The Problem
Before DeliveryPage, customers had no choice in how their orders were delivered. They were locked into whatever delivery arrangement a business had, with no visibility into pricing or delivery windows, and no say in the matter.
DeliveryPage's answer was to become the bridge: onboard delivery partners directly, let them manage their own pricing and availability, and give customers a say in how their package gets to them — without the business absorbing the operational risk of running logistics itself.
That single decision — DeliveryPage as bridge, not operator — shaped almost every design decision that followed.
The System
DeliveryPage serves delivery partners on one side and customers on the other:
- Delivery Partners — execute deliveries, managed through the Partner Dashboard
- Customers — place and track orders via the Mobile App and Web Dashboard, which function as parallel experiences rather than divergent ones
- Landing pages for each surface, focused on conversion and positioning rather than product decisions
Both the Partner Dashboard and the customer-facing Mobile App/Web Dashboard carried real design decisions — the Partner Dashboard around serving different business sizes on one surface, the customer side around handling edge cases and keeping the experience consistent across devices.
Design Decisions: Partner Dashboard
1. DeliveryPage as bridge, not operator
Delivery partners set their own prices, routes, and availability windows. Rather than DeliveryPage dictating terms, partners operate with autonomy — DeliveryPage's role is to connect them to demand, not manage their business for them.
2. Splitting Shipments by origin, not filtering a unified list
Partner businesses range from small operations with no in-house logistics to established businesses running their own fulfillment alongside DeliveryPage orders. The first approach was a single shipment list with filters to separate marketplace orders from a partner's own shipments — in practice, this added tabs and filter logic that smaller partners had to navigate just to find orders that mattered to them.
The dashboard was restructured around two distinct sections instead: Shipments, for orders originating from DeliveryPage customer orders, and Direct Shipments, for a partner's own in-house fulfillment. A partner with no direct shipments never sees that section at all — the interface stays simple by default rather than requiring a mode to be toggled. Established partners get a dedicated space to manage their own logistics alongside marketplace orders, without either use case cluttering the other.
3. Curating partner autonomy against customer experience
Giving partners free rein over pricing and availability created a risk: customers could end up choosing between too many options, or options that weren't genuinely competitive. Rather than restricting what partners could set, the dashboard surfaces a curated selection of three delivery options to the customer at checkout — partners retain full control over their own pricing and terms, while customers are protected from being overwhelmed or exposed to weak options. Autonomy on one side, curation on the other.
Design Decisions: Customer Experience (Mobile App & Web Dashboard)
4. Handling weight discrepancies without blocking the order
Delivery pricing is tied to the weight a customer declares at checkout. When a package's actual weight exceeds what was entered — a user error, not a partner dispute — the order isn't silently held or cancelled. Instead, the customer gets an "Action required" notification with a choice: top up the payment difference to proceed, or cancel and receive a refund. The customer stays in control of the resolution rather than the system making the call for them or the order stalling with no clear next step.
5. Making the web dashboard feel like the app, on purpose
Rather than treating the Web Dashboard as its own product with its own conventions, it was deliberately designed to mirror the Mobile App's structure and interactions as closely as responsive layout allows — same flows, same terminology, same mental model. Since there are no functionalities exclusive to either surface, a customer who's used the app shouldn't have to relearn anything switching to web, or vice versa.
Other Surfaces
Landing pages — conversion and positioning-focused, not a source of product decisions.
Reflection
DeliveryPage doesn't have outcome data attached yet. Looking back, the strongest work spans both sides of the product: the Partner Dashboard, where two genuinely different business types had to be served by one surface, and the customer experience, where edge cases like weight discrepancies had to be resolved without leaving the customer stuck or the order in limbo. Across both, the harder problem was less about visual design and more about deciding where to give people control and where to protect them from complexity they didn't need to see.
If revisiting this today, the next test would be validating whether the curated three-option selection at checkout is actually the right number, or whether customer behavior data would justify showing more or fewer choices.