Better, every release.

The tracking pipeline now uses newer Grok and Meta models. Existing brands move through the update as fresh answers arrive. Historical answers keep their original source, and the interface uses assistant names consistently so reports remain easier to read as model versions change.

The report pipeline now handles Chinese and Japanese text more reliably when finding brand mentions and matching competitors. Prompt suggestions for non-English brands have also been improved to read naturally in the chosen language. Each question keeps its own market, so a wider report scope does not move it into another country.

WordPress publishing can use the site's normal connection without requiring the Trakkr plugin. WordPress SEO support now includes Slim SEO alongside supported existing plugins, with each connection showing the changes it can apply. Webflow and Shopify publishing use readable URLs based on your title. If a publishing response is interrupted, Trakkr checks for the existing post before creating another copy.

Send a language with POST /get-brands/markets to track a country in the language your buyers use. Leaving it out keeps the brand language as the default. Existing market languages and historical reports stay in place, and missing question translations are prepared for the next scan. Requires REST API access, brand editor access and an available market slot. REST API access is normally included with Scale and eligible Enterprise accounts.

Citations distinguishes a page that mentions you, a checked page that does not, and a page that could not be checked. Unreadable pages no longer count as confirmed gaps. When a page is blocked, an available excerpt from that same AI answer can support a mention. Open the evidence or request a recheck from the page menu. Reddit threads and cited YouTube videos are also read through more suitable sources.

Explore adds a Brands view for accounts with more than one readable brand. Compare visibility and other measures across a shared window, split the table by another dimension, and open the evidence behind a figure. Save a question or copy its view link to return to the same setup. CSV and copied tables follow the rows, filters and calculations on screen.

The Reports export menu can build a single Excel workbook for your portfolio. It includes monthly results and a separate scan-history sheet for each brand, with scan times in your time zone. Monthly figures come from the same saved records as the Months tab. The export shows its progress and names the brands included, making client reporting easier to prepare in one download.

Choose a content opportunity and turn it into an article without starting a paid plan. The allowance is shared across the account, and Create shows when another article becomes available. You can read and edit the generated draft; publishing to a connected site still depends on your plan and connection.

Documents keeps the Agent's reports beyond the conversation. Choose a completed month, audience and sections, add your business context, then review a draft with linked figures. Edit or rewrite one section, resume interrupted generation, and publish a saved version when it is ready. Download PDF, Word or Markdown, or save the structure as a template for another report. The library supports search, accessible brands, drafts, published versions and archiving.

Reports now has a Months tab alongside the scan history. Open a month to see its visibility, presence, rankings and supporting sections, compared with the month before. Saved coverage shows what contributed to the result. Corrections are recorded as restatements, so a changed historical figure has an explanation. Export the closed months as CSV or use them as the source for a document.

Sheets sync calculates scores from the same answered prompts as CSV exports. Failed provider requests no longer pull visibility down, and a day with no usable answers leaves an empty cell. Brand names and competitor groups are combined consistently. Tag Trends now shows your own brand for each tag.

• Presence still includes failed attempts in its base, so this change affects visibility rather than hiding missing answers
• Existing sheet columns stay in place

Set your branding, connect a custom domain, send branded email and share client reports as part of Scale. White-label access no longer needs a separate per-brand add-on. Client invitations do not use team seats or increase the bill.

• Includes client portals, custom domains, branded email and report sharing
• Agency trials can configure and preview a portal; publishing a live portal requires paid Scale

Reports brings the scan history, score trends and changes into one workspace. Compare a previous period or a custom window, inspect movement by prompt, model or tag, and add dated notes beside the chart. A like-for-like comparison helps separate movement from changes to your tracked set. Save a personal or shared view, then create a report link with the sections, password and expiry you choose.

The report market picker includes stopped markets, so ending tracking does not hide their saved results or exports. When reducing paid market slots, choose the markets to keep for each affected brand and review the billing change before confirming. The primary market stays selected, and saved reports remain available.

Diagnose now explains one buying question through a linked evidence trail. See which pages shaped the answer, who appeared on each assistant, and whether your brand was named with search or from model knowledge. The brief shows your tracked visibility beside the fresh samples and calls out disagreements. Pick fixes for Actions, track the question, export the evidence or scan it again later to see what changed.

The question matrix makes it easier to find a wrong price, stale fact or unsupported answer. Open a question to read what each assistant said, compare it with retrieved evidence, and confirm or correct an unsettled fact. Recheck a single question without scanning the whole brand again. Corrections and check history show what improved, what is new and what went wrong again. Missing answers remain visible as gaps in coverage.

The Actions queue groups work by what it needs next: your decision, this week's plan, suggestions, measurement, or a later date. Each item opens its evidence, draft and next step in one brief. Starting work saves what will be measured, for how long and what counts as an improvement. Results records earned outcomes, no movement, drops and cases that could not be measured, without claiming the work caused a rise or fall.

Clients are grouped by the reason to look at them, such as losing ground, waiting for a decision or having a report due. The roster also surfaces steady accounts where little work has shipped and the client has not been looking. Select two to four clients for a side-by-side comparison, adjust the reporting window, or export the rows you need. Client workspace, invite and report controls stay close to each row.

Agency pitches keep a saved brief, editable question set and measurement history. Write your own questions or edit suggestions, choose the competitors and test the prospect in a specific market. Read captured answers and sources before using the report. A new scan keeps the shared version in place until you publish the update. Pitches remain one-off tests and do not use a tracked brand slot.

Prompts has focused views for the questions gaining or losing ground and the ones where you barely appear. Search terms such as declining, inactive or score above 50 filter the table directly. Group rows by tag, audience or intent, and choose the columns you want. Adding prompts uses a staging table for pasted lists and CSV files, with duplicates marked before you save.

The public Prompt Tracker Fingerprints study examines recurring questions, language patterns and timing in AI search traffic. It shows the evidence used to separate likely tracking activity from buyer conversations, with the limits of that distinction made clear. Use it alongside the free AI Query Scanner when interpreting unusual queries in your own Search Console data.

The Trakkr integration connects through a sign-in and brand picker. Send priority work to your task system, add visibility readings to a sheet, or notify your team when a report is ready. Zaps can also add tracked prompts and complete Actions when the connected user has editor access. A catch hook in Integrations remains available for simple one-way event delivery.

Automatic exports now check the saved schedule before starting another sync after a report. Pending syncs are not started again, and temporary failures wait before another attempt. Choose manual, daily or weekly in the Google Sheets integration. Connecting and syncing requires a paid plan; changing the connection requires brand editor access.

Connect Google Search Console with read-only access and choose a property. The scanner reviews up to 16 months of queries, separates likely machine patterns from buyer questions, and shows the evidence for its findings. Results appear as the scan runs, and the completed report is saved for the Google account that requested it. You can also receive it by email.

The AI Crawler Log Analyser reads server or CDN logs and groups recognised user agents, paths and response codes. The AI Traffic Analyser reads a GA4 export to compare identifiable AI referrals, landing pages, engagement and conversions. The Share of Voice calculator works from answers you have already collected and shows its calculation. Each tool runs locally and lets you download the results.

• Crawler user-agent matches are signals, not proof of a verified bot
• Traffic analysis reports identifiable referrals and does not relabel Direct traffic as AI

Security now brings together your sign-in methods, password controls, two-factor authentication and trusted devices. Sign out other sessions or forget trusted devices when you need to. Team owners can require two-factor authentication for the whole team, and the activity list gives you a record of recent security changes.

Opening a brand in Settings takes you to its own page, with more room for identity, market and tracking controls. Existing edit links still open the right brand. The Ask search bar can also take you straight to settings such as billing, language or two-factor authentication when you ask how to change them.

The plan view shows where you are in the billing period and what comes next. Paused plans show when billing resumes, with a reminder before the pause ends. Receipts and billing emails explain the payment or change more clearly, so you can check your plan without piecing together separate screens.

Actions brings a weekly brief, suggested work and follow-up results together. Each item explains what was found, why it matters and what to do next, with links to its evidence. Completing work starts a check of the relevant page or measure. Results show improvements, no movement or missing evidence.

• New suggestions include broken cited pages, review sites shaping answers and competitor-only forum threads
• Supported technical fixes can be checked overnight and marked as verified
• Duplicate or outdated suggestions are reviewed as new evidence arrives

Accuracy brings the report and claim record onto one page. Open a claim to read the assistant transcript behind it, check the evidence and decide what needs attention. Recheck the report, download claims as CSV or print a record for someone outside Trakkr.

Site Optimization opens with a plain explanation of the findings, followed by ranked work and its evidence. The full checklist sits inside Findings, so the same problem is easier to follow from summary to detail. Each item includes a suggested fix and a description you can copy into a ticket.

• Page counts say how much of the site was checked
• Unmeasured items stay distinct from zeroes
• First scans, progress and stopped scans use the same report structure

Assign topics to tracked prompts to compare your position across categories. Trakkr suggests a starting set from your existing questions, which you can review and edit. Topics selected during brand setup carry into your prompt list.

Integrations groups connections by category and puts connected tools first. Each connection shows its status, recent sync and available next steps. WordPress connections list what Trakkr can do on that site, and developer credentials have one home in Settings.

CSV exports offer a summary time series, detailed prompt results or the ranked brands from each answer. Filters and a preview help you check the scope before building a file. Completed exports stay in a history list so you can download them again.

AI Pages tests the site as an AI crawler would and reports the result. The setup guide covers common problems, including domains that are not proxied through Cloudflare and therefore do not run the worker.

Explore shows how each model contributes to a prompt score and how the total changed. Source profiles in Citations make it easier to continue from a score to the pages used in the answers.

Visit trakkr.ai/open-source to see which AI crawlers fetch our site, which pages they visit and where those pages appear in citations. It is a read-only example of the workspace you can use for your own site.

Perception replaces broad tone labels with attributes supported by assistant quotes. Open the evidence to see how the description was formed. The Agent and Actions use the same descriptions when suggesting work.

Each prompt has a full page for its score, trend, model breakdown, cited sources and measured phrasings. Search Console data appears when connected. Use j and k to move through prompts in the order of your filtered table.

The public Answers library explains citations, model recommendations and ways to measure visibility. Browse the questions at trakkr.ai/answers and follow the supporting material when you want more detail.

The free scanner at trakkr.ai/data/security checks common paths used to look for exposed environment files and configuration data. Enter your domain to see which tested paths can be read without signing in.

Choose the client, recipients and schedule, and Trakkr prepares the report in your agency branding. You can require approval before each send so someone on your team reviews it before delivery.

The agency home brings client visibility and portfolio movement together. Open a client workspace, compare performance or focus on accounts needing attention without switching between separate lists.

Connect a Search Console property to compare search performance with AI citations and recommendations. The same data helps Actions identify pages worth improving, including pages that receive Google traffic but rarely earn AI citations.

The free rank tracker suggests buying questions from the site, samples AI answers and keeps the raw answer beside each result. Use it for your own domain or a competitor's as a first look before setting up ongoing tracking.

Link an OpenAI ad account to a brand and check the connection. A free context-hint tool helps prepare hints for buying questions, and an open-source MCP server provides access to the OpenAI Advertiser API from compatible assistants.

Perception no longer treats a qualified answer as a failure to recognise the brand when the assistant describes the business correctly. This separates disagreement with a stale profile from a lack of awareness.

The public Bias Index asks models the same questions repeatedly without web search. Browse by model or question, compare two models, and inspect the published method and the limits of what was measured.

The Crawlers Map tab turns your site into a path tree. Compare access, traffic, citations and freshness, or follow crawler journeys and return patterns. Use the worklists to find pages crawlers visit without citing and pages they do not reach.

The Shortlist publishes sampled AI recommendations across categories, with the answers and the products named in them. Use it to inspect how a category is described and who appears in the sampled results.

Accuracy asks assistants about your business and checks factual claims against available evidence. Keep the answer, source and check history together, then turn a finding into an action or a printable record.

• Check identity, pricing, products and other facts buyers ask about
• Read the evidence and history behind a finding
• Take unresolved findings to the Agent or Actions

Agency pitches use measured visibility signals to show a prospect where they appear and where they are missing. Review the report, apply agency branding, and share it through a link or PDF with the access settings you choose.

Guided setup explains how to connect an assistant to your Trakkr account and inspect the brands you can access. Use the hosted sign-in flow or a personal connection token where the client requires one. Manage connected assistants from Settings.

• Setup guides for supported assistants and developer tools
• Personal credentials and revocable connections
• Brand access follows your Trakkr permissions

Weekly emails add context to the numbers, including competitors gaining ground and questions where your brand is missing. The report links back to Trakkr so you can inspect the data and follow up.

Prompt health reviews the questions you track and suggests what to add or prune. Movement and overtake indicators help you choose which prompts to investigate first.

Citations combines source profiles, query context, video citations and outreach opportunities. Filter the evidence, inspect your own channels and export a useful cut for the team.

• Read sources and the queries that surface them
• Follow up through outreach or content work
• Filter and export citation evidence

The Agent workspace keeps threads, brand context and tool activity together. Ask about visibility, competitors, citations or site findings, then inspect the lookups behind the answer. Proposed changes appear as cards you can review and confirm.

• Keep investigations in saved threads
• Read tool activity and attached evidence
• Confirm suggested changes in the conversation

Brand setup shows site findings, crawler readiness and assistant progress as results arrive. If a scan or checkout is interrupted, the recovery flow helps you continue from the saved state.

Crawler analysis feeds suggestions into Actions, linking a recommended change to the page and traffic behind it. Review access problems, pages missing useful content and places where better coverage could help.

• Inspect the page content received by crawlers
• Review recommendations with their evidence
• Follow site work through Actions

The public status page gives you a place to check whether a Trakkr service is delayed or unavailable and read any posted incident or maintenance updates.

Crawler reporting adds page drawers, bot details, session filters and date ranges. Explore the visits behind a trend, check connection status and download rows as CSV.

Create can check selected signals on a schedule and turn suitable opportunities into drafts. Set the pace, sources and quality checks, then review the work. Automatic publishing is available only where your saved settings and connection allow it.

• Choose sources, schedule and quality checks
• Review drafts in the queue
• Set publishing permissions before allowing automatic delivery

Site Optimization prepares a before-and-after change from an audit finding. Approve it to apply through a supported CMS or code connection, then review the verification result. Supported changes can also be rolled back.

• Read the proposed edit before applying it
• Use connected CMS or GitHub destinations where supported
• Follow applied, verified, declined and rolled-back changes

The audit workspace brings findings, page detail and history together. Expand a finding to inspect its evidence, compare scans and choose the work most relevant to your site.

Site connections support publishing and update workflows. GitHub connections can map a page to its likely source file, explain the choice and open a pull request. Review or override the mapping when the evidence is uncertain.

• Configure a connected publishing destination
• Review GitHub file mapping and proposed changes
• Use manual review when the target is uncertain

Create brings content opportunities, campaigns and article drafts into one workspace. Start from citation gaps, tracked questions or connected traffic data. Set brand voice, knowledge and templates, then follow citations to the articles you publish.

• Review opportunities with their supporting evidence
• Group related work into campaigns
• Generate and edit drafts
• Track citations to published articles

Choose destinations and routing for the tools your team uses, such as Linear, GitHub, Slack, Trello, Notion, Asana or Jira. Available actions depend on the connection, and linked work keeps its status and context together.

Submit eligible URLs through IndexNow after an update and check the submission status. A successful submission confirms receipt; it does not guarantee that an AI assistant will cite the page.

Dashboard, Competitors, Diagnose and Prompts use the selected market. Citations remains brand-wide. Brands tracking multiple locations can keep regional questions and their results together while moving between views.

The Agent can keep a plan for tasks that need several steps. Progress, approval requests and recovery controls stay attached to the work so you can continue an investigation or change without starting over.

Crawler reporting recognises Google-Agent as an AI agent rather than grouping it with search or training crawlers. For installations using an edge template, regenerate the setup snippet to include the updated detection.

• Separate Google-Agent rows in crawler reporting
• Updated Cloudflare and Netlify setup templates

Prism has been renamed AI Pages. It serves an alternative version of a page to supported AI crawlers, with structured information and useful text drawn from the page. Manage it under AI Pages in Trakkr.

Perception adds a visual comparison across trust, expertise, value, innovation and market presence. Choose a competitor to inspect where the model assessments differ.

The source-type filter shows counts for each category and narrows the citation list. Use it to understand the mix of sources appearing in your answers and focus on the ones you want to investigate.

The REST API adds these product areas with request and response documentation. Use your account's feature access and brand permissions to retrieve data or run supported operations.

• Perception and narrative data
• Diagnoses and content ideas
• Reports and crawler activity
• Scores by prompt and market, plus brand market and name settings

Add markets to a brand and read their results separately. Supported assistants use the selected location for search, and reports keep each market's prompts and answers together.

• Choose a primary market and add others
• Set the market for a prompt
• Compare results within the same regional scope

The content workspace combines the material and preferences used when writing for your brand. Store knowledge, choose the voice and build reusable templates for common article types.

• Brand knowledge and messaging
• Reusable article templates
• Writing style and audience settings

Perception brings saved assistant assessments into one dashboard. Inspect how your brand is described, follow changes over time and compare the results with competitors.

Use the builder to choose a trigger, conditions and delivery through Slack, email or a webhook. Templates provide a starting point for common monitoring tasks.

• Templates for common rules
• Conditions and filters
• Slack, email and webhook delivery

The Agent reads your brand data to answer questions, investigate movement and propose next steps. Keep the conversation in a thread and review a suggestion before confirming a change.

Citations combines the sites and pages used in tracked answers with query context. Inspect a source, compare its coverage and find opportunities where competitors appear but your brand does not.

The combined audit checks page access, structure and content, keeps scan history and suggests changes to review. Use the findings to choose which pages need work first.

Competitor reporting adds configurable breakdowns by prompt and model, with position detail for each brand. Match the scope when comparing your results.

• Choose the breakdown you need
• Read prompt and model results
• Compare competitor positions

Traffic reporting brings visitor analytics and crawler activity into Trakkr. Follow identifiable AI referrals, then inspect which crawlers visited and which pages they requested.

Team controls support brand permissions, assigned client access and brand groups. Activity records help you follow changes made across the workspace.

• Assign client access to specific brands
• Organise related brands into groups
• Review team activity

Demand scoring compares the estimated interest in prompt topics. Use it alongside relevance to your buyers when choosing a tracked set. It is an estimate rather than a count of private conversations.

Prompt Ideas suggests questions to review before adding them to your tracked set. Use the suggestions as a starting point and edit the wording to match your buyers.

Choose a report format and sections, review the figures and share the result with stakeholders. Shared dashboard views provide another way to let someone inspect the data.

Trakkr's learning resources explain how to measure visibility, interpret reports and choose useful work. Follow the guides and examples alongside your own results.

• Product guides and tutorials
• Strategy playbooks
• Customer examples

A WordPress connection lets Trakkr prepare supported changes against your site. Inspect the proposed edit and confirm it before applying.

Activity lists events in time order and lets you filter the record. Use it to check what finished and who made a change.

Outreach matches opportunities to relevant publications and helps prepare a draft with the source context attached. Review the pitch before sending it.

The V2 design updates page layouts, typography and dashboard components, with improvements for smaller screens. Related data and controls are easier to find together.

The V2 update improves page loading and data fetching, reducing repeated work when moving between views.

The V2 API updates access to brand data and supported product operations, with clearer endpoint documentation. Use the reference to check credentials, parameters and limits for each request.