Closing the loop between collecting experience data and acting on it.

Treat questions as first-class objects in a library.

A brand-marketing director who needs to know what their guests actually think isn't a survey designer. They have a meeting in 20 minutes and they need to ship a question that returns useful data. Question Builder is the surface that gets them there.

The product treats questions as first-class objects in a library. Operators build a question once, name it, configure it, translate it, and reuse it across every experience the brand runs. Standardized question types (NPS, Marketing Opt-in, Purchase Behavior, Birthdate) come with industry-validated wording so the data they collect is comparable to peer benchmarks downstream. Custom questions cover everything else, with response formats sized to what the operator is actually asking.

The latest enhancement adds conditional logic on top of the library: sub-questions branching off a parent answer with an AND/OR composer for compound rules.

• 01
  Library as the home. Every question authored is a saved entity with a name, type, preview, and the count of experiences using it. Authoring becomes inventory, not one-off work.
• 02
  Standardized types unlock benchmarks. NPS, Marketing Opt-in, Purchase Behavior, and Birthdate ship with industry wording. 84% of AnyRoad customers use them by default, which is what makes cross-customer comparison possible in Industry Benchmarks downstream.
• 03
  Custom for the long tail. Six response formats (Free Text, Multi-Select, Radio, Checkbox, Dropdown, Date Input) for everything the standardized library doesn't cover.
• 04
  Translations as a first-class concern. Spanish, Mandarin, and other languages live on the same screen as the source. When the source changes, an "Update Language Translations" notification fires automatically.
• 05
  Lifecycle without surprises. New, Rename, Duplicate, Archive, Restore. Destructive actions (deleting collected data) require explicit confirmation.
• 06
  Conditional logic as the latest layer. Sub-questions, conditions, and AND/OR composition for the authoring patterns that needed to grow up.

From thousands of comments to a handful of topics that matter.

A single mid-sized brand could collect tens of thousands of free-text responses a year. Reading them was impossible. Tagging them was a part-time job. The hard part wasn't sentiment. It was knowing which topics were worth a meeting and which were noise.

Pinpoint AI runs every response through a topic-clustering model and ranks the resulting topics by NPS impact, not just frequency. A topic in the negative column flagged "Critical" means it's not just complained about. It's measurably depressing the brand's score. Every topic is one click from the verbatims that drove it, with the full guest profile attached.

• 01
  Two columns, two stories. Positive and negative topics are split by default, so operators see both what's working and what's hurting on the same screen.
• 02
  NPS impact, not response count. Topics are sorted by how much they shift the brand's score, so a small loud problem ranks above a big quiet one.
• 03
  Severity badges. Critical / Mild labels surface on each verbatim, so an operator can triage urgency at a glance without reading every comment.
• 04
  Pre / post NPS in the margin. Topic Details shows each guest's expectation versus their actual rating, turning a verbatim into a measurable delta.

The four surfaces above let a brand understand themselves. Industry Benchmarks lets them understand themselves relative to peers. NPS, brand conversion, opt-in rate, revenue per visitor: every operator's first question is whether their number is good or just a number. Benchmarks answers it.

The surface overlays anonymized peer data on top of a brand's own metrics, sliced by industry, experience type, demographic, and season. A whisky distillery sees how their post-tour NPS compares to the alcohol category average. A CPG brand sees how their festival opt-in rate compares to the industry. The number stops being abstract and starts being directional.

• 01
  Peer overlay. Every brand metric carries an industry benchmark by default, so the operator immediately sees over- and under-performance.
• 02
  Slice by what matters. Compare against the right peer group: industry, experience type, season, demographic. Different cuts surface different gaps.
• 03
  Privacy-preserving by construction. Benchmarks are aggregate and anonymized. Brands contribute to a shared baseline without exposing individual data.