Building a Business Intelligence Architecture at Aventine

Situation

Aventine operated in a volatile, narrow margin business environment. Employees needed be armed with the most up-to-date information so that they could make data-informed decisions on the fly.  Unfortunately, decision making and strategic planning capabilities were severely diminished due to limited visibility to financial data. 

Static reports, which were prepared by the finance department, were:

  • Delivered once per month, which was not frequent enough
  • Backward-looking, always a month behind
  • Limited,  excluding critical financial metrics
  • Summarized, showcasing numbers in aggregate with no ability to drill down into data

Not only were these reports inadequate, they were manual and time-consuming to produce. Further, because the reports did not meet needs, the finance department often had to spend additional time completing ad-hoc reports and analysis for specific requests.

Action

I led effort to implement business intelligence software tools. Once tools were implemented, I developed a series of scorecards, dashboards, and reports to allow teams to monitor operational and financial performance. 

Specific activities are summarized here and detailed below:

  • I worked with business teams that needed access to data to understand requirements
  • I worked with finance department to understand data architecture of financial system and their process.
  • I partnered with IT to identify and evaluate business intelligence tools that could integrate with the company’s financial system (Oracle ERP)
  • I led effort to integrate financial system with business intelligence software
  • I developed robust, dynamic scorecards and dashboards for each business audience which allowed teams to monitor operational & financial performance.

Results

  • Increased visibility to real-time, actionable data
  • Better, faster decision making
  • Improved strategic planning and financial reporting activities
  • Reduction in time spent by finance teams; increase in accuracy and usefulness of data

Additional Details

Step 1: Conducted one-on-one stakeholder interviews with all parties:

  • C-Level Officers
  • Senior executive team members
  • Management team members (multiple departments)
  • Front-line employees (multiple departments)

Step 2: Organized all information, preparing various self-referencing documents that articulated business goals and requirements:

DOCUMENTSCONTENTS
GOALS & OBJECTIVES

·       Enterprise business objectives

·       Departmental strategic objectives

·       Supporting KPIs and metrics

PERSONAS

·       Persona details

·       Goals

·       List of tasks (What decisions does persona make with data?)

TASK ANALYSIS

·       Steps involved in each task

·       Data needed for each task

DIMENSIONS & MEASURES

·       List of dimensions and measures required

·       For each dimension and measure, indicated:

·       Related tasks

·       Related personas

·       Data permission level (to restrict visibility of sensitive data)

·       Must have vs. nice-to-have

Step 3: Created wireframes which illustrated possible dashboard layouts

This was an iterative process which required continuous feedback from stakeholders

 

Step 1: Conducted stakeholder interviews with select members of finance team to discuss:

  • Process used to create monthly reports
  • Data points included on reports
  • Financial assumptions
  • Manual calculations
  • Process used to conduct ad-hoc analysis
  • Most common ad-hoc requests
  • Pain points and challenges to compiling report and conducting ad-hoc analysis

Step 2: Updated previously created requirement documents with new information

 

Step 1: I developed scoring criteria based on business and technical requirements

Step 2: I evaluated multiple business intelligence tools using a pugh matrix

Step 3: I conducted a NPV analysis on the top 3 options

Step 4: I provided analysis and recommendations to C-level executives.

Ultimately, my recommendations were selected:

Step 1: Worked with members of IT team to construct “as-is” data schema. 

This schema visualized data architecture of the company’s Oracle ERP system (Oracle SQL Modeler)

Step 2: Conducted gap analysis; compared data schema to requirements to identify:

  • Missing data points
  • Missing dimensions
  • Data that needed more granularity

Step 3: Worked with members of IT to construct and document the “to-be” data model.

This model included several considerations:

Developed 20+ custom dashboards and reports for various stakeholders:

  • C-Level officers
  • Senior executive team members
  • Management team members (multiple departments)
  • Front-line employees (multiple departments)

Each dashboard included:

  • Filters to change dimensions (date, categories, etc.)
  • Drill-down data for deeper analysis

To minimize cognitive load and increase understandability, every dashboard had 4 common sections.

  1. Data points related to enterprise objectives
  2. Data points related to departmental objectives
  3. Custom data points and/or views specific to each audiences needs
  4. Industry benchmarks and data points

Image: not actual work example

SITUATION

ACTION

RESULTS

Find this information interesting?​

There is more where that came from!

Scroll to Top