01
Location gated the conversation
The flow couldn't progress until the expected location input was provided.
FARMAKO
CONVERSATIONAL UX
Helping customers order medicines through a more useful WhatsApp conversation
Customers came to WhatsApp to find out whether their medicines were available. But the ordering flow required information (location, address details, etc.) before it could respond to their needs.
I explored how to make the experience more helpful for users.
CONTEXT
User WhatsApp conversational experience
Farmako's one of the order channels was WhatsApp. But the flow felt nothing like a conversation. Users messaged expecting an interaction that would help them get medicines urgently, but instead they were asked a series of questions like a form filled out before being admitted.
SELF CONVERSATIONAL AUDIT
Understanding the existing experience
I started auditing the conversation by myself by trying different chat prompts.
02
No recovery for unexpected input
When I responded again, the flow repeated the same prompt instead of helping me recover.
ANALYSIS
Understanding how people actually ordered
I reviewed real WhatsApp orders to understand how users navigated the existing flow.
01
Users prioritised the medicine request
Instead of responding with location, some users continued by sharing the medicine they wanted.
02
Availability was an early user question
Some users wanted to know whether a medicine was available before progressing through the ordering flow.
03
The flow requested multiple details before fulfillment was established
Users were asked for billing and delivery information alongside their medicine details early in the journey. Information collection was front-loaded.
INITIAL CHALLENGE
How might we help customers get their experience easier and faster through WhatsApp?
IDEATION
Exploration
Initial analysis as per research of user behaviour.
01
Letting user know the reason before asking for location.
Cart confirmed → Address/details → Payment
02
Supporting different ways users express medicine needs
Prescription / Medicine name / Previous order / symptoms
03
Collecting fulfillment details progressively
Cart confirmed → Address/details → Payment
04
Designing for unavailable medicines
Unavailable → Alternatives → Revised cart
CONSTRAINTS
When customer needs met system constraints
Farmako needed customers' locations upfront to check nearby medicine stock and delivery availability. At the same time, WhatsApp's interaction limits restricted how much flexibility we could introduce into the conversation.
The challenge was to make the flow feel responsive to customer needs while working within these constraints.
CUSTOMER EXPECTATION
Do you have Telma 40 mg?
Wants to check medicine availability
V/S
SYSTEM REQUIREMENT
Please share your location to continue.
Needs location to check nearby stock
DESIGN DECISIONS
Designing for conversations that don't follow a script
The existing bot expected customers to follow a fixed sequence. But customers often started with medicine names, availability questions or unrelated messages. I explored how the bot could respond to their intent while still collecting the information required to check stock.
DESIGN DECISION 1SHIPPED
Location couldn't go, so it got a reason
Before asking for location, the bot explains why Farmako needs it: to check nearby stock and delivery availability.
Principle
Explain requirement before asking
DESIGN DECISION 2PROPOSED
Ask for details when they're relevant.
Instead of presenting every order requirement at once, the proposed conversation collects additional details as they become relevant, while retaining mandatory checks.
Principle
Collect progressively, not all at once
DESIGN DECISION 3PROPOSED
A stock-out became a next step, not a dead end.
When a requested medicine isn't available, the bot can offer an available same-salt alternative and explain the next step instead of leaving customers to figure out what to do.
Principle
Offer a useful next step when the answer is no.
The proposed experience
OUTCOME
What shipped
Following the conversational audit, the team implemented an improved location-request message that explained why customers needed to share their location.
The broader conversational flows were explored but were not launched on WhatsApp because of platform constraints. The four entry-path concepts also informed the in-app AI chat exploration.
Early observations
Early internal observations suggested that more customers were sharing their location after the change.
LEARNINGS & REFLECTION
Make each step relevant in conversational UX
Working on Farmako's WhatsApp experience changed how I approached conversational UX. The challenge wasn't removing every step. It was making each step feel relevant to what the customer was trying to accomplish.
What I'd Do Differently
I would have involved the team earlier around WhatsApp's technical constraints. I designed the four entry paths based on user behavior and the right instinct but didn't know what the platform allowed. Earlier alignment would have kept the solution within real constraints from the start.