Brands were sending broken campaigns to hundreds of thousands of customers - unrendered variables, wrong audience, broken links. The platform had a test-send feature that almost nobody used because it required leaving the composer.
STUDY 12 / 30 · ANONYMISED · NDA · MARTECH · INDIA · SERIES A
MARTECH · INDIA Campaign QA that catches the mistake before the send.
RoleLead designer
Timeline8 weeks
Team1 designer, 1 PM, 2 frontend
VerticalLifecycle messaging platform · 900 D2C brands
Built inspection into the send path as an unskippable gate that renders the message against real customer records, diffs the audience against the last send, and validates every link.
Broken sends effectively stopped. The gate added twenty seconds to a workflow that had been producing brand-damaging errors weekly.
0.3BROKEN SENDS / MONTHfrom 14
1,840UNRENDERED VARIABLES CAUGHTpre-send, 6 mo
2%GATE ABANDONMENTusers who backed out
THE ARGUMENT
Why the obvious solution was wrong.
The study matters because the product problem was reframed before the interface was polished.
Every messaging platform has a test-send feature and every messaging platform sends broken campaigns, because a test send is an optional detour at the moment a marketer is most eager to be finished. The failure is not carelessness. It is that the safe path costs more than the risky one at exactly the wrong moment, and the interface never renders the consequence of skipping it.
The gate is unskippable and takes about twenty seconds. It renders the message against five real customer records chosen for variable diversity, so an unrendered token shows up as visibly broken rather than as a correct-looking placeholder. It diffs the audience against the previous send and states the change in plain language - this send reaches 84,000 more people than the last one, are you sure. It validates every link. In six months it caught 1,840 unrendered variables that would otherwise have shipped. Two percent of users back out at the gate, which is the number that matters.
THE INTERFACE CRAFT
The interaction, rendered as a working product surface.
The specimen below is code-native and uses the study's own design logic. The client interface remains protected.
MARTECH
MODEL CONFIDENCE
Paid searchHIGH
EventsLOW
UnknownDECLARED
DESIGN DECISIONS
Positions we would defend.
Each decision names the principle and the product consequence, not a stylistic preference.
Make the safe path the only path
An optional check at the moment of maximum eagerness is a check that does not happen. The gate is unskippable by design.
Real records, not placeholders
A preview with placeholder data cannot show a missing-field failure, which is the most common failure.
Twenty seconds is affordable
We measured the cost. Twenty seconds against a weekly brand-damaging error is not a trade-off requiring debate.
PRODUCT LEADER READOUT
What transfers, and what should remain specific to this product.
A case study is useful when its operating principle travels without turning the original interface into a template.
Read the operating condition
For Lifecycle messaging platform · 900 D2C brands, the transferable lesson is not a copied screen. It is the condition the interface had to make legible: Five actual customer records selected for variable diversity, including the edge cases with missing fields. Not a generic preview. Rebuild that visibility for your own roles, risk, terminology, and operating cadence.
Protect the design rule
An optional check at the moment of maximum eagerness is a check that does not happen. The gate is unskippable by design. Keep that rule in the acceptance criteria, component states, and production QA record so later visual cleanup cannot erase why the interaction exists.
Measure behaviour after ship
The evidence record is 0.3 for broken sends / month, from 14. Recreate the baseline and outcome window before rollout, segment the result by role and context, and state clearly what the measure cannot prove.
RESEARCH RECORD
The work behind the interface.
These artefacts connect the final interaction back to the evidence and product model that produced it.
Broken-send post-mortem
Reconstructed 14 broken sends across six brands; 11 would have been caught by rendering against a record with a missing field.
Test-send usage analysis
Instrumented the existing optional test send; used in 7% of campaigns, and almost never by the most prolific senders.
Record selection algorithm
Designed the five-record sample to maximise variable and edge-case coverage rather than picking at random.
Gate friction measurement
Measured campaign completion rate before and after the mandatory gate; completion was unchanged, errors fell to near zero.
“We used to have a broken send every couple of weeks. It has been six months.”
Head of CRM, D2C beauty · under NDA
NEXT STUDY · 13 / 30 · D2C & COMMERCE