What Is Data
Analytics, Exactly?
At the lowest level Data Analytics begins with raw data collection; at the highest, its about structuring discussion of business heath, piecing together accurate assessments, and getting to actionable recommendations.
Book a Consultation
Sometimes a problem or question just has too many pieces for a BI team, or it requires advanced methods, or it’s very murky and unstructured at its beginning. Data analytics teams can lead these longer analyses, and the results and recommendations may shape company strategy for years to come.
AB tests have many limitations, especially for operational adjustments (pricing, manufacturing process changes, etc) and multi-touch marketing campaign adjustments. More advanced experimentation methods are often required.
The data team is in a spot where at least one team member can specialize in supporting a single department. Their “solid line manager” is on the data team, their “dotted line manager” is the stakeholder department lead. The central data team continues to develop/maintain foundations and best practices which the embedded team member fully adheres to and may also contribute to.
Robust and complete frameworks are accepted across the company as the most accurate and most succinct quantitative picture of business/initiative health. The most basic forms of these can include retention, cross-selling, and LTV frameworks adapted for business specifics. We work with a lot of “category defining” companies where the correct answer isn’t yet known.
Using a BI tool (ex: Looker, Tableau), some business stakeholders are creating custom reports/dashboards and digging a layer or two into metric shifts independently. Every business stakeholder feels comfortable opening a relevant dashboard and interpreting it at a high level. BI operators are not making wrong decisions based on erroneous data interpretations or practices.
Operators are using data frameworks - consisting of guiding metrics and a syntheses of contributing factors - to make short-term decisions.
Data team partners with (a) biz teams to standardize company-wide definitions, calculations, and basic interpretations and (b) SWE to establish process guidelines for new work that touches data collection or sharing.
A core data analytics warehouse (ex: Snowflake, BigQuery, Redshift) is receiving raw data from all data collection sources. A central data function owns organization and stitching of this data into initial data models. Practices are implemented (ex: automated alerting, analysis verification processes) to increase trust and depersonalize remediations.
Data is generated and shared across various tools and platforms, both internally built and otherwise. Challenges in ensuring complete, accurate, and fully integrable data collection and sharing can be considerable.
Deep Dive Analysis
Sometimes a problem or question just has too many pieces for a BI team, or it requires advanced methods, or it’s very murky and unstructured at its beginning. Data analytics teams can lead these longer analyses, and the results and recommendations may shape company strategy for years to come.
Advanced Experimentation
AB tests have many limitations, especially for operational adjustments (pricing, manufacturing process changes, etc) and multi-touch marketing campaign adjustments. More advanced experimentation methods are often required.
Dotted-line Embedding
The data team is in a spot where at least one team member can specialize in supporting a single department. Their “solid line manager” is on the data team, their “dotted line manager” is the stakeholder department lead. The central data team continues to develop/maintain foundations and best practices which the embedded team member fully adheres to and may also contribute to.
Higher-Level Frameworks
Robust and complete frameworks are accepted across the company as the most accurate and most succinct quantitative picture of business/initiative health. The most basic forms of these can include retention, cross-selling, and LTV frameworks adapted for business specifics. We work with a lot of “category defining” companies where the correct answer isn’t yet known.
Initial Self-serve and skills transfer
Using a BI tool (ex: Looker, Tableau), some business stakeholders are creating custom reports/dashboards and digging a layer or two into metric shifts independently. Every business stakeholder feels comfortable opening a relevant dashboard and interpreting it at a high level. BI operators are not making wrong decisions based on erroneous data interpretations or practices.
Operational Metrics and Lower-level Frameworks
Operators are using data frameworks - consisting of guiding metrics and a syntheses of contributing factors - to make short-term decisions.
Stabilization
Data team partners with (a) biz teams to standardize company-wide definitions, calculations, and basic interpretations and (b) SWE to establish process guidelines for new work that touches data collection or sharing.
Centralisation, Reporting, and Best Practices
A core data analytics warehouse (ex: Snowflake, BigQuery, Redshift) is receiving raw data from all data collection sources. A central data function owns organization and stitching of this data into initial data models. Practices are implemented (ex: automated alerting, analysis verification processes) to increase trust and depersonalize remediations.
Raw Data Tooling
Data is generated and shared across various tools and platforms, both internally built and otherwise. Challenges in ensuring complete, accurate, and fully integrable data collection and sharing can be considerable.
