Inventory Forecast

Ardent Mills relies on a custom application to support critical processes in its flour milling operation, such as inventory management and scheduling. As the product designer, I led this project, collaborating closely with the product manager to address specific pain points in inventory forecasting.

Client
Ardent Mills

Role
Designer

Background

Inventory woes

Managing inventory has always been a struggle at Ardent Mills due to the complexity of the milling process. It has been a constant goal to improve this issue and give users better and more accurate data.

Inventory forecast is as it sounds; it gives users a prediction of what will be available based on a specified “days out” value. This information is especially critical for schedulers planning the packing of finished material to be shipped. Detailed shipment and packing schedules are available via links in the table, and users use this information to manually estimate when inventory will run out. As you can imagine, this is not only frustrating but inefficient.

Current inventory forecast page

The Problem

Lack of visibility

Operations users lacked a quick and intuitive way to know which exact day inventory would run out. The existing process required users to manually calculate depletion dates by cross-referencing data hidden in modals.

The Goal

A detailed view

The goal was simple in concept; provide users with a detailed, daily view of inventory to heighten their understanding of the forecast. In addition, we aimed to eliminate the need for the user to perform calculations themselves. To achieve this, visual indicators were needed to highlight when inventory is low or depleted.


Process

The initial idea

Although no formal user research was conducted, I worked closely with the product manager and business users to understand the workflow and pain points. Business users voiced that they would like a solution similar to an existing feature which provides forecasts for milling.

I always consider user suggestions; it gives me insight into how they think and work. Ultimately, it’s up to the designer to find the right solution, and my final solution ended up in a different place than where I started. I quickly mocked up a design based on the users’ suggestion to see how it might work in this context. This hands-on approach helps me understand the information and find flaws and opportunities.

Technically, this solution accomplished the goal. The page defaulted to a summary view, but now users had controls to display the individual days of the forecast for each category. This addition allowed us to add visual indicators to indicate the specific days where inventory was insufficient for shipping.

The noticeable drawback was the amount of horizontal space potentially required to display the data. Mills typically forecast 7 to 14 days out, so we added a limit of 14 days out. While doing so would help, but it didn’t fully resolve the space issue.

Going a different direction

As I explored different solutions, one nugget of information that was tugging at me. The product manager has explained that the forecast is calculated with four key data points:

Shipping - (Available + Packing + Transfers) = Daily Forecast

It seemed like a no brainer that these values should be grouped together by day. This would bring clarity by telling the full story of an item on any given day. Currently, these values were separated in an unhelpful way, forcing the user to hunt for them in this giant table. Bringing the data together so the formula was intuitive became the driving force behind my solution.

The Final Solution

To optimize the presentation of data, I pivoted towards a vertical layout. This approach helped resolve the space issue while offering a clearer picture of the data by placing data points in relation to one another. The key accomplishments of this solution include:

  • Enhanced data insight: Daily forecasts were added, providing a more comprehensive overview.

  • Improved information hierarchy: Expandable rows reveal the breakdown of the values that compose the daily forecast.

  • Reduced cognitive load: Visual indicators easily identify low and short inventory levels.

  • Detailed info: Links give access to detailed order information for each day and category.

The Result

A positive future for inventory management

Our solution was presented to business users, and they were very enthusiastic about the new approach. For now, our solution has been added to our development board and is awaiting implementation. We delivered a solution that not only addresses the immediate problem but also sets the stage for broader operational improvements in PRISM.

Further needs to consider

This solution has the potential to drive meaningful progress, significantly increasing user efficiency. Schedulers represent a distinctive user group, often needing to see an extended forecast view—typically 30 to 60 days out. That volume of data can negatively impact user experience due to increased load times. This remains an open question as we investigate with developers to find a resolution.

Challenges

  • Needs of different user groups: Limiting the forecast to 14 days keeps the interface manageable, but other users require a 30-day view. This remains an open question for future iterations.

  • User-suggested solutions vs user needs: Users often request features based on what they know, but the design process revealed more effective ways to present the data.

Lessons Learned

  • Knowledge is key: Understanding business processes and data relationships is crucial. One key insight about how forecasts are calculated shaped the entire solution.

  • Find the opportunities: There is often a gap between what users ask for and what they truly need; designers must look beyond initial requests to deliver more effective solutions.