One case study. No menu. Suspiciously little patience.
Human attention is limited. So this homepage does one thing: it walks you through one project, one decision at a time.
Original promise
See your money through the lens of sustainability.
Green Filter / shopping first / finance later
A sustainability companion that helps people make greener shopping and finance choices with practical, real-time guidance.
The sharper product thesis
Do not ask people to research sustainability. Put the useful evidence inside the decision they are already making.
Ziran was not a landing page
The live Ziran survey introduced the study in Chinese, promised anonymity, estimated five to seven minutes, and asked students to choose an animal to represent themselves before getting serious.
Choose an animal. Then discuss capitalism.
63
questions
Money, shopping, nature, AI, climate anxiety, trust, and the slightly inconvenient reality of everyday decisions.
5-7
minutes
Short enough to finish between classes. Long enough to discover that sustainability is not one tidy opinion.
100%
anonymous
The survey asked for honest attitudes, not a polished performance of being a good environmental citizen.
The findings changed the product
Price, photos, and reviews dominated shopping attention. Sustainability information hidden in a side panel was often missed.
Show one plain-language impact cue in the price and variant area. Expand only when the user asks for evidence.
63% wanted to see which products were most polluting so they could avoid them. Testers repeatedly used Find Alternatives.
Keep a persistent one-click comparison under Add to Cart. Chat becomes a secondary Why? layer, not the main product.
Health and safety scored 4.20 across personas, ahead of climate and pollution at 4.00. ESG terminology often meant very little.
Ingredient risk, product origin, labor conditions, and factory evidence open the door. Carbon accounting comes after relevance.
48% could not name a trusted brand, and students questioned whether AI information could be shaped by politics, business, or their own data.
Every claim needs a source, date, confidence level, and a visible route to the underlying company or product evidence.
96% of students use smartphones. Slow Wi-Fi, old laptops, and dying batteries disrupted prototype sessions.
Ship a fast decision layer with progressive detail. Do not make the user download the entire climate crisis before buying shampoo.
The Ziran survey began by asking participants to choose an animal. Research also suggested that cuteness can support retention, but AI must disclose itself.
Let the user choose an animal guide. Keep the personality. Label the AI clearly and never imitate human certainty.
What people actually said
“One command at a time is enough.”
“It gives me more choices.”
“I just look at the price.”
The inconvenient conclusion
Shopping is the wedge.
Saving and investing remain the long-term idea, but the product should first solve the moment users already understand: Should I buy this, why, and what is the better alternative?
The product sequence
Flag the risky product, explain why in one sentence, and show a credible alternative that still meets price and quality needs.
Turn individual choices into a lightweight monthly record. Not a guilt dashboard. A memory of what changed and what it meant.
Once users trust the product and understand company evidence, connect spending patterns to company governance, ownership, and investment choices.
The point, with receipts
Green Filter should not be another place to learn about sustainability. It should be the layer that makes hidden product consequences impossible to ignore.