> For the complete documentation index, see [llms.txt](https://boundaryai.gitbook.io/boundaryai-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://boundaryai.gitbook.io/boundaryai-docs/whats-new.md).

# What's New

New features, improvements, and fixes in BAI Analytics.

The latest features, improvements, and fixes in BAI Analytics. Dates reflect when each update went live in production. Have feedback or a request? [Get in touch](/boundaryai-docs/help/getting-help.md).

{% updates format="full" %}
{% update date="2026-09-16" %}

##

**New**

* Ask your feedback: an in-app assistant that answers questions about your workspace and your customers' feedback from your own data, with clickable citations, saved conversations, and page-aware suggestions. Available when enabled for your organisation.
* SurveyMonkey joins the native connectors, with real-time delivery of new responses when you connect with a client ID and secret.
* Deliver feedback files over SFTP: BAI Analytics generates the key pair, you register the public key with your vendor, and new files are imported on a schedule. Medallia connections can now choose between the Query API and file delivery.
* The Integrations page is organised into Connections, AI assistants, and Developer tools, with your connections grouped by category and per-sync controls.
* Connect several accounts of the same platform (for example one Zendesk per region) and pick which one to import from inside a Feedback Group. The Add connection catalogue is searchable.
* Connect Claude or ChatGPT by signing in: choose your assistant under AI assistants, sign in, approve read-only access, and ask your first question. A publish kit walks workspace administrators through ChatGPT, Claude, Gemini Enterprise, and Microsoft 365 Copilot.
* Import a batch of spreadsheets with one shared column mapping; each file becomes its own source.
* Merge sub-themes and monitor sub-categories the same way you merge themes, with an optional new name.
* Grouped theme and monitor pages redesigned around a source-by-source view: presence, sentiment, and comparison with the overall picture for every source.
* API: synchronous bulk push returns the completed result in the same call, and a new MCP tool lets assistants quote the real comments behind any grouped theme.

**Improved**

* Grouped themes keep the same names from one period to the next and match the Overall view on tracked groups.
* Grouped theme and monitor pages gather the short and detailed summaries in one Summary area: use **Generate detailed summary** once, then switch between **Overview** and **In depth**.
* Executive summaries and action plans only conclude that no action is needed when the negative figures support it.
* Canvas editor navigation: one page dropdown with a thumbnail strip, a page actions menu, and a labelled back button.
* The coding-matrix export lists columns alphabetically with no duplicated All columns.
* The API reference names the host that actually serves the Input API on every operation, and `GET /me` returns the organisation name.

**Fixed**

* "Your site cited" counts every citation in an answer and only divides by answers that cite a source.
* Applying a monitor while a source's analysis is still running no longer reports zero matches; it waits for the analysis to finish.
* Changing a question's type no longer leaves a blank result card or a false "no topics found".
* A single network blip during a multi-file upload no longer discards the whole batch.
* `source_reference` on bulk API pushes is stored on every row.
* Source comparison labels on grouped theme pages read **Above overall** and **Below overall**, so they no longer contradict the Sentiment column.
* Grouped theme PDFs print the same number of sources as the page.
* Report schedules that fail now record the error and pause after repeated failures instead of retrying silently.
  {% endupdate %}

{% update date="2026-09-08" %}

##

**New**

* Report designs: build reusable branded designs with your colours, fonts, logos, section order, and custom canvas pages, then pick one whenever you generate a report.
* Export any report as PowerPoint alongside PDF and Word, and download the underlying data as Excel.
* Share a report with a public link that needs no login, expires after 30 days, and can be revoked at any time.
* Source reports: select a single source and generate the full analysis of that source as a PDF saved with it.
* Bring your own fonts: drop font files or a folder into a design and apply the family in one click. PDFs embed them.
* Logos are saved with each report design, so every report using it prints them for everyone in your organisation.
* Remove any report section, including the cover, and add it back later with its settings intact.
* Multi-location benchmark report layouts, for organisations that have them enabled.
* MCP self-serve setup: create an assistant key, copy a ready-made configuration for Claude Code or Claude Desktop, and check access from the dashboard.

**Improved**

* Four brand colour roles (Primary, Accent, Text, Page background) drive every surface of a report; each section shows whether it follows the design or has a custom appearance.
* Canvas blocks snap to alignment guides while dragging and arrow keys nudge a selection.
* Organisation add-ons carry clear customer-facing names in Settings, and the Features tab is now called Analysis settings.
* Assistant keys inherit the access of the person who created them and stop working when that person leaves the organisation.
* MCP `list_themes` reads the same grouping the dashboard shows for time-tracked groups and can read a single period.
* Qualtrics incremental syncs now import only what is new; S3 credential errors are diagnosed more precisely.
* The interviewer's speech is kept out of analysed text and quotes for transcript and audio uploads.
* Monitor Library usage counts include monitors applied from the Edit monitors dialog.

**Fixed**

* Surveys with older rating questions that had an invalid range analyse normally again.
* Report designs applied by the BAI Analytics team no longer render as the benchmark layout in organisations without it.
* Background task notifications can be dismissed with the X.
* API analysis read-back for sources in time-tracked groups returned no figures; it now returns the aggregated result.
  {% endupdate %}

{% update date="2026-08-31" %}

##

**New**

* Medallia syncs start paused with a Check setup panel and a Preview that shows exactly what would be imported before anything lands in your group.

**Improved**

* AI Visibility is now an add-on enabled per organisation, alongside Web & Social listening.

**Fixed**

* Fetch new data and the source card work for every native connector source, not just support and work tools.
* The AI Visibility scorecard on a period page covers that period only, and existing AI Visibility data stays viewable when the add-on is off.
* Manage sources rows with nothing to open say "Not analyzed yet" instead of failing on click, and agent rows are clickable too.
* A cancelled period analysis no longer blocks later analyses on the group.
* The monitor trend shows only the group's current period granularity, so totals no longer double-count.
* Changing your interface language no longer blanks the account window.
  {% endupdate %}

{% update date="2026-08-24" %}

##

**Fixed**

* AI Visibility buttons that looked active but did nothing now take you where they say.
  {% endupdate %}

{% update date="2026-08-23" %}

##

**New**

* Native connectors for Qualtrics, Medallia, InMoment, Amazon Redshift, Amazon S3, Snowflake, Google BigQuery, and Databricks: create the connection once, then add scheduled syncs that feed any Feedback Group.
* One Medallia connection can feed several sources: add "Limit this source to" rules to route survey responses and reviews into different groups.
* Indeed joins the employer-review platforms in Social Listening.
* Mark a scraped comment as Not relevant to help tune your agent's filtering.
* Redesigned AI Visibility workspace: a verdict, what AI says about you, recommended next steps, and evidence by model, competitor, and cited source.
* Named phases: group feedback into chapters you name (Pre-season, Launch weekend, Playoffs) with an open last phase and a Start next phase button on the timeline.
* One place for periods: Periods and phases replaces the separate grouping and phase dialogs and tells you exactly what a change will reset before you save.
* Automatic period reports live under Reports on the group page, with a Period-close reports tab and an Automatic badge on the reports they produce.
* Choose who receives each monitor's alerts, per channel, instead of the whole roster.
* Monitor Library: applied monitors are frozen copies, Library edits are offered as an explicit Use update, Create draft replaces Save for later, and new monitors can be saved to the Library for reuse.

**Improved**

* Monitors with no matches in a period stay visible, greyed out, instead of disappearing.
* The Library shows a Back button and lists your own folders above Built-in and Uncategorized.
* Report source weighting moved under Report settings, and a period-close report's send delay is checked when you shorten a group's periods.
* Recordings whose language is hard to detect are re-checked in segments instead of being dropped.
* Agents keep same-category, different-brand content out of your imports.
* The product is consistently called BAI Analytics in all copy.

**Fixed**

* AI-suggested Glassdoor links are checked against the company, and the "suggested by AI" label clears once you edit them.
* A competitor's name can no longer be counted as a mention of your brand in AI answers.
* Library names are exact, so "Food quality" and "Food Quality" no longer collide.
* Manage sources no longer opens two upload windows at once.
* Phase editors stay usable with many phases.
  {% endupdate %}

{% update date="2026-08-18" %}

##

**New**

* Import a whole folder of recordings or documents in one action; subfolders become Folder and Subfolder filters.
* Click a source row in Manage sources to open it.

**Fixed**

* Import a whole folder opened the plain file picker.
* Long Amazon product URLs with tracking parameters are accepted.
* The AI Visibility wizard describes AI answers rather than social posts.
  {% endupdate %}

{% update date="2026-08-17" %}

##

**New**

* Organisation access replaces plans: Settings shows the add-ons enabled for your organisation and your monthly allowances. Plan tiers, upgrade prompts, and self-serve subscriptions are gone; access is managed with your BAI Analytics representative.
* Request demo access: people who sign in without an organisation can request an industry-tailored demo or contact us.
* Personal data redaction: remove emails, phone numbers, card numbers, names, addresses, and more from feedback as it arrives, organisation-wide or per Feedback Group.
* Track app and product reviews from the Apple App Store, Google Play, Amazon, and Glassdoor in Social Listening.
* A suggested reply is drafted for every scraped review or comment you choose to answer.
* Monitor Library folders: group, reorder, hide, duplicate, and export monitors, and import a CSV straight into a folder.
* Slack and Microsoft Teams alerts: add an incoming webhook as a third notification channel next to email and SMS.
* Compare segments side by side on any choice question, with percentage-point differences, sample-size warnings, and a CSV export.
* Change a question's type after upload from the card's 3-dots menu, without re-uploading.
* Weighting: weight by up to four fields, import your own respondent weights, and export computed weights with the metadata columns you pick.
* Upload several audio recordings at once and analyse them as one source, with recording language and enhanced transcript cleanup options.
* Surveys: close a survey automatically at a response count you set, schedule a send for later, and send automatic reminders to non-responders by email, SMS, or both.
* Surveys: embed a survey on your own website (inline, popup, or iframe) and redirect respondents to a page of your choice when they finish.
* Rename any period from its Details row; the name shows everywhere the period appears.
* One timeline for every group: all-time groups use the same timeline bar, Manage sources, Manage monitoring, and settings menu, with Track feedback over time one click away.
* Manage sources shows Create survey, Upload file, Add connector, Web & social, and AI Visibility directly.
* Regenerate a single theme's summary in one click.

**Improved**

* Agents filter every post and comment against your monitoring goal, and suggested replies are short and natural in the reviewer's language.
* Segmentation is offered on datasets made of open-ended questions plus metadata fields.
* Coding-matrix export: totals row, written English headers, and a note that counts are unweighted.
* AI-generation buttons acknowledge instantly and show as Queued in Background Tasks; repeat clicks no longer start duplicate runs.
* Grouped themes and AI insights are prepared as soon as a period's analysis finishes, so opening a period no longer starts a wait.
* New Feedback Groups start in your interface language, and the upload wizard shows which language the group's AI summaries will use.
* Simpler timeline choice at upload, and uploaded files are checked by content rather than extension.
* Removing a source is a clear Delete source action; nothing is left behind in a hidden state.
* Weighting wizard warnings are shown in your interface language.

**Fixed**

* Alerts fire for the first match after alerting is enabled, and for a new source's first batch.
* Monitor counts refresh correctly after a re-scan.
* Recompute grouped themes targets the period you are viewing.
* Renaming a source renames it everywhere, including period views.
* The prefilled name of an Overall report named the wrong months for groups east of UTC.
* Weighting no longer reports a false "targets cannot be met" when targets match the sample to rounding.
* Ranking answers can no longer carry stray Other text into results.
* The credit check before a preset agent run matches the estimate shown.
* The sources panel no longer takes up to 45 seconds to load right after publishing a survey.
* Grouped monitor source counts agree across the row, the drawer, and the detail page.
  {% endupdate %}

{% update date="2026-08-03" %}

##

**New**

* Surveys: two new question types, Dropdown and Ranking, plus an Answer format (Email, Phone number, Date) for Short answer questions.
* Statistical significance on segment breakdowns: chi-square and cross-tabs with confidence intervals for choice questions; group means and t-test or ANOVA for scales and NPS.
* Click any source on the Overall view to see which periods it appears in and jump straight to that period's analysis.

**Improved**

* API: request and response examples on every operation, exact Retry-After on rate limits, and every write states the key permission it needs.
* Single-speaker recordings and long paragraphs are split into passages so short monologues still get an overview.
* Every comment in a theme belongs to a sub-theme (an Others bucket catches the rest), and mixed-sentiment themes are re-audited automatically.
* Custom Monitoring is no longer offered on Short answer questions, which are never scanned.

**Fixed**

* The same customer quote no longer illustrates two different grouped themes in one report.
* Social replies are attached to themes by their own wording, not the parent post's caption.
* Connector-fed sources no longer expose a public answer link.
* Hour-long interview recordings can be uploaded.
* Grouped monitor source counts no longer count constituent entries twice.
* Full sub-theme names in coding-matrix headers, and numeric feedback values in the API returned as numbers.
  {% endupdate %}

{% update date="2026-07-22" %}

##

**New**

* SCIM 2.0 user provisioning: let your identity provider create, update, and deactivate BAI Analytics users automatically.
* Per-competitor Prominence in the AI Visibility comparison.
* Evolution view charts get an Auto / Full Y-axis toggle so real movement in NPS, sentiment, and scores is visible.
* Your organisation's logo, with an optional second logo, can appear in the sidebar lockup on request.

**Improved**

* The default Individual Surveys group displays in your interface language.

**Fixed**

* Opening an outdated analysis link lands on the latest analysis with an info banner instead of a permission error.
* In-depth summaries no longer render as one giant heading when line breaks were lost.
* Switching organisations no longer briefly empties a Feedback Group's sources.
* Regenerating a report from another tab or a schedule refreshes the open viewer.
  {% endupdate %}

{% update date="2026-07-20" %}

##

**New**

* Reply to any scraped review or comment straight from the analysis, with a guided handoff to the platform.
* Ongoing agents in tracked groups sync their schedule to the group's periods automatically.
* AI Visibility: visibility rate and share of voice trends per period, percentages and the prompt behind each answer, a per-model summary, recommended models pre-selected, and several phrasings of the market question sharing one answer budget.
* Monitor Evolution widgets show mention rate as a percentage of responses, with volume bars and low-sample markers.

**Improved**

* Accepting an invitation opens the organisation you were invited to, and links to another organisation's feedback switch you there automatically.
* Analysis settings simplified to two switches: multiple themes per comment and multiple monitors per comment.
* Brand mentions in AI answers recognise acronyms, short forms, and aliases, and citations resolve to the real source page.
* Date columns are detected reliably in English, French, and Spanish exports, and an upload can no longer be silently mis-dated.
* The dashboard's Continue card always reopens a group at its Overall view, and the onboarding Add monitors step unlocks as soon as a group exists.

**Fixed**

* Removed the dashboard Import data tile that led nowhere.
* Long-answer questions with zero text responses in a period show a clear empty state instead of looking like a failed analysis.
* Star ratings are shown only for review-platform sources, and the agent wizard opens each step at the top.
* Monitor re-runs on AI Visibility sources no longer get stuck on "analysis in progress".
* Changing a group's grouping no longer leads to a blank page, and groups with data no longer show the "Start tracking" placeholder by mistake.
* Feedback that arrived while a period was being analysed is picked up and analysed instead of silently skipped.
* Progress notifications show accurate counters and specific, translated messages.
  {% endupdate %}

{% update date="2026-07-15" %}

##

**New**

* Monitors now cover new sources automatically. Add a source to a feedback group and every monitor with auto-coverage on watches it from its very first analysis. Monitoring setup steps are gone from the upload, agent, and connector flows; there is nothing to configure at import time.
* Score monitors: get alerted when new NPS or rating scores match a condition you set (for example, 5 new NPS scores at or below 6). Pure number matching, so they use none of your AI allowance.
* Evolution Widgets: build your own metric-over-time cards (sentiment, volume, NPS, ratings, choice share, monitor mentions) on a tracked group's Overall view, with segment split, live preview, and per-view half or full width.
* AI Visibility: ask ChatGPT, Claude, Gemini, Perplexity, Grok, DeepSeek, and Mistral about your brand and analyse their answers, with a visibility scorecard and per-model drill-down.
* Import feedback from thousands of other apps by connecting Zapier to a BAI Analytics API endpoint.
* Put your own logo (plus an optional co-brand logo) on every report format, and save it as your default.
* Redesigned PDF reports for review-platform sources: ratings and trend, low-star reviews to answer, and a customer-voices wall.
* Upload JSON conversation exports holding one conversation or thousands, choose how they become rows, and join an Excel or CSV metadata sheet on a shared key.
* High-volume survey sending: 10,000 included invites per month, or connect your own AWS account and send under your own identity with no platform cap.
* Connect AI assistants to your feedback analytics through a read-only MCP endpoint, authenticated with an API key.

**Improved**

* Agents created from a tracked feedback group now default to Ongoing (weekly), and ongoing agents can set a custom start date for their first look-back.
* Email sending settings moved into the organisation profile, next to the other organisation-level options.
* Detached sources can now be deleted permanently from Manage sources.

**Fixed**

* Excel uploads keep original dates in more edge cases (1904 date system, custom-formatted numeric cells).
* Grouped themes no longer keep the old language after you switch a group's analysis language.
* Source labels and counts on feedback-group pages now match the underlying data.
* Audio transcription automatically retries transient provider errors.
  {% endupdate %}

{% update date="2026-07-09" %}

##

**Improved**

* Monitor detection is now sorted by relevance.
* Evolution periods now analyze and close automatically as new data lands, with no manual step required.
* Feedback-group weighting can be based on how much each source actually contributes, not just its metadata.

**Fixed**

* The dashboard's "Continue" card no longer points at a feedback group you've already deleted.
* Various Custom Monitoring reliability fixes.
  {% endupdate %}

{% update date="2026-07-06" %}

##

**New**

* Subscribe to get an automatic report email whenever a tracked Evolution period closes.
* Monitor alerts can fire in near real time as feedback arrives, instead of waiting for a full analysis run.

**Improved**

* Weighting now applies to in-depth AI results too (theme mentions, sentiment, criticality), not just raw counts.
* More reliable Custom Monitoring detection.

**Fixed**

* Restored full comment volume from Twitter/X and Facebook scraping.
* Grouped-theme summaries no longer go stale after incremental updates.
* Fixed PDF report previews failing to render for some surveys.
  {% endupdate %}

{% update date="2026-07-03" %}

##

**New**

* Sample weighting (redressement / post-stratification) for surveys and feedback groups, so results reflect your target population, not just your raw sample.
* Upload audio recordings or call transcripts for automatic transcription and analysis.
* Build survey questions that show or hide based on respondent metadata, not just a previous answer.

**Improved**

* Select several saved monitors at once when adding them from your library.
* Per-question response counts now show right on the Survey Result page.
* More accurate AI-usage metering.

**Fixed**

* Fixed slow loading on surveys with very large metadata breakdowns.
  {% endupdate %}

{% update date="2026-07-01" %}

##

**New**

* TikTok joins Social Listening: track videos and comments by account, hashtag, or search term.
* Scale and NPS questions can now display as a star rating.

**Improved**

* Smoother Custom Monitoring setup, with a library picker and background progress instead of a blocking spinner.
* Faster Evolution period switching, even on large feedback groups.
* Faster, more reliable sign-in.

**Fixed**

* Various survey, segmentation, and report-preview fixes.
  {% endupdate %}

{% update date="2026-06-28" %}

##

**New**

* Send survey invites directly from BAI Analytics, with your own sender identity or a shared address.
* Bulk-import monitor templates from a CSV or Excel file.
* Choose exactly which columns to include in a Raw Data Excel export.
* YouTube scraping can now pull an entire channel within a timeframe, not just individual videos.

**Improved**

* Faster Survey Result and source page loads.
* More accurate credit charging and renewals across the board.

**Fixed**

* Clearer per-source credit warnings when adding sources.
  {% endupdate %}

{% update date="2026-06-26" %}

##

**New**

* Custom Monitoring: apply a shared library of detection rules across every source in a feedback group, with its own alerts and trend line over time.

**Fixed**

* PDF reports are now available for scraper and Evolution-tracked surveys.
* Various Evolution, segmentation, and reporting fixes.
  {% endupdate %}

{% update date="2026-06-23" %}

##

**New**

* Run in-depth AI analysis for a single segment on demand, and switch between Evolution periods instantly.
* YouTube scraping adds AI-assisted autofill, relevance scoring, and exact comment dates.

**Fixed**

* Retired the recurring-survey rotation feature (scheduled weekly/monthly auto-cloning); use Evolution to track a survey over time instead.
* Various survey logo, upload, and segment-view fixes.
  {% endupdate %}

{% update date="2026-06-19" %}

##

**New**

* Evolution groups can lock all AI analysis, themes, summaries, and time labels, to a single language across every period.
* Connecting a data source to a tracked feedback group now defaults to continuous syncing, with one-time import as a secondary option.
* Added Avis Vérifiés as a new review-scraping platform.
* "Recompute grouped themes" is now available on every feedback group, not just tracked ones.

**Fixed**

* More accurate quote anchoring in AI-written report summaries.
* Various Evolution and segmentation display fixes.
  {% endupdate %}

{% update date="2026-06-12" %}

##

**New**

* Redesigned survey experience: a familiar Google Forms–style flow (section-per-page for respondents, single-page builder).
* Combined-corpus analysis on the Overall view: analyze all of a group's sources together.
* Pinned feedback-series navigation bar with a period picker and drop-down trend detail.
* Summary-based analysis for long comments, on by default.
* Follow-up questions with WYSIWYG customization and improved display logic.

**Improved**

* Faster loading across analysis views, with results that stay cached reliably until you open them.

**Fixed**

* Run timelines now read left-to-right (oldest → newest) on every surface.
* Numerous Manage Sources refinements, including reattaching detached web-scrape sources.
* Rebranded to "BAI Analytics" across emails, SMS, and PDF reports.
  {% endupdate %}

{% update date="2026-06-08" %}

##

**New**

* Call-transcript ingestion with dedicated Sales-cycle and Callback-priority views.
* Per-comment priority grading.
* Named, organization-shared monitor collections on an analysis.
* Large multi-file uploads via resumable upload.

**Improved**

* Substantial speed-ups in comment loading and analysis throughput.

**Fixed**

* More reliable handling of very large analyses.
  {% endupdate %}

{% update date="2026-06-07" %}

##

**New**

* In-platform email template editor with personalization, conditionals, and ZIP/CSV export.
* "Recompute grouped themes" per period bucket (reset and relaunch from scratch).
* Period-scoped reports and multi-source bucket aggregation.
* Web-scraping raw data now renders as a social feed.
* Specific date-range picker and clearer prompt guidance.
* Segment by NPS category and rating band, persisted across pages.
* Raised the per-platform URL limit from 20 to 500.

**Improved**

* About 3 seconds faster on first open.

**Fixed**

* Clearer error and progress messages, with a no-data vs. failure distinction; "no data in time range" now shows as info rather than an error.
  {% endupdate %}

{% update date="2026-05-20" %}

##

**New**

* Optional relevancy filter and a Raw Data date filter.
* Per-survey font selection.
* Connector sync progress (last/next fetch) in the sources drawer.
* "Remove from group" on every source row.

**Fixed**

* More stable Grouped Themes across re-analyses and recomputes.
  {% endupdate %}

{% update date="2026-05-18" %}

##

**New**

* Time-bucketed analysis for feedback groups: track how feedback evolves over time.
* Continuous scraping with incremental fetch and auto top-up.
* Identified-respondent lists.
* Per-option custom colors for multiple-choice charts.
* Editable comment sentiment with async narrative refresh.

**Improved**

* Faster bucket and source loading.
  {% endupdate %}

{% update date="2026-05-15" %}

##

**New**

* Question-aware sentiment scoring.
* Inline citations with a verbatim popover and deep-link.
* Mixed-sentiment detection.
* Segment surveys by QR code.
* Bulk theme/monitor download and survey PDF preview.
* Summaries, in-depth analysis, and Q\&A now respect your organization's language.

**Fixed**

* Faster, more honest PDF report progress; multilingual exports in the survey's own language.
  {% endupdate %}

{% update date="2026-05-04" %}

##

**New**

* Richer Quick Overview with AI in-depth analysis, sub-categories, Q\&A, and a PDF report.
* Option to include recommended actions in PDF and Word reports.
* Accurate "Continue where you left off."
* Summaries grounded strictly in your feedback: no fabrication.

**Fixed**

* Auto-fit charts and wrap long axis labels.
  {% endupdate %}

{% update date="2026-04-28" %}

##

**New**

* Robust large-file and multi-sheet upload pipeline.
* JSON interview-transcript ingestion: drag and drop alongside .docx/.pdf.
* Usage metering switched from row counts to AI tokens (APS).

**Improved**

* Moved AI tasks to a faster model for quicker responses.
  {% endupdate %}

{% update date="2026-04-22" %}

##

**New**

* Monitors rendered as theme-style cards in the survey PDF, with sub-themes and sub-categories.

**Fixed**

* Trustpilot URL validation accepts country subdomains.
* Scraper wizard now surfaces save-blocking validation errors.
  {% endupdate %}

{% update date="2026-04-20" %}

##

**New**

* Higher monitor limits (up to 150) with AI de-duplication.
* Manual custom-monitor creation mode alongside AI mode.
* Added Reddit as a web-scraping platform, with relevance filtering.
* Enabled Facebook scraping.

**Fixed**

* Complete EN/FR/ES translation coverage across the platform.
  {% endupdate %}

{% update date="2026-04-12" %}

##

**New**

* Survey information blocks, acknowledgment gates, and conditional reveal.

**Fixed**

* Support for larger Excel uploads (raised column limits).
  {% endupdate %}

{% update date="2026-04-08" %}

##

**New**

* Product analytics wiring, so we can see which features get used.

**Fixed**

* Clearer scraper progress messaging; content-shaped skeleton loaders replace spinners.
  {% endupdate %}

{% update date="2026-04-07" %}

##

**New**

* AI usage and cost tracking for every analysis.
* Onboarding questionnaire with a personalized checklist.
* Super-admin dashboard with three-tier permissions and analytics.
* Bot protection on the embed widget.
* Monitor notifications with email/SMS alerts.
* Per-comment translation.

**Fixed**

* Back-button overhaul: no more dead-ends or loops.
  {% endupdate %}

{% update date="2026-03-30" %}

##

**New**

* Background dataset uploads with a progress toast.
* Facebook and Instagram scraper platforms.
* LinkedIn as a web-scraping platform.
* Dedicated scraper results dashboard.

**Improved**

* 2–3× faster analysis for large surveys.
  {% endupdate %}

{% update date="2026-03-12" %}

##

**New**

* Subscription billing.
* Google News and TripAdvisor scraping platforms.
* Redesigned Custom Monitoring UI with inline creation.
* Enhanced PDF reports with AI insights and a priority matrix.
* Sign-up page.

**Fixed**

* Improved mobile sidebar and menu.
  {% endupdate %}
  {% endupdates %}