browsers into
regulars.
How Swiggy Instamart could convert more cart-builders into repeat customers, and the one test I would run first.
Payable totals from one late-night session. Not like-for-like baskets.
My short answer
Test the clarity before blaming the price.
Show shoppers the payable total and fee conditions earlier in the cart, and test it against a control group before calling it the root cause. Instamart’s checkout already shows bill details. The question is whether earlier and clearer changes what people do.
What I wanted to answer
- What decides which quick commerce app people open?
- Where do discovery and checkout create friction?
- How do fees, minimum-order thresholds and offers change perceived value?
- Which intervention is both testable and measurable?
Three personas, three ways to define value
- Compares apps and orders where the price is lowest
- Reported abandoning orders over pricing, extra-fee worries and one 33-minute delivery
- Reading: total price and speed expectations may decide completion
- Uses several apps by price and availability
- Minimum order matters; Instamart felt less affordable because of extra charges
- Reading: value includes fees, thresholds and service recovery
- Compares first for essentials, wants low minimums
- Calls Instamart a last resort for some orders; says it looks too much like Swiggy
- Reading: value perception and differentiation may both matter
Interview reports, n=15, anonymised. Directional only.
Where the journey gets heavy
- 1Need
- 2App choice
- 3Discovery
- 4Cart and checkout
- 5Delivery and after
| Stage | What I found | Source |
|---|---|---|
| Need | Urgent top-up or routine restock? One participant checks a nearby store first. | Interview report |
| App choice | Price, availability, minimum order and convenience get compared across apps. | Interview report |
| Discovery | Water and chips took scrolling. Amul Taaza and regular Amul Vanilla were not available on Instamart. | Walkthrough |
| Cart and checkout | “What will I actually pay?” Fees and thresholds looked different on each app. | Walkthrough + interview |
| Delivery and after | One 33-minute delivery. One wrong-item order, fixed by support. | Interview report |
Testable now: uncertainty about total cost. Discovery and assortment come next, in a separate test.
Same night, three different bills
| App | Items | Handling | Extra fee | Delivery | Payable | Estimate |
|---|---|---|---|---|---|---|
| Instamart | ₹317 | ₹12 | ₹35.40 | ₹0* | ₹364 | 23 min |
| Blinkit | ₹410 | ₹13 | ₹30 | not listed | ₹453 | 22 min |
| Zepto | ₹357 | ₹0 | ₹0 | ₹0 | ₹357 | 5 min* |
*Instamart’s delivery fee was waived through Swiggy One. Instamart showed a ₹35.40 late-night fee; Blinkit a ₹30 rain surge charge; Zepto showed delivery, handling and late-night fees as waived. Product variants and pack sizes differed, so this is not a like-for-like price comparison, and one session does not establish typical delivery speeds.
Instamart and Blinkit showed extra charges; Zepto showed them as waived in this basket. Instamart didn’t have Amul Taaza or regular Amul Vanilla in this session.
That one app is cheaper or faster overall, or that fees cause abandonment. Seeing a fee isn’t the same as leaving because of it. Or what fees and stock look like at other addresses or times.
Three ideas, one to test first
Two interviewees raised price or fee concerns; one reported abandoning orders over price. Main doubt: does clearer presentation change behaviour, or is the price itself the barrier?
Some items needed scrolling and two products were missing in this session. Main doubt: how often does this cause exits or substitutions?
One participant said it looks too similar to Swiggy. Main doubt: does a distinct look drive purchase or loyalty?
“Test first” means worth testing, not proven. If funnel data shows discovery or stock-outs lose more orders, the ranking changes.
The test: Clear Basket
| Design | |
|---|---|
| Hypothesis | If Instamart shows the expected payable amount and fee conditions earlier, more carts become orders, without hurting contribution margin. |
| Primary metric | Cart-to-order conversion: eligible cart users who place an order, divided by all eligible cart users, over a sample size and duration agreed in advance. |
| Control | Today’s cart and checkout. |
| Treatment | A cost summary in the cart, before payment: subtotal and discounts, applicable charges, estimated total and fee conditions in plain language. |
| Randomisation | Eligible users are randomly assigned to control or treatment, and each user keeps the same group. |
| Guardrails | Contribution margin: if conversion rises but margin falls past an agreed limit, don’t scale. Also watch abandonment, order value, repeat purchase, complaints and errors. |
How I would decide
A proposed 30 days:
- 1Week 1: check funnel data and fee-related exits
- 2Week 2: prototype the summary, test comprehension
- 3Week 3: set up events, thresholds, sample size
- 4Week 4+: launch the random test, then scale, iterate or stop
The timeline is a proposal. No uplift target yet: there is no baseline conversion rate to set one against.
What could make this wrong
Not how it’s shown.
Funnel data may show search misses or stock-outs lose more orders than checkout.
15 interviews and one late-night basket can’t speak for Instamart’s customers.
Showing fees earlier could scare people off as easily as reassure them.
My closing take
The goal isn’t more orders. It’s an informed decision, an order when the value works, and a reason to come back.
Next: check checkout abandonment and fee-related exits in Instamart’s funnel data. If material, run Clear Basket. If discovery misses or stock-outs lose more orders, prioritise those instead.
Want the full case study?