Balancing Tactical and Strategic tasks as a Data Analyst | by Sandeep Uttamchandani | Jun, 2022

How to maximize business outcomes without burning

In linking data to business outcomes, Data Analysts play the most critical role. They are typically the only people on the Data team that understand both the data context as well as the business context.

In my experience, on a typical day, a Data Analyst will get pulled into multiple ad-hoc tactical requests by the business:” Can you send the 2-year trend for our product CAC by this afternoon.” While these requests are a distraction from the strategic roadmap tasks, a subset of these can be critical for business outcomes. Also, a good percentage of these requests tend to be repetitive or debugging oriented since the “CAC values in the Tableau dashboard seem to be different from the quick email report.” Some of these ad-hoc requests may need much more than just a “few hours of work” and need to be planned.

As a data analyst, how to balance the time across tactical requests and strategic roadmap activities with the goal to maximize the impact on business outcomes? Among the tactical tasks, how do you separate the wheat from the chaff and address ones with higher business impact?

This framework is based on my experiences leading data analyst teams. Without the framework, everything becomes a priority, and can lead to Data Analyst teams working 24 x7 before finally burning out!

Let’s start off with a sample of typical tasks of a data analyst:

• Building reports/dashboards/notebooks for analysis
• Supporting ad-hoc requests for reports/analysis
• Analyzing the data/Explaining the “why”/storytelling for the business team to make insights actionable
• Understanding business requirements and context
• Debugging business logic inconsistencies across reports
• Debugging data quality and pipeline issues/Debugging data models/Dashboards
• Creating/updating dashboards
• Standardizing business logic for the metrics
• Improve data literacy within the business team
• Finding data sources, metadata details, and data documentation
• Preparing/wrangling data/verifying governance
• Change management related to datasets and business logic
• Operationalize data models/views/data apps/dashboards
• …and more…

Consider diving these tasks into a 2 x 2 matrix of Impact (on business outcomes) and Complexity (to technically accomplish the task).

• High Impact and High Complexity (Bucket 1): Manage as a roadmap
• High Impact and Low Complexity (Bucket 2): Manage as tactical backlog with rapid turnaround
• Low Impact and High complexity (Bucket 3): Examples include data quality, dashboard inconsistencies, pipeline performance, etc. Manage using the emerging ecosystem of tools in the Modern Data Stack.
• Low Impact and Low Complexity (Bucket 4): Manage as self-service tasks done by the business team (with guidance as required)

Revisiting the task list with the context of the 2 x 2 matrix:

• Building reports/dashboards/notebooks for analysis — Bucket 1
• Supporting ad-hoc requests for reports/analysis — Bucket 2 (needs to be further analyzed — more details later in the blog)
• Analyzing the data/Explaining the “why”/storytelling for the business team to make insights actionable — Bucket 1
• Understanding business requirements and context — Bucket 1
• Debugging business logic inconsistencies across reports — Bucket 2 or 4
• Debugging data quality and pipeline issues/Debugging data models/Dashboards — Bucket 4
• Creating/updating dashboards — Bucket 2 or 3
• Standardizing business logic for the metrics — Bucket 1
• Improve data literacy within the business team — Bucket 1
• Finding data sources, metadata details, and data documentation — Bucket 4
• Preparing/wrangling data/verifying governance — Bucket 4
• Change management related to datasets and business logic — Bucket 2 or 3
• Operationalize data models/views/data apps/dashboards — Bucket 4

The time spent on each of the buckets varies based on the overall analytics strategy and maturity of the data platform. For instance, during the covid pandemic, given the significant number of unknowns/uncertainties, Bucket 2 was the higher priority. For organizations with mature data platforms, Bucket 4 is increasingly either automated or specialized into Analytics Engineering roles.

Tactical tasks start off by being categorized as Bucket 2 (High Impact, Low Complexity). Upon working on the ask, four outcomes are possible:

• Complete the task
• Realize the task is “more than just a few hours of work” and needs to be planned in roadmap/sprint planning
• Complete the basic work to get the required KPIs as a time-bound activity — instead of spending time building dashboards, make it self-service for the business team to further analyze the data using excel, etc.
• Defer the task to data engineering or analytics engineering if they are better suited to accomplish it using the modern data stack.

How to maximize business outcomes without burning

In linking data to business outcomes, Data Analysts play the most critical role. They are typically the only people on the Data team that understand both the data context as well as the business context.

In my experience, on a typical day, a Data Analyst will get pulled into multiple ad-hoc tactical requests by the business:” Can you send the 2-year trend for our product CAC by this afternoon.” While these requests are a distraction from the strategic roadmap tasks, a subset of these can be critical for business outcomes. Also, a good percentage of these requests tend to be repetitive or debugging oriented since the “CAC values in the Tableau dashboard seem to be different from the quick email report.” Some of these ad-hoc requests may need much more than just a “few hours of work” and need to be planned.

As a data analyst, how to balance the time across tactical requests and strategic roadmap activities with the goal to maximize the impact on business outcomes? Among the tactical tasks, how do you separate the wheat from the chaff and address ones with higher business impact?

This framework is based on my experiences leading data analyst teams. Without the framework, everything becomes a priority, and can lead to Data Analyst teams working 24 x7 before finally burning out!

Let’s start off with a sample of typical tasks of a data analyst:

• Building reports/dashboards/notebooks for analysis
• Supporting ad-hoc requests for reports/analysis
• Analyzing the data/Explaining the “why”/storytelling for the business team to make insights actionable
• Understanding business requirements and context
• Debugging business logic inconsistencies across reports
• Debugging data quality and pipeline issues/Debugging data models/Dashboards
• Creating/updating dashboards
• Standardizing business logic for the metrics
• Improve data literacy within the business team
• Finding data sources, metadata details, and data documentation
• Preparing/wrangling data/verifying governance
• Change management related to datasets and business logic
• Operationalize data models/views/data apps/dashboards
• …and more…

Consider diving these tasks into a 2 x 2 matrix of Impact (on business outcomes) and Complexity (to technically accomplish the task).

• High Impact and High Complexity (Bucket 1): Manage as a roadmap
• High Impact and Low Complexity (Bucket 2): Manage as tactical backlog with rapid turnaround
• Low Impact and High complexity (Bucket 3): Examples include data quality, dashboard inconsistencies, pipeline performance, etc. Manage using the emerging ecosystem of tools in the Modern Data Stack.
• Low Impact and Low Complexity (Bucket 4): Manage as self-service tasks done by the business team (with guidance as required)

Revisiting the task list with the context of the 2 x 2 matrix:

• Building reports/dashboards/notebooks for analysis — Bucket 1
• Supporting ad-hoc requests for reports/analysis — Bucket 2 (needs to be further analyzed — more details later in the blog)
• Analyzing the data/Explaining the “why”/storytelling for the business team to make insights actionable — Bucket 1
• Understanding business requirements and context — Bucket 1
• Debugging business logic inconsistencies across reports — Bucket 2 or 4
• Debugging data quality and pipeline issues/Debugging data models/Dashboards — Bucket 4
• Creating/updating dashboards — Bucket 2 or 3
• Standardizing business logic for the metrics — Bucket 1
• Improve data literacy within the business team — Bucket 1
• Finding data sources, metadata details, and data documentation — Bucket 4
• Preparing/wrangling data/verifying governance — Bucket 4
• Change management related to datasets and business logic — Bucket 2 or 3
• Operationalize data models/views/data apps/dashboards — Bucket 4

The time spent on each of the buckets varies based on the overall analytics strategy and maturity of the data platform. For instance, during the covid pandemic, given the significant number of unknowns/uncertainties, Bucket 2 was the higher priority. For organizations with mature data platforms, Bucket 4 is increasingly either automated or specialized into Analytics Engineering roles.

Tactical tasks start off by being categorized as Bucket 2 (High Impact, Low Complexity). Upon working on the ask, four outcomes are possible:

• Complete the task
• Realize the task is “more than just a few hours of work” and needs to be planned in roadmap/sprint planning
• Complete the basic work to get the required KPIs as a time-bound activity — instead of spending time building dashboards, make it self-service for the business team to further analyze the data using excel, etc.
• Defer the task to data engineering or analytics engineering if they are better suited to accomplish it using the modern data stack.