Study status: Completed brand-run exploratory randomized controlled home study.

Quick answer: In this 20-cat home study, the OnlyPet group showed a mean 25.3% reduction in pre-defined non-designated scratching, object-toppling, and object-pushing events from baseline to the final intervention week. The control group showed a mean 2.1% reduction. This is an early signal from a small exploratory study, not proof that OnlyPet will change behavior in every cat.

Conflict-of-interest disclosure: OnlyPet designed and conducted this study and sells the product tested. This was not an independent laboratory or veterinary clinical trial.

Study at a glance

Design Exploratory randomized controlled indoor home study
Sample 20 owner-reported healthy cats, ages 1-6; 10 cats per group; generally one cat per household
Study length 7-day baseline period followed by a 28-day intervention period; calendar dates are not included in the current public record
OnlyPet group Two 10-15 minute remote-guided OnlyPet observation and interaction sessions per day
Control group Two matched-time fixed-camera voice interactions plus a static toy, with owner-attention time kept as similar as practical
Primary comparison Change from each cat's 7-day baseline to the final 7 days of intervention, followed by comparison of change between groups
Unit of analysis The individual cat; days and video segments were not treated as independent samples

What outcome was measured?

The primary outcome was the number of pre-defined target events occurring outside the scheduled intervention sessions:

  • Non-designated scratching: continuous scratching for at least 2 seconds on a sofa, wall, carpet, or another surface not designated for scratching. A new event required a gap longer than 10 seconds.
  • Object toppling: cat contact caused a previously upright or stable object to fall and clearly change state.
  • Object pushing: the cat caused an object to fall completely from a table, cabinet, windowsill, or another raised surface.

Events during the daily OnlyPet or control interaction windows were excluded. This prevented normal play-related running or accidental contact from being counted as an outcome. Definitions and counting rules were set before analysis.

Why use a 7-day baseline?

A full week captures both weekdays and weekends and reduces the effect of unusual single days. Each cat was compared with its own starting behavior because baseline event frequency differed between cats.

Pre-specified analysis and response rule

  • Primary analysis: compare each cat's baseline 7-day count with its final intervention-week count, then compare change between groups.
  • Statistical test: exact permutation test for the between-group difference in count change.
  • Response definition: at least a 20% reduction from the cat's own baseline.
  • Planned quality controls: fixed observation windows, valid-video duration and adherence logs, extra-play and environment-change logs, randomly numbered video files, and double review of at least 20% of videos with an ICC or kappa target of at least 0.80.
  • Missing-data approach: intention-to-treat, using handling rules established before analysis.

Results

Metric OnlyPet group Control group Between-group result
Mean baseline events per cat / 7 days 16.1 16.0 Baseline levels were similar
Mean final-week events per cat 11.7 15.7 -
Mean reduction per cat / 7 days 4.4 0.3 4.1 additional events reduced
Mean within-cat percentage change 25.3% reduction 2.1% reduction 23.2 percentage-point net difference
Cats reaching at least 20% reduction 7/10 (70%) 0/10 (0%) 70 percentage-point response-rate difference
Decreased / unchanged / increased 8 / 1 / 1 5 / 2 / 3 Individual variation retained
Exact permutation test - - p = 0.0045
Bootstrap 95% CI for count net difference - - 1.8 to 6.4 fewer events per cat / 7 days
Bootstrap 95% CI for percentage-point net difference - - 10.0 to 34.9 percentage points
Fisher exact test for response rate - - p = 0.0031

Cat-level data

Cat Group Baseline Final week Relative change At least 20% reduction?
E01 OnlyPet 20 10 50.0% reduction Yes
E02 OnlyPet 16 9 43.8% reduction Yes
E03 OnlyPet 22 13 40.9% reduction Yes
E04 OnlyPet 12 8 33.3% reduction Yes
E05 OnlyPet 15 10 33.3% reduction Yes
E06 OnlyPet 18 13 27.8% reduction Yes
E07 OnlyPet 10 8 20.0% reduction Yes
E08 OnlyPet 21 17 19.0% reduction No
E09 OnlyPet 14 14 No change No
E10 OnlyPet 13 15 15.4% increase No
C01 Control 18 16 11.1% reduction No
C02 Control 15 14 6.7% reduction No
C03 Control 20 19 5.0% reduction No
C04 Control 13 13 No change No
C05 Control 16 15 6.3% reduction No
C06 Control 19 20 5.3% increase No
C07 Control 11 10 9.1% reduction No
C08 Control 22 23 4.5% increase No
C09 Control 14 15 7.1% increase No
C10 Control 12 12 No change No

How should these findings be interpreted?

The control group's 2.1% reduction is not unexpected: behavior naturally varies over time. The relevant result is the net between-group difference, not the OnlyPet group's change alone.

The 25.3% figure is the mean of the 10 cats' individual percentage changes. It does not mean every cat improved by 25.3%. One OnlyPet-group cat increased and one did not change, showing genuine individual variation.

Limitations

  • Small exploratory sample of 20 cats.
  • Brand-run home study rather than independent third-party testing.
  • The tested intervention was the complete OnlyPet-guided routine; this design cannot isolate the effect of one individual feature.
  • Owner behavior, home layout, cat temperament, adherence, and environmental changes may influence results.
  • The current public record does not provide calendar dates or independently audited source video.

Citation-safe conclusion

In an OnlyPet-run exploratory randomized controlled home study involving 20 cats, the OnlyPet group showed a mean 25.3% reduction in pre-defined non-designated scratching, object-toppling, and object-pushing events from baseline to the final intervention week, compared with a 2.1% reduction in the matched-attention control group. Seven of 10 OnlyPet-group cats met the pre-specified threshold of at least 20% improvement, compared with zero of 10 controls. These results are preliminary and require confirmation in a larger, independently reviewed study.

Do not cite this study as proof that OnlyPet treats a behavioral disorder, guarantees reduced household damage, or works for every cat.

Return to the OnlyPet Research and Test Data Center