HelloTrace

To represent a business process using metric trees, we need to capture three distinct types of relationships between metrics: 1) Metrics that are connected to other metrics 2) Metrics that can be segmented into sub-metrics 3) Metrics with undefined or unknown relationships to other metrics We explore these connections here: https://lnkd.in/gJMQXit6

There is a perennial debate whether the data function is overly obsessed with tooling vs their end impact on the organization. This is a false dichotomy. This tension stems from a fundamental insight: analytics has evolved into a dual mode discipline, a fusion of science and engineering. Doing data well is doing science well and positively impacting the organization. But, in practice, this process is deeply intertwined with solving engineering problems. Efficiently building and maintaining data models requires strong engineering solutions. In fact, constructing even stronger engineering abstractions like metric trees can revolutionize data consumption across the organization. More on this here: https://lnkd.in/esw8GtfB

With the rapid advancements in data platform building, it sometimes feels like business consumers are driving a Ferrari to buy milk at the neighborhood store. Are organizations truly leveraging the advancements in our data platforms - the effortless data ingestion, the robust pipelines, and the reliable production of data assets like metrics? If we find ourselves still writing numerous ad-hoc queries, managing a clutter of dashboards, or frequently exporting data to spreadsheets for quick analysis, it’s a sign that our data consumption hasn't kept pace with the sophistication of our data production infrastructure. So, how can we evolve our data consumption? https://lnkd.in/eqqjCrr3

Analytics is a dual mode discipline, a fusion of engineering and science. We need both tools/processes AND creative hypotheses/critical thinking. We examine this in our blog: https://lnkd.in/esw8GtfB

While we may never fully achieve assembly-line efficiency in analytics, breaking down the process into distinct components and codifying them can streamline the work. More on this here: https://lnkd.in/e4U7e_dB

There are 3 distinct types of analytical workflows - React, Explore and Plan - and getting your REPs in is how an organization builds its analytical muscle. React - regular business reviews, pre/post or experimentation analysis Explore - develop hypothesis, explore to validate or refine Plan - run scenarios, forecast, set goals, track progress, reforecast https://lnkd.in/emzZVyEG

Business reviews are surprisingly under-invested, despite being one of the most common organizational rituals. As we embark on a new month and quarter, let's look into the 4 stages of evolution within organizations. Which stage are you at? https://lnkd.in/eMG_smet

There was a time when generating a dashboard was sufficient. Here's a dashboard, there's your metric. Today, our data use cases are increasingly varied and complex. We seek automated monitoring, root-cause analysis, experimentation and segmentation insights—all with self-serve capabilities across the organization. While leading tech companies like Uber, Airbnb, and DoorDash leverage their metrics layers to support these use cases, enterprises of all sizes must also streamline and operationalize them to drive effective data-informed cultures. Vijay Subramanian of HelloTrace joined me on the Zero Prime Ventures Podcast to unpack how organizations are tackling these new data challenges with the metrics layer. It's a simple yet powerful concept: a single layer that defines, calculates and utilizes metrics—ensuring metric consistency org-wide while enabling emergent features like metric trees. This empowers executives and leaders to drive alignment and streamline use cases across data and business teams. We're in the early stages of adoption, with enterprises crafting bespoke solutions to define, manage and operate on their metrics effectively. Vijay, thanks for the knowledge!

Data Engineers -> Kimball Data Consumers -> Metric Trees A simple way to think about metric trees: they have the potential to revolutionize data consumption, similar to how modeling frameworks like Kimball transformed the work of data engineers and BI developers. Read more about this analogy here: https://lnkd.in/eE5KaH9m

What are the two key components of a metrics-first operating system? 1) Metrics Modeling 2) Metrics Applications 1) Metrics Modeling: This refers to modeling the right set of metrics that capture the nuances of the business model. But, it extends beyond just capturing the metric definitions. A robust system has the ability to flexibly compute metric cuts covering a wide variety of use cases. Moreover, the relationships between metrics have to be modeled as a first class citizen. HelloTrace is leading the way in designing and implementing these metric relationships: https://lnkd.in/eDB5xaNs 2) Metrics Applications: Metrics modeling alone serves as just a "back-end." The emerging crop of metric or semantic layers include these modeling capabilities (barring the advanced modeling of metric relationships). Metrics applications that utilize this back-end to streamline workflows for users across the organization are the second crucial piece of a metrics-first operating system. These applications should ideally be low/no-code, empowering users from both data and business teams to extract meaningful insights on business performance and operate effectively. If you're aiming to cultivate a thoughtful, metrics-driven operational culture, you need to consider both of these components as integral to unlocking value.