← Back to Portal
Dataset 03 · Observation Method

Floor Observation Data

Structured field observations of associate-customer interactions at Rosie Home Tech retail display locations across 12 partner stores.

Interactions Observed62 Sales Interactions
Locations12 Retail Partners
Observation Hours84 Hours Total
MethodNon-Participatory Observation
Section A

Observation Protocol & Overview

Two trained observers visited 12 retail partner locations over a six-week period. Observers posed as general shoppers and used a structured observation checklist to record associate behaviors during customer interactions at the Rosie display area. 62 complete interactions (from initial customer approach to departure) were observed and coded.

62
Complete interactions observed
4.2
Avg interaction length (minutes)
11%
Resulted in observed purchase
76%
Associate spoke first without asking questions

The observation checklist tracked 12 specific behaviors across four categories: conversation initiation, product knowledge delivery, objection handling, and conversation close. Each behavior was recorded as observed / not observed / partially observed for each interaction.

Section B

Behavioral Frequency Data

The following table summarizes the frequency with which each observed behavior was recorded across all 62 interactions.

Observed BehaviorFrequency% of Interactions
Associate referenced product card or signage during explanation4877%
Associate opened with product information (not a question)4776%
Associate asked at least one open-ended question about customer's household1118%
Associate recommended a specific model with a stated reason1423%
Customer asked a question about AI data or privacy3861%
Associate provided a substantive answer to AI privacy question46%
Associate redirected or gave vague response to privacy question3455%
Associate asked about customer's existing smart home devices711%
Associate mentioned Rosie Elite's tutoring feature unprompted915%
Associate paused to let customer respond after recommendation813%
Associate provided a clear next step or call to action at close1219%
Interaction ended with customer appearing interested but undecided4166%

Key Finding: In 61% of observed interactions, customers raised AI privacy questions. In only 6% of those cases did the associate provide a substantive response. The gap between customer need (information about privacy) and associate behavior (vague reassurance or redirect) is the single most consistently observed failure pattern.

Section C

Representative Field Notes

The following are condensed field notes from select observed interactions, representative of the patterns identified across the full dataset.

Observation 07 · Location C · Rosie Pro Display
Associate approached customer immediately and began describing Rosie Pro features using the product card as a visual reference. Customer listened for approximately 90 seconds, then asked: "Does it record what I say in my house?" Associate responded: "It only listens for the wake word, so you're good." Customer nodded but appeared unconvinced. No follow-up question from associate. Customer thanked associate and left 40 seconds later.
AI privacy question raised · Response: vague · Customer departed without purchase
Observation 19 · Location F · Rosie Elite Display
Customer approached display with a child (approximate age 8–10). Associate asked: "Are you looking for something for the whole family?" Customer said yes and mentioned homework help was a priority. Associate explained the K–12 tutoring feature in detail, described the learning profile system, and then asked how old the child was and what grade. Customer became notably more engaged. Asked three follow-up questions about tutoring. Did not purchase but asked for associate's name and said she'd be back with her husband.
Open question asked · Tutoring feature highlighted · Customer highly engaged · Likely return visit
Observation 33 · Location A · Rosie Classic Display
Associate began with: "This is our entry-level model, the Classic, it handles cleaning, cooking, and laundry." Then listed six additional features from the product card without pausing. Customer asked: "Which one do most people get?" Associate said: "Pro is popular." No explanation of why. Customer asked about price difference between Classic and Pro. Associate listed both prices. Customer said "Okay, thanks" and left. Total interaction: 3 minutes 10 seconds.
No questions asked · No recommendation rationale · No next step offered · Customer departed
Observation 51 · Location J · Rosie Pro Display
Customer asked immediately: "What does Rosie do with my grocery data — like does she sell it?" Associate said: "I don't think so, I mean, she just uses it to order stuff." Customer frowned and said: "You don't think so?" Associate laughed nervously and said: "I'm pretty sure it's safe." Customer said he wanted to look it up online first and left. Privacy policy was visible on the display stand but was not referenced by the associate at any point.
Privacy question raised · Associate response undermined confidence · Resource not utilized · Customer departed
In 77% of observed interactions, associates referenced product cards or display signage — indicating heavy reliance on external aids over internalized product knowledge
Only 18% of associates asked an open-ended question about the customer's household before beginning their product explanation
Interactions that included at least one open-ended question averaged 6.8 minutes vs. 3.1 minutes for interactions without — and were 3x more likely to result in a purchase or stated return intent