PintellyPintelly

Dashboard Workflows

Select the project before investigating data or changing settings. Available actions depend on your project role and plan; a feature's absence or empty state does not by itself indicate a collection failure.

Surveys

For setup, expected behavior and troubleshooting, see Surveys.

  1. Create a survey in Surveys. New surveys can remain in draft while you prepare questions.
  2. Add single-choice, multiple-choice, 1–5 rating, NPS (0–10) or free-text questions. Mark required questions and add English/Turkish translations where needed.
  3. Choose a page-path pattern and display delay in seconds. * matches any sequence of characters; an empty pattern applies to every page. For example, /checkout* targets paths beginning with /checkout.
  4. Activate the survey; pause it to stop new displays.
  5. Review impressions, responses, response rate and question results. Export responses as CSV for further analysis.

Questions appear one step at a time. Visitors who complete or dismiss a survey are remembered in that browser's local storage; this is not a cross-device guarantee. If storage is unavailable, the widget avoids displaying it because repeat suppression cannot be ensured. When several surveys match, the first eligible survey is selected.

Results include option counts, rating/NPS distributions and text answers. NPS is the percentage of promoters (9–10) minus detractors (0–6); 7–8 are passive. Response rate compares submissions with impressions, not with all site visitors. Survey answers are stored with page path and time rather than analytics identity.

Performance

For setup, expected behavior and troubleshooting, see Performance.

Use Performance to investigate real-user FCP, LCP, TTFB and load measurements by page and period. FCP marks the first content paint; LCP describes the largest content paint; TTFB measures time to the response's first byte. Where p75 is shown, it is the value at or below which 75% of measured observations fall.

Compare the same page, time range and population when evaluating a change. Missing timings can reflect browser support or incomplete observations, and are not zero-duration measurements. The AI agent performance tool can read these diagnostics too.

SEO insights

For setup, expected behavior and troubleshooting, see SEO Analysis.

Enable SEO Analysis for on-page collection, then use SEO Insights to inspect page scores, issues and analysis history. Checks cover title/description, headings, alternative text, canonical and indexing signals, structured data and content quality. Crawl-related information includes robots and sitemap checks where available.

Connect Google Search Console through Settings → Integrations to add search performance and keyword context. Search Console metrics and Pintelly page checks answer different questions; a page score is not a promised search ranking. Check the analysis timestamp and collected page coverage before treating results as current or site-wide.

Experiments and version comparison

For setup, expected behavior and troubleshooting, see Experiments and Versions.

Define an experiment with variant keys, a control and a primary metric: element click, page view or custom event. An optional guardrail metric measures a secondary outcome. Draft, running, paused and completed states control the experiment workflow.

Tag the variant actually shown to a visitor using the configured experiment tag or pb_variant / pb_tag URL parameter. Exposure measurement is consent-gated. Define and serve the variation in your site; an experiment definition alone does not allocate visitors or change your page content.

Review exposures, metric hits, conversion rates, lift, confidence intervals and significance where available. Device segments and sample-ratio mismatch (SRM) help identify uneven populations. Low sample counts or a flagged mismatch require investigation before declaring a winner. Version comparison can help relate product changes to outcomes without proving causation by itself.

Anomalies

Use Anomalies to investigate recorded difficulty signals alongside page activity and feedback. Repeated or unusual interactions are diagnostic hints; corroborate them with replay or a reproducible issue before concluding that the interface is broken.

Monitoring

For setup, expected behavior and troubleshooting, see Monitoring.

Monitoring checks website and service health independently of visitor analytics. Supported monitor types include HTTP, API, SSL, domain expiry, cron heartbeat, DNS, TCP port and ping.

Choose the target and available interval/validation options. Cron monitoring expects periodic heartbeats and supports a grace period; SSL and domain checks report certificate validity or renewal timing. Review recent checks, response times, availability and incident detail. A paused monitor is not performing active checks, and missing history is not evidence of continuous uptime.

Integrations and team access

  • GitHub: connect an account, select a repository for the project, create issues from feedback, and review issue activity. Issue import can be enabled for the linked project.
  • Slack: connect the desired workspace/channel for notifications.
  • Google Search Console: connect an accessible property for search context.
  • Team: invite members and assign project roles. Custom roles provide permission levels for individual areas; choose access based on the person's responsibilities.
  • API tokens: create scoped tokens for assistants; revoke them when access is no longer needed. See AI agents.

Reports, usage and retention

Account settings include timezone, language and notification preferences. Report frequency can be daily, weekly, every 14 days, monthly or off. Use a consistent timezone when comparing dates.

Settings → Usage shows current consumption and allowances. Features and retained history vary by plan; do not assume changing plans will restore expired data. When a report is empty, check the selected project, filters, collection settings, consent, plan access and retention before concluding that no visitors performed the action.