The locker delivery experience currently suffers from serious usability issues, particularly when orders contain multiple sub-orders. This impacts a large number of users. Customers often don't find out whether their order is eligible for locker delivery until late in the checkout process. The delivery promise may change without clear communication, and more critically, users can lose the option for locker delivery entirely if just one of their packages is ineligible. Addressing this problematic experience would likely increase locker usage, which is a key strategic goal for the business. Specifically, our aim was to raise usage from 35% to 50%.
Problem Statement
The current experience for locker selection in multi-suborder deliveries is not optimized, despite being a frequent user path. Some of the key pain points users face include:
Lack of Early Visibility: Users don't know which suborders are eligible for locker delivery (and the associated delivery fees) until they reach the order details page — or in some cases, the payment stage.
Inconsistent Locker Selection: If all suborders qualify for delivery to a Skroutz Point, users can select the same pickup point. However, if any suborder is ineligible for locker delivery, users lose the option to select a locker for all suborders. This forces them to proceed without their preferred delivery option.
Limited Modification Options: On the payment screen, when locker delivery is suggested for eligible suborders, users cannot modify their locker selection directly within the modal.
Lack of Delivery Transparency: Changes in delivery promises are not clearly communicated during the selection process.
This poor experience creates confusion for users and can lead to cart abandonment, ultimately reducing overall satisfaction and engagement with locker delivery services.
Goal
This task focuses on improving the experience for users who place orders with multiple suborders and want their packages delivered to a Skroutz Point. The changes aim to increase transparency and simplify the user journey.
Research
The problem was quite clear, but before jumping into solutions, I wanted to ensure I had a solid understanding of the users' pain points and that the entire team was aligned on both the problems and potential solutions.
Fortunately, we had a good amount of data related to locker selection and locker eligibility. Some key data points we considered were:
35% of orders contain multiple suborders.
On average, 50% of eligible locker orders are completed with a locker as the selected delivery method.
It’s important to clarify what we mean by “eligible locker orders.” An order is considered eligible for locker delivery if it meets the following criteria:ker selection and locker eligibility. Some key data points we considered were:
It falls within the acceptable size dimensions (e.g., a sofa isn’t locker-eligible).
The merchant partners with Skroutz Last Mile (90% of merchants do).
The product category is locker-eligible — only a few categories are excluded from locker delivery.
Current Flow
Heuristic Evaluation
Before identifying problems and brainstorming solutions, I ran a heuristic evaluation session with the team to uncover usability issues in the current locker selection experience. My main goal was to surface all the pain points a user might encounter and assess their severity, but more importantly, to align the team on the key problem areas.
This shared understanding would make it easier to later define the user journey, mapping out what users do, what they feel, and the obstacles they face throughout the process.
User Journey
As a next step, by using insights from the heuristic evaluation along with qualitative data from previous user interviews, I designed the user journey. This helped define what users do during the process of sending their suborders to a Skroutz Point, how they feel at each stage, what their goals are, and the pain points they encounter.
By clearly mapping out these aspects, we can identify key opportunity areas that will help us shape user-centered solutions.
Opportunity Tree
Since one key aspect of the problem is the lack of early visibility, we decided to map the opportunities across the three main phases of the user funnel:
Product discovery: The user is still browsing and searching for products.
Cart review: The user has finalized their cart and is preparing to check out.
Checkout: The user is entering personal details and proceeding to payment.
Starting from the data we already had, I collaborated with the product manager to build the opportunity tree, organizing insights and potential improvements based on each stage of the journey.
Iterations
The product trio decided which solutions to prioritize first:
Suggest alternative carts that are locker-eligible
Display locker eligibility at the line-item level and explain why certain items cannot be delivered to lockers
Clearly show delivery fees for each available option
While the solution for displaying delivery fees is being handled by another team, it remained a key goal that guided the solutions we designed.
1st Iteration
Although the user journey begins much earlier, the team decided to focus on the post–product discovery phase to deliver a quick, MVP-level solution. Since changes to the listing filters were already planned, we concentrated our efforts on designing solutions specifically for the cart and checkout pages.
The main idea behind the first iteration is to allow users to select, directly from the cart, that they want their items delivered to a Skroutz Point. If some of the suborders are ineligible for locker delivery, the order is then split during the delivery options step in checkout: eligible items are assigned to locker delivery, while ineligible ones are assigned to home delivery.
Testing
We tested this solution at the wireframe level internally, but the checkbox in the cart didn’t perform well. Most users didn’t notice the option, and those who did select it, opting to deliver their products to a locker, were confused by the label “This item is not eligible for locker delivery.” While rephrasing the message could partially address the issue, the deeper problem was that users didn’t seem interested in making this decision during the cart step. Our user research team confirmed this insight.
2nd Iteration
Having decided to communicate locker eligibility earlier in the funnel—specifically by adding a “Delivery to Skroutz Point” filter on the listing page—we also chose to make no changes to the cart for now. As a result, the user reaches the checkout page and, if they choose to have their products delivered to a locker, we need to support that selection seamlessly.
The flow is as follows:
As you can see, this iteration was created as a mockup rather than a wireframe, as I wanted to test it in a more finalized, realistic state.
Testing
Once again, we tested the prototypes internally. While this iteration performed better than the previous one, it still presented some challenges. On one hand, it successfully addressed the core problem, allowing users to split their suborders and choose delivery to a Skroutz Point. On the other hand, the new flow introduced several unfamiliar patterns and seemed to require significant cognitive effort from users. It was a step in the right direction, but it needed further refinement to become more intuitive and user-friendly.
Plot Twist
While I was in the process of finalizing the designs, a strategic business development provided timely support: Skroutz partnered with ACS, the largest courier company in Greece, making it our exclusive courier collaborator alongside Skroutz Last Mile. This meant that all our merchants now collaborated with either Skroutz Last Mile or ACS, making their products eligible for locker delivery. As a result, with the ACS collaboration, the percentage of total orders that could partially be delivered to a locker dropped to just ~3.5%.
Final Design
To simplify the experience, when one or more items couldn't be delivered to a locker, I followed the previously designed and tested flow, but streamlined the way it was presented to users:
This time, we tested the flow with external users. While it initially caused some confusion, it ultimately proved effective in addressing the problem of partial locker delivery.
Success Metrics
The feature launch showed promising results, as the share of suborders delivered to lockers increased by 1.8 percentage points. While other factors may have influenced this trend, the increase is considered reliable, as it aligns precisely with the release date and has remained consistent since then. Additionally, the share of partial locker delivery orders out of total orders nearly doubled (from 0.75% to 1.35%), further indicating that the new flow addressed a real user need and improved adoption.
