Rebate Management
Rebates are funds that Tiki's retail business leveraged to lower product prices and boost sales. Using rebate is like making a bet, and over-utilization had caused tens of billion VND in losses to Tiki's operating costs.
Our tech team developed a tool to help better estimate/predict and manage the rebate budget, however, due to poor design, the team in charge did not adopt our product.
This project is an example of how Usability, especially Learnability, in complex operational systems is crucial for effective operations, and ultimately, for driving positive business outcomes.
Summary
Challenge
Users did not adopt our product, which causes risks to Tiki's operating cost due to budget over-utilization.
What I did
Identify the causes through user interview.
Then redesign the product to achieve better usability and adoption.
result
The submitting rebate process was completely transitioned to our tool.
Budget over-utilization was reduced from tens to only few billions VND.
what i've learnt
Usability, especially Learnability, in complex operational systems is crucial for effective operations, and ultimately, for driving positive business outcomes.
Rebates—a risky bet
Tiki has a retail business where they buy products from suppliers, store them in Tiki's warehouse, and sell them on the Tiki marketplace under the name Tiki Trading.
A rebate is a
discount from suppliersfor purchasing in bulk or meeting specific conditions. Some conditions are simple to estimate the receiving amount (e.g., buying 1,000 units for a 10% discount), some are more tricky (e.g., 1$ discount for each item Tiki sells).Tiki uses rebates to lower product prices on its platform and stay competitive. However, to take advantage of the rebate, Tiki needs to cover the promotion costs upfront using its own funds
beforereceiving the rebate.
As you can see, this solution carries risks:
If the rebate amounts are
overestimated, Tiki may face losses if the selling prices plus rebate amount are less than the cost of acquiring the products.Conversely, if the rebate amounts are
underestimated, Tiki will miss opportunities to invest more in promotions.
To mitigate risks, rebate approval has two layers:
The Procurement team secures contracts and reports expected rebates.
The Finance team, who provides the company’s money (before the supplier actually disburses it), double check to ensure the reported rebate matches contractual terms. Only after this approval can the Procurement team use the rebate amount for promotions.
Adoption problem
At that time, our Retail IQ team has provided the Procurement team with a tool (not designed by me) that:
Allowed the Procurement team to input rebate conditions, making it easier for the Finance team to verify them against contracts.
Once approved, the tool
automatically trackedwhether conditions were met, reducing the need for initial predictions and allowing the Procurement team to adjust expectations asconditions (sales, revenue, etc.) changed in real-time.
However, three months after launch, the Procurement team still relied on the previous method—manually tracking eligibility and sending reports to the Finance team—leading to continued over-utilization issues (tens of billions of VND).
Identify the causes through user interviews
I met some representatives from the Procurement team to discuss the issue. It appeared that the team found the tool hard to use, the interface was unclear and sometimes confusing. They felt like it takes too much effort to figure out, so they’d pretty much given up on it. I summarized the findings into following main problems along with few examples.
The improved design
The detailed design process is not covered here but can be explained further during interview/upon request. Many activities were taken before jumping into & during Figma. Please reach out and I'll share a more comprehensive presentation of the project.
Improvements in information presentation
Existing data was restructured or re-grouped for better learnability.
Additional information was added for better clarity.
Important alerts were made more prominent for better efficiency.
The rebate amount summary was visualized using a stacked chart for more meaningful insights.
Improvements in Error Handling
Possible error cases and corresponding error messages were also defined. I created a guideline on how to display these error messages for the tech team to follow.
Result
✔ Within a month, the rebate approval process fully transitioned to our tool.
✔ Budget over-utilization dropped from tens of billions of VND to just a few billion.
✔ Approval speed was increased—confirmed by Procurement & Finance team.
✔ As part of a larger initiative to stabilize Tiki Trading’s profit post-COVID, this project directly improved financial control and helped maintain profitability.