workflow redesign
Redefining the decision making workflow for Sales and Market strategy teams
The Sales and Marketing teams used 4+ dashboards to analyze a portfolio of 300+ pet food products.
Teams were repeating analyses across different tools to answer a fairly straightforward set of questions:
What's going wrong?
→
What's influencing it?
→
What actions needs to be taken?
→
How do we anticipate it?
This fragmented workflow meant business decisions could take up to 3 weeks.
I redesigned the workflow around how these teams actually make decisions, bringing a fragmented analysis process into one continuous experience.
Look out for the feedback I received from the users!
project setup
Mars
Lead UX Designer
1 PM · 1 Developer · 2 Designers
Jan to Mar 2024
2024
Couldn’t have done it without this stellar team.
The final product
One dashboard that brought portfolio analysis, investigation and decision-making into a continuous workflow instead of a fragmented one.
4 dashboards. One decision.
Teams jumped between dashboards, repeated analyses, and manually connected insights before they could make a decision.
before
One continuous analysis workflow.
The new experience brought the analysis, investigation and next steps into one place.
after
What I needed to solve
01
Simplify the decision making workflow.
Reduce the repetitive steps and tool switching required to answer a business question.
02
Design the IA around the business process.
Mirror how users analyse data, from a high level signal down to granular detail, in one continuous path.
03
Design dynamic visualisations.
Let users explore the portfolio by market, brand, metric or level without being limited to Top 10 / Bottom 10 views.
Constraints
A few constraints shaped the approach I took for this project.
Tight timeline, different priorities
Different managers, KPIs and reporting lines.
One developer
The solution had to be feasible for a single engineer to build.
Approach
Step 01: Understand the users
Before designing screens, I needed to understand how a business question actually became a decision.
Methodology
I spoke with Sales and Market Strategy stakeholders (N=12) across three levels, walked through their existing dashboards and mapped how they moved from identifying a performance issue to taking action.
Stakeholder interviews
Heuristic Evaluations
Moderated, remote interviews
the two teams
Sales Team
Focused on product and retailer performance. Sales needed to identify where performance was dropping, which products were responsible and what action to take with the retailer.
Market Strategy Team
Focused on market share performance and competitor benchmarking. Market Strategy needed to understand category and brand movement, identify what was driving it and find opportunities across the portfolio.
Step 02: Understand the business process
I mapped how a business question moved across Sales, Market Strategy, Data & Tools, Finance, Business Leads and Retailers. The goal was to understand why one decision required multiple dashboards, where teams depended on each other and where the process slowed down.
What I learned
Key friction points slowing down decision-making
1
Tool fragmentationTeams jumped between 4+ dashboards and data sources to answer one question.
2
Context reconstructionEvery switch meant rebuilding filters, hierarchy and data levels from scratch.
3
Cross-team dependencySales and Market Strategy couldn't complete the analysis independently.
4
Data reconciliationDifferent sources and formats led to mismatches and extra validation loops.
5
Decision lagBy the time decisions were made, the data was already starting to change.
Step 03: Define and prioritise use cases
PM collaboration was huge here. We constantly pressure tested these against real business questions.
I mapped every use case against impact and effort to decide what to build first.
QUICK WINS
MAJOR PROJECTS
FILL-INS
TIME SINKS
Bring the relevant portfolio data into one view for each analysis.
Assess a brand across demand, market share evolution, content and reviews.
Look at all products and brands, not only top 10 and bottom 10.
An AI assistant generating quick answers on market share for the weekly MS&D clinic.
Recommendations on buy box, sell-in and sell-out actions, ranked by priority.
Intelligence to anticipate and act on negative contributors before they hurt performance.
Investigate why a specific cat brand underperforms peers on Amazon.
Identify and present drivers behind a dip in category market share.
Identify contributors to slower growth and action them with the retailer.
Simulate long-term innovation strategy and its impact on the product mix.
See projections for leading-indicator metrics to act on ahead of time.
LOW EFFORTHIGH EFFORT →
MS&D use cases
Sales use cases
Time sinks, deprioritise
Most Market Strategy opportunities fell into quick wins or major projects, while several Sales and process-heavy requests were more expensive relative to their immediate impact.
design explorations
How do you turn all of that into one dashboard?
Breaking the business process down by team
The business process showed where decisions slowed down. I then mapped the Sales and Market Strategy flows separately to understand what each team actually did across the different tools.
The starting points were different, but both teams followed a similar decision-making pattern.
Iteration 01: Organizing around the data
✓What worked
The first approach gave each team a dashboard built around its own data and reporting needs. It reflected how the teams already worked.
✕What didn't
Two dashboards meant more development effort and still required leadership to switch views to understand the complete picture.
Iteration 02: Organizing around the decision
The teams had different mental models at the data level, but followed the same pattern at the decision level. So instead of structuring the product around teams, I structured it around three shared actions:
Analyse
What's happening?
Investigate
Why is it happening?
Anticipate
What should we do next?
This became the information architecture for the dashboard.
goal 03
300+ products don't fit neatly into a Top 10 chart.
Traditional dashboards reduced the portfolio to things like Top 10 / Bottom 10, which meant users constantly changed filters or views to understand the whole picture. I wanted users to be able to ask:
Which market, brand, metric, or portfolio level , without rebuilding their analysis every time.
Moving from portfolio level down to a single SKU
impact
3 weeks - 3d
Decision making time, down from three weeks.
4 to 1
Dashboards unified into a single workflow.
$500k
Engagement converted into a year long contract.