Item Data Audit Tool (IDAT)
Background:
IDAT is a web application used specifically by auditors to enforce item data accuracy for products to be redistributed by a large home retailer. We (balanced team) were tasked with creating an interface that would decrease errors/'bad audits' and increase efficiency.
Process:
1. User Research
2. Prioritization + Alignment
3. Ideation
4. Design, Test, Iterate
Skills:
Remote and in-person user interviews, persona creation, ideation workshops, wireframe + prototype (Sketch), usability test
Results:
• 45% decrease in bad audits
• 23% increase in rate of audits
• Successful implementation of IDAT 2.0 nationwide
Process Deep Dive
1. User Research
→ Interviews:
10 remote/in-person user interviews with Auditors
Explored pain points, needs, and opportunities. Additionally interviewed several other secondary users who use audit data from IDAT, such as Supply Chain Data Analysts, Supply Chain Managers, and Support Specialists. Interview key takeaways below:
Top User Pains
• "I am not aware when I am auditing a 'special' item with specific auditing guidelines."
• "I input the data for the wrong hierarchy. For example, I believe I am auditing the case, but I am auditing the each, or vice versa."
• "The error messages are not helpful- I do not know what to do to correct the error."
• "After opening up a case for auditing, I close it, forgetting that I need to scan the UPC of the each again."
• "I mistype numbers for items that are too large for the CubiScan."
• "I do not understand how certain items should be oriented when I submit an audit."
Internal Stakeholder Needs
• Accurate audit data to provide to vendors in the case of failed audits
• Descriptions of any atypical circumstances surrounding a failed audit
• Indication that the item audited was a 'special' item
Business Goals
• Decrease the rate of 'bad audits' (auditor error resulting in falsely incorrect audits) to less than 1%
• Increase the rate of audits performed by at least 25%
• Implement IDAT smoothly across distribution centers nationwide with little to no training
→ Persona creation:
Collaborated with team to determine that it was necessary to focus on newer, less experienced auditors due to high turnover rate (auditors are typically promoted after 2 years) and because when implemented nationwide, IDAT would be utilized by many day-one users.
IDAT User Persona
Alex the Auditor
Wants and Needs
• Quick audit process
• As few clicks as possible
• To conduct audits accurately
• Simple and straightforward process
• Ability to self-correct errors (understand error messages)
Pains
• Doesn't know when he is auditing a 'special' item with unique auditing guidelines
• Typos when entering data manually
• Adjusts orientation of items incorrectly
• Mixes up hierarchies (Each, Pack, Case, Pallet)
• Forgets to scan UPC of each again when auditing case; has to reopen
About
• Has been auditing for less than 1 month
• Stands at auditing station all day
• 10-hour shifts, 4-5 days per week
• Conducts about 500 audits per week
• Places items on CubiScan to capture package dimensions and weight which are automatically uploaded into IDAT
2. Prioritization + Alignment
→ Prioritization:
Organized auditor pains based on frequency vs. level of pain.
Very Painful; Occur Frequently:
• Mama-Baby errors (most frequently occurring 'special' item)
• Having to open case and scan all hierarchies twice
• Too much clicking to complete an audit
• Auditing pallets- auditors enter case quantity instead of each quantity
• Unhelpful error messages
Very Painful; Occurs Infrequently:
• Not knowing when auditing 'special' items with unique auditing guidelines
Less Painful; Occur Frequently:
• Errors while adjusting orientation
• When lowest sellable unit ("Quantity of Eaches") is the case/pack, but auditors count individuals
Least Painful and Occur Infrequently:
• Switching back and forth between screens to look up item information
• Typos when entering data manually
→ Alignment:
Presented all findings and prioritization results to business stakeholders; discussed and collaborated until we truly aligned on the top pain points we
3. Ideation
→ Design Studio:
Ideation activity for the core team to collaboratively brainstorm potential solutions for the auditors' pain points.
Focused on one of our prioritized pains that could not be easily fixed within the interface (shown below)
User Scenario/Problem Statement:
"As an auditor, I have trouble distinguishing between the hierarchies (Each, Pack, Case, Pallet) of items that I am auditing."
Most common examples:
• "I mistake the CASE for the EACH" (Case of Hammers - Individual (sellable) hammer is the each, but the case of hammers is also sellable)
• "I mistake the EACH for the CASE" (4-pack of Light Bulbs- 4-pack is the each, not the individual light bulbs)
Design Studio Workshop with Balanced Team
Top Proposed Features from Design Studio:
• Pop-up notifications for special items (after scanning UPC)
• Display item photos
• Default certain hierarchies to have a quantity of 1
• Display full product name instead of short SKU description
• Set sanity checks for dimensions, quantity, and weight that do not physically make sense
• Bring back last UPC audited
→ Hypothesis generation:
After solution generation, team discussed and came to agreement on the following 3 hypotheses to determine success on the ideated designs.
Success Metrics Hypotheses:
We believe that adding notifications of 'special' items (Mama-Baby, sellable case, etc.) in IDAT
for auditors in distribution centers
will achieve increased awareness and accuracy while auditing these special items.
We will know this to be true when the number of bad audits involving special items is reduced by at least 50%.
We believe that displaying the product image in IDAT
for auditors in distribution centers
will achieve increased clarity on which hierarchy is currently being audited.
We will know this to be true when the number of bad audits is reduced by at least 50%.
We believe that replacing the SKU description with the full product name in IDAT
for auditors in distribution centers
will achieve increased clarity on which hierarchy is currently being audited.
We will know this to be true when the number of bad audits is reduced by at least 50%.
4. Design, Test, Iterate
→ Design process:
Went through multiple iterations of the wireframes, then prototype, tweaking based on feedback from users (remote usability tests), core team members (concept validations), and UX designers, PMs, and engineers from other teams who had no prior knowledge of the application (design critique to mimic day-one users).
Prototype 1: 2-Page Audit with Notification for Special Items
"Day-One User" Feedback from UX, PMs, and Engineers from Other Teams
Final Prototype: Auditing a Mama-Baby Item
Results and Final Thoughts
After implementation of IDAT 2.0 to many more distribution centers nationwide, there was a 45% decrease in bad audits- especially for those involving special items like Mama-Baby items. This was a huge win because it also decreased the time the internal Support Specialists spent investigating and resolving these failed audits. There was also a 23% increase on average in the number of audits performed, meaning auditor efficiency was increased.
Overall, I believe our team collaborated very well in order to research and design IDAT 2.0. This was our first project together and the team had not had any prior experience working with UX. It was great introducing them to user research methodologies and seeing them understand the importance of empathizing with the end-users of their applications.