their self-service data product, which any data person can sign up for and try it.