Smart insight, faster decisions
中文Book a demo
Case studies (anonymized)

Every case is a complete evidence chain

Client names are anonymized; the process and outcomes are intact. Challenge → approach → results, every step reproducible.

01
Home AppliancesProducts used · TrendAI · ezData

A leading home-appliance brand · water-softener category

Challenge

A high-value, high-repeat but slow-selling niche category. Years of conventional research only surfaced ideas the client already knew — framed surveys cannot reveal opportunities outside the client’s own assumptions.

Approach

Skipping framed surveys, the team mined authentic social UGC: TrendAI captured and parsed ~100k posts; ezData cleaned them to 50k quality texts. A co-created 1–4 level coding frame quantified the unstructured text, then crosstab analysis peeled down to segments, scenarios and purchase barriers — reverse-engineering product and channel strategy.

✦ VOC parsing accuracy
Quick Decision AI parsing86–92%
Industry norm~60%
Results
  • VOC parsing accuracy 86%–92% (industry norm ~60%)
  • Surfaced entirely new usage scenarios the client never anticipated
  • Pinpointed the core purchase barrier — extending filter life from 6 months to ~2 years
  • The coding frame became a reusable, transferable data asset
Read the full teardown →
02
Consumer ElectronicsProducts used · ezChat · ezData

A global PC & smart-device brand

Challenge

Around 75% of growth came from overseas, yet frontline retail staff were dealer employees — so HQ never heard the real in-store signals. Flying in was costly and partial; dealer relay was slow and distorted.

Approach

ezChat asynchronous interviews collected daily frontline feedback (text / voice / photo) from 15 countries at minimal effort. AI multimodal processing turned multilingual voice and store photos into readable text; ezData and AI opinion clustering distilled cross-market deal-drivers and local differences — then made the channel a standing daily feed.

✦ 15-country frontline channel
Each tile = one country's frontline feed, lighting up daily
Results
  • Frontline feedback from 15 countries aggregated and parsed daily
  • Distilled deal-drivers (young buyers, accessory completeness, gaming-meets-work) and market-level nuances
  • Established a standing intelligence channel reaching 15 markets without dealer relay
Read the full teardown →
03
Food & BeverageProducts used · ezTalk

A major food & beverage brand · new-concept screening

Challenge

The brand had 30 new-product concepts across multiple innovation briefs. It needed to identify promising directions quickly — and know exactly how to improve the winners, not merely receive a ranking.

Approach

The team first defined what “promising” meant and the action attached to each outcome. ezTalk then used sequential monadic testing so each consumer saw only one concept from a given brief, pairing comparable scores with open feedback and a Keep / Revise / Watch out recommendation for every concept.

Results
  • 30 concepts screened against one shared definition of success
  • 300 target consumers tested, yielding analyzable samples per concept
  • Every winning concept received actionable Keep, Revise and Watch out guidance
Read the full teardown →
04
Internet & RetailProducts used · ezTest · ezData

An e-commerce platform · young-male BHT

Challenge

Sales among men aged 20–30 were declining, but the platform did not know where those customers were going or why its existing short-form video communication was failing to stick.

Approach

BHT tracked the brand and competitors among the target audience, first locating competitive loss and then comparing recall, relevance and specific content cues — turning a sales decline into a communication problem that could be addressed.

Results
  • Identified men aged 20–30 as the audience requiring priority recovery
  • Showed that competitor content was closer to men’s interests — a key driver of loss
  • Triggered a shift from generic fashion messaging toward more male-relevant video content
Read the full teardown →
05
Food & BeverageProducts used · TrendAI · ezData · ezSummary

An international seasoning brand · sauce innovation

Challenge

The brand wanted to find new sauce opportunities in real market demand, but flavour signals were scattered across dishes, regions and consumer language — too fragmented for conventional desk research to prioritize.

Approach

TrendAI assembled restaurant data and filtered for opportunity dishes. AI then coded flavour, ingredient, staple texture, usage and sauce texture in layers; regional preference and social verbatims converted fragmented dish data into developable “usage × staple × flavour” combinations.

Results
  • 78,005 restaurant-data records entered opportunity screening
  • 17,006 dishes formed an analyzable opportunity pool
  • 4,607 “priority flavour × priority province” records received deep AI coding
  • Produced a structured map of flavour needs, representative dishes and development priorities
Read the full teardown →
06
Beauty & Personal CareProducts used · ezTalk

A body-care brand · elevator TVC diagnosis

Challenge

Before launch, the brand needed to know which of four elevator TVCs would be remembered, support purchase intent and build brand appeal — and how to improve the creative, not simply select the prettiest film.

Approach

Single viewing recreated the elevator context. The study first captured unaided brand, product and creative recall, then had the same respondents complete forced comparison — combining recall, appeal, purchase intent and brand appeal into an actionable creative diagnosis.

Results
  • 4 TVCs diagnosed and compared under a realistic single-exposure condition
  • Identified the creative direction with the strongest combined KPI performance
  • Translated recall, message-delivery and credibility issues into optimization guidance
Read the full teardown →
07
Home AppliancesProducts used · ezTalk

A home-appliance brand · post-launch frontline interviews

Challenge

Two laundry-care launches were underperforming. Headquarters needed fast, direct feedback from store promoters and consumers to determine whether the issue was product, in-store presentation, selling script or brand trust.

Approach

ezTalk interviewed frontline promoters from stores with and without completed sales, cross-checking product competitiveness, consumer feedback, competitors, in-store barriers and brand image before translating the findings into product, retail and marketing actions.

Results
  • 82 frontline promoters completed valid interviews across two launches
  • Included both converting and non-converting stores, not only the success stories
  • Located conversion barriers in price, ease of use, appearance and in-store demonstration
  • Delivered next actions across product, retail execution and marketing
Read the full teardown →
08
Home AppliancesProducts used · ezTest

A home-appliance brand · new-product e-commerce simulation

Challenge

During new AC development, the brand needed to see attraction, conversion barriers, design and benefit gaps against its core competitor in a realistic e-commerce path — not just in a static survey.

Approach

E-commerce simulation captured click, save, cart and purchase behaviour; design, benefit and PSM price testing then located full-funnel conversion barriers and informed product definition and communication.

Results
  • 300 target consumers completed simulation and evaluation
  • Located the competitor gap in e-commerce attraction and conversion
  • Turned design, price and benefit gaps into specific optimization actions
Read the full teardown →
09
Beauty & Personal CareProducts used · ezTest

A premium haircare brand · brand-health study

Challenge

An emerging premium haircare brand had gained online growth, but did not know whether traffic was becoming durable brand equity. The team needed to test whether its positioning was truly perceived and find purchase barriers and marketing actions.

Approach

The BHT compared brand and competitors across awareness, consideration, purchase, first choice, brand associations and marketing reach, then connected purchase drivers and barriers to brand-building and marketing action.

Results
  • 1,000 target consumers established the brand-equity baseline
  • Showed that online growth had not automatically created a clear, distinctive brand association
  • Identified professional efficacy and category breadth as priority brand-building directions
Read the full teardown →
10
Food & BeverageProducts used · ezTest

A seasoning brand · Lunar New Year campaign pretest

Challenge

Before Lunar New Year launch, the brand needed to select the combination of storyboard, creative idea and KV that would engage consumers while clearly carrying brand and product value.

Approach

Different stories, creative ideas and KVs were tested in groups. Appeal, purchase intent, brand impression and clarity were compared, then the reasons behind winning content were unpacked by audience.

Results
  • 500 seasoning decision-makers tested five creative groups at 100 respondents each
  • The winning story direction lifted purchase intent by 18%
  • Identified the preferred character and creative combination balancing clarity, brand association and purchase intent
Read the full teardown →
11
Health & WellnessProducts used · ezTest

A health brand · short-drama marketing effectiveness

Challenge

Short-drama reach does not equal brand memory. The brand needed to assess drama appeal, brand recall and placement quality among genuinely exposed viewers, finding content worth further investment and placements that work.

Approach

The study followed real exposed audience packs: using 5+ seconds of valid exposure to approximate the delivery environment, then comparing drama appeal, brand recall, key-message memory and placement formats.

Results
  • 8 short dramas and 800 valid-viewer samples were assessed
  • Unaided brand recall was 22%, rising to 46% with a prompt
  • Plot-critical-object placement reached 70.8% average recall
  • Produced guidance for drama selection and placement optimization
Read the full teardown →
12
Home AppliancesProducts used · ezTest · ezTalk

A home-appliance brand · ducted-AC strategic NPS

Challenge

The brand needed to rebuild ducted-AC NPS from the user perspective: not only knowing the recommendation score, but what drives it across awareness, purchase, installation, use and after-sales — and which needs deserve priority investment.

Approach

JTBD mapped full-journey measures. Quantitative NPS and competitor comparison came first, followed by online and in-home interviews to validate causes and unmet needs; “importance × performance × KANO” set improvement priority.

Results
  • 720 quantitative responses, 131 online qualitative interviews and 12 home visits formed one evidence loop
  • Established a JTBD-NPS system spanning awareness, consideration, purchase, use and after-sales
  • Classified improvements as must-solve, differentiation opportunities or basic maintenance
  • Converted recommendation drivers into investment priorities across product, service and marketing
Read the full teardown →
Industries served
Home AppliancesConsumer ElectronicsBeauty & Personal CareFood & BeverageHousehold CareHealth & NutritionInternet & RetailAutomotive & Mobility

Want a full teardown for your industry?

Smart insight, faster decisions
Book a demo