Predicting the Why · ARP Survey
Practitioner survey.
16 questions, ~5 minutes. Your answers save instantly as you go — no submit button required. Leave and return; nothing is lost.
How saving works. Each answer is written to your browser the moment you choose it, with a timestamp. When the backend is configured, those same writes also go to the response database in real time — capturing the rhythm of how practitioners think, not just the final state. You can leave and come back; nothing is lost. Use the toolbar at the bottom to export your own copy.
About you
A little context first.
So we can tag your responses and follow up if you're open to it. Name and email are recommended; profession helps us interpret answers; company is optional.
—
Required
Your name
Saved
—
Required
Email address
Saved
—
Recommended
Your role or profession
Saved
—
Optional
Company or organisation
Saved
Thanks. Let's dive in.Four short sections — you'll be through in under five minutes.
Section A
Current-State Benchmarking
Which ABSD layers your system currently operates at.
01
Single select
Which best describes your primary professional domain?
Saved
02
Multi-select
Which of the following does your AI system currently capture or analyse?
Saved
03
Likert 1–5
How well does your current system understand why users behave the way they do?
1
2
3
4
5
Not at allVery well
Saved
Section A done.Three more to go. You're already shaping the data.
Section B
Signal Recognition Prevalence
Which of the seven behavioural signals you observe in your product.
04
Multi-select
Which of the following behavioural signals do you observe or track in your product?
Saved
05
Top 3 — tap in order
Select the top three signals (from the list above) that are most informative for understanding your users.
Tap to rank 1 → 2 → 3. Tap again to deselect.
·
Hesitation or pause patterns·
Rapid navigation or switching·
Re-engagement after inactivity·
Help-seeking behaviour·
Changes in interaction pace·
Regression to completed content·
Error tolerance patternsSaved
06
Open text
Are there behavioural signals available in your data that your team does not currently use? If yes, what prevents you from using them?
Saved
Halfway through.The next set is about interpretation — reading meaning into behaviour.
Section C
Signal-to-State Mapping
Whether practitioner inferences match the framework's mappings.
07
Single select
A user pauses for 20 seconds before completing a key action. In your domain, the most likely explanation is…
Saved
08
Single select
A user rapidly switches between three features without completing any task. This most likely indicates…
Saved
09
Single select
A user who previously engaged daily has reduced to weekly sessions over three weeks. The most likely driver is…
Saved
10
Likert 1–5
How often does the same behavioural signal mean different things depending on the context?
1
2
3
4
5
RarelyAlmost always
Saved
You're nearly there.One last section on the framework itself — then a quick yes/no.
Section D
Framework Utility
After a brief introduction to ABSD.
11
Likert 1–5 · framework intro first
How useful would the ABSD layered structure be for your product or analytics work?
The ABSD framework proposes four layers of user understanding: Action (what users do), Behaviour (patterns in how they do it), State (what they feel), and Drive (why they do it).
1
2
3
4
5
Not usefulVery useful
Saved
12
Single select
Which ABSD layer represents the biggest gap in your current system?
Saved
13
Single select
If your system could reliably infer user states, the primary way you would alter its response is…
Saved
14
Single select
The single biggest barrier to implementing deeper user understanding in your system is…
Saved
15
Single select
Would you use the ABSD framework as a reference in your product design or analytics discussions?
Saved
Section E
Stay in touch?
One final thing — then you're done.
16
Optional · Multi-select
Would you like to hear back when the paper is published?
Saved