E308 – See how to allow data teams to answer questions and make business-critical decisions in minutes with Narrator, a data modeling platform | with Ahmed Elsamadisi

---

Ahmed explains that data is information and information can be used to make things better. From a marketing background, you would want to know which ad source is bringing consumers to our website and which ad source is keeping them there. The idea of making big decisions with data is not looking at individual customers, but seeing how people behave and what gets your customer to make the decisions they want. It’s about you getting more detailed in making a more specific decision that affects your customers.

It starts with the question that you’re asking and then figuring out how to get there. If you are thinking about investing your time in getting more women or getting more elderly, for instance, Ahmed will tell you the conversion rates and give you a scale of where to dedicate your energy. All good data starts with a good question and a good question leads to a good analysis, which leads to a good decision. You must first have correct and accurate data that you trust and then get down to the individual customer journeys. When they built Narrator, the idea was to build something you can trust and understand, something that makes sense and has an intuitive nature.

When you’re using tools like Facebook Ad Manager and Google Ad Manager, they are lying to you. They’re incentivized for you to give them money. Be aware of the fact that all these different tools are very misleading. Words are very misleading, especially since Facebook is notoriously been sued every single year for lying about its metrics or using words that are deliberately misleading. At Narrator, they try to nudge you into being very clear about your question and building a tool that enables you to get that clarity in the questions and in the answers.

The Narrator standardizes all data as one of its functions. Being able to reason about data by having it in a consistent format with common assumptions. Some factors that could go and mess up your data and AI, causing you to have inconclusive results are bad Al algorithms. Ahmed explains what’s happening in the industry now, that everyone is shoving in machine learning models and hoping it sticks out. The machine learning model is being trained and it’s not understandable. When AI makes a decision; a human being cannot evaluate whether it makes sense or not. The AI then behaves badly, at which point they need to go and train it again. This represents a failure of the data team as well as data science. You have to do good data analysis and use suitable data algorithms in order to base your conclusions on what you are learning. 

In this episode:

[01:46] Ahmed explains how Narrator, a data modeling platform works 

• The Narrator is not just a data platform but helps and guides you to make the best decisions 
• Asking the right question to understand the analysis

06:11] Seeing how data is color coding and ordering an Excel sheet

• Prioritizing what’s important and how to understand the scale
• Where to dedicate your energy

[11:18] Thinking of being statistically significant, but also actionable

• How narrative standardizes all data
• Built a way to capture how a data analyst will deduce a decision

[15:42] Messy algorithms, how everyone shoves in machine learning models, hoping it works

• You need to consistently check if features are still driving the value that you expect

[21:06] Narrative will let you know as it’s making decisions, adjusting to customer’s behavior 

• When you grow and get more customers, how does their behavior change
• Providing speed and quality and being able to answer questions and make decisions

Watch the Interview:

Learn more and reach out:

• LinkedIn: Ahmed Elsamadisi
• Website: The Narrator