What’s new

Every entry below renders from _data/changelog.yml — the same data that drives the “Recent changes” block on each feature page it names.

31.07.2026

  • Limadata joins the bring-your-own people-data providers — title, country, city and skills all filtered at source, LinkedIn profile and headline returned, and mobile phone numbers alongside LeadMagic. One API key. — see Bring your own people-data provider

29.07.2026

  • HarvestAPI joins the bring-your-own people-data providers — LinkedIn lead search with title and country filtered at source. It has no email product, so it returns people and job titles only, and a project with “Email is required” on will not spend against it. — see Bring your own people-data provider

  • Wiza joins the bring-your-own people-data providers — title, country, city and skills all filtered at source, plus mobile phone numbers. Email and phone are reveals that bill only on a hit, spent against one-off API credits rather than a seat plan. One API key. — see Bring your own people-data provider

09.07.2026

  • Agent data plane Clay integration published: a company’s hiring motion as flat, filterable columns from a single enrichment call. — see Agent data plane — Clay integration

  • Agent data plane MCP server published: the same primitives as agent tools – code-complete, not yet hosted at a public endpoint. — see Agent data plane — MCP server

  • Agent data plane REST API published: the full /v1 reference for every endpoint, request, and response shape. — see Agent data plane — REST API

  • The agent data plane launches: a machine-facing surface reporting what changed in a company’s hiring activity, not just what’s posted now – a REST API, an MCP server, and a Clay integration. — see Agent data plane

30.06.2026

  • API access launches: a read-only REST API for pulling your own projects and leads out of SignalsAPI as JSON or CSV. — see API access

18.06.2026

  • Job description required words – the “Job description search terms” field is now a hard requirement and has been renamed accordingly. A signal is kept only if its description contains at least one of the words you list (OR within the list), on top of any job-title requirement. This is the description counterpart of “Job title required words”. Use case: keep only remote-friendly roles by requiring remote, home, or hybrid in the description. Previously these words only widened the search, so jobs that didn’t actually mention them could still slip through. Read more: Advanced search — see Advanced search

16.04.2026

  • Accurate signal status badges – the Signals page now correctly distinguishes between companies where decision-makers were actually found vs. companies that only passed qualification filters. Previously, all qualified companies showed “Leads found” even when no decision-makers existed. Now: — see Find decision-makers

  • Company name in CSV exports – the company name column was missing from CSV downloads; now included. — see API access

  • “Available on paid plans” on sample pages – sample lead pages and CSV downloads now show “available on paid plans” instead of blank fields for decision-maker data (name, title, LinkedIn, email). This encourages trial users to upgrade for full lead details. — see Find decision-makers

  • Company website always in exports – CSV exports now always include the company website and domain, even when no decision-makers were found. Companies from sources that don’t provide website data are automatically enriched in the background via LinkedIn lookup. While enrichment is in progress, the signal detail page shows an “Enriching…” badge on the Website and LinkedIn URL fields. Use case: you see an interesting signal and want to dig into the company before finding the right people. — see Find decision-makers

14.04.2026

  • Recruit CRM integration — push qualified leads into your Recruit CRM as contacts (sales-side prospects, not candidates), linked to the hiring company. Duplicate contacts are detected by email, so re-exports do not create duplicates. Add it in Settings → Integrations → Recruit CRM (requires a Business-tier Recruit CRM plan or higher, as the Recruit CRM Open API is not available on lower plans). — see Integrations

17.03.2026

  • Automatic retry for failed personations – companies that initially returned “no leads found” (e.g., because their LinkedIn profile wasn’t resolved yet) are now automatically re-checked. When enrichment data improves, tasks are reset and decision-maker search runs again — no manual intervention needed. — see Find decision-makers

  • Company stop words are now applied during search (previously only applied after fetching, wasting credits). — see Filter leads with AI

  • Job posting age now works correctly – the “Maximum age of job posting” setting is now properly applied. — see Advanced search

  • Hiring surge – only show companies with 5+ new roles in 30 days at 1.5x rate. — see Advanced search

  • Hard to fill – only show roles open 30+ days or reposted 3+ times. — see Advanced search

  • Exclude staffing agencies – built-in filter that’s more reliable than manual stop words. — see Filter leads with AI

  • Location patterns – match specific cities or regions (e.g., “London”, “Manchester”) instead of filtering by country only. — see Advanced search

  • Job families – filter by structured job classification regardless of title wording. — see Advanced search

  • Skills filter – require or exclude specific skills mentioned in the job posting (e.g., PMP, PRINCE2, Agile). — see Advanced search

  • Description search – search within job descriptions separately from titles. Catches signals where relevant terms appear in the description but not the title. — see Advanced search

  • Semantic search – AI-powered search that finds similar job titles beyond exact keyword matching. “Project manager” now also finds “delivery manager”, “program director”, “PMO lead” and other semantically related roles. — see Advanced search

18.01.2026

  • Telephone numbers for qualified decision-makers are now provided as an add-on — see Find decision-makers

  • We now support more jobs including TechInAsia, Reed, Justjoin, Jooble, Finn.no, Jobindex, Monster, RocketJobs + more — see How it works

11.04.2025

  • Unspent credits expire in 60 days — see Pricing

09.04.2025

05.04.2025

04.04.2025

  • Option “Auto-reject people without emails” was removed from Lead lists, option “Emails required” was added to personas instead. If turned on, all people will have emails already on the “Personas” stage. “Unlock email” link removed from lead lists, as all emails are already unlocked. — see Anatomy of a lead

06.03.2025

  • Option to not filter out companies with unknown number of employees — see Advanced search

22.01.2025

17.01.2025

14.01.2025

  • Bugfix: Low search result count — see Create a search

  • Bugfix: Bullhorn integration silently continued sending leads to nowhere after integration has expired — see Integrations

  • Fewer fields in CSV export — see API access

13.01.2025

  • Bugfix: Trial credits were added twice — see Pricing

09.12.2024

  • Indeed searches are temporarily limited to US and UK — see How it works

03.12.2024

  • Bugfix: not exporting leads that are in process of unlocking email — see API access

01.12.2024

  • Bugfix: AI prompts with job_location field did not work — see Filter leads with AI

  • Bullhorn export field mapping must include all fields to be exported, nothing is exported by default anymore — see Integrations

30.11.2024

  • Bullhorn export has a field mapping setting — see Integrations

  • Exporting date fields to Bullhorn — see Integrations

29.11.2024

  • Field mapping for Bullhorn integration — see Integrations

  • Showing integration type in export list — see API access

28.11.2024

27.11.2024

26.11.2024

  • Bugfix: personator produced no results if both Person location and Max decision-makers per company are specified — see Find decision-makers

  • Default view mode of filter matches is “All” instead of “Qualified” — see Advanced search

  • Showing pagination instead of a date filter in matches lists in Filters, Personas, and Lead lists — see Find decision-makers

25.11.2024

  • Search can be limited by a US state or UK country — see Advanced search

  • Jobs with match in title show first in run view — see Create a search

  • Jobs with match in title have higher fetching priority than without — see Create a search

  • Manual runs have higher priority than automated — see Create a search

24.11.2024

  • New searching core working faster, but not allowing more that 4 words as search term, quotes and special characters, words “or” and “and” — see Create a search

16.11.2024

  • Normalizing job title on save – trimmed, lowercase, no special chars, quotes, “and”, “or” etc — see Advanced search

13.11.2024

  • Improvement: showing some warning messages that search will take long to run after a search is created — see Create a search

08.11.2024

  • Bugfix: disabled search schedules were shown as if they are enabled — see Create a search

  • Improvement: Scheduling the next run after a search with a running schedule has been created or edited and if you press a “Save” button (not “Save and run”) is behaving more predictably — see Create a search

02.11.2024

30.10.2024

  • Bugfix: Daily searches sometimes did not plan the next run. — see Pricing

29.10.2024

  • Bugfix: The new search engine introduced September 29th was not properly fed the “remote” parameter so it was ignored in LinkedIn searches. — see Work arrangement filter

26.10.2024

  • When search is scheduled for the first time, it will not require a manual “Run now” anymore but will start running after a defined interval — see Create a search

23.10.2024

  • Automated daily testing of new user registration, verification, welcome email — see Operator quickstart

22.10.2024

  • New engine for docs.signalsapi.com — see Pipeline

20.10.2024

  • Added more location clusters to LinkedIn searches, like North America, Latin America etc. — see How it works

  • Added more country options into both LinkedIn and Indeed searches. — see Advanced search

19.10.2024

  • Users on expired and cancelled plans are still able to go to billing and download invoices — see Pricing

18.10.2024

  • Fixed bug – manually initiated search never stopped but created a new run right after previous one has ended — see Create a search

29.09.2024

26.09.2024

  • Searching and fetching engine were completely re-written. The new one works slower, but much more stable. — see Pipeline

08.07.2024

  • Scheduled searches that were paused because of low balance continue automatically after balance top-up — see Pricing

20.06.2024

  • Fixed webhook exporting — see Integrations

  • Export showing integration name and exports per day stats. — see API access

19.06.2024

18.06.2024

  • Search has a “Job posting age” setting, which controls how recent the job posting should be to get matched. — see Advanced search

17.06.2024

  • Filtering jobs based on where the match was found. Solving two most popular problems – “too few matches” and “too many junk matches”. — see Advanced search

16.06.2024

  • Highlighting relevant keywords in the job title and description when viewing the match — see Advanced search

  • Showing in the run which part of the signal matches search terms. Some matches do not contain search terms at all, but the search engine (Indeed or LinkedIn) still shows them. — see How it works

14.06.2024

07.06.2024

30.05.2024

29.05.2024

20.05.2024

  • Do-Not-Contact list allows to make sure you do not contact companies or people you do not want to. Set it up in Leads -> DNC. — see Do-Not-Contact list

  • Limit the number of decision-makers per company. If more decision-makers are identified, they are prioritized in the order specified in the ‘titles’ field, and those with valid email addresses come first. — see Find decision-makers

18.05.2024

  • Fixed: Opening a filter and a persona took significant time. — see Find decision-makers

  • Fixed: Resetting a failed search did not reset match counts in that search so search match count stopped matching sum of success counts of its runs. — see Create a search

  • Fixed: Searches get stuck forever in “running” status if run has failed. — see Create a search

  • List views of searches, filters, personas, lead lists and exports show where they are taking data from and where they are sending it to. — see API access

14.05.2024

  • Leadlist has an AI field which applied to each person, available templates are {first_name}, {last_name}, {title}, {email}, {linkedin_url} — see Personalize emails with AI

13.05.2024

  • Leads are downloaded based on the applied filters — see API access

  • Leads within a leadlist, matches within personas and filters can be filtered by date — see Anatomy of a lead

12.05.2024

  • Accepting registrations with personal emails, but asking for company info and limiting free runs in this case — see Find decision-makers

  • Anti-bot protection — see Pipeline

11.05.2024

10.05.2024

09.05.2024

  • Fixed: runs in “Fetching” status were showing as “Completed” — see Create a search

  • Regions of England in search location — see Advanced search

  • New run status: Fetching — see Create a search

  • LinkedIn match limits per search run increased. — see How it works

  • By default, deduplication of matches occurs within each run, not across the entire search. This avoids confusion regarding the actual match count for a given run. Deduplication on the search level and globally is still available as a lead list setting. — see Anatomy of a lead

08.05.2024

  • Fixed: deleting filters — see Advanced search

  • Fixed: deleting filter fails with a missing variable — see Advanced search

  • If leadlist is reset while running, the run is terminated — see Anatomy of a lead

  • Preventing double submissions in forms by disabling buttons on first click — see How it works

07.05.2024

06.05.2024

  • Lead view has clickable hyperlinks for job posting, company website and LinkedIn, person LinkedIn — see How it works

  • Company description is shown with line breaks instead of smashing all lines together — see Anatomy of a lead

  • Showing job description — see Anatomy of a lead

05.05.2024

  • Speed-up under the hood: searching and fetching+enriching happens in separate processes, allowing new matches to appear and get processed before the whole search run is completed — see Pipeline

04.05.2024

03.05.2024

02.05.2024

01.05.2024

30.04.2024

  • Fixed: searches are getting stuck in an infinite loop — see Create a search

  • Fixed: it is now possible to approve a lead before emails finished unlocking, export will happen automatically when unlocking is finished — see API access

  • Fixed: it is now possible to reject a lead that does not have unlocked emails — see Bring your own people-data provider

  • When you click “approve” or “reject” it shows the next lead, saving a few clicks — see Anatomy of a lead

  • Showing running status in search list — see Create a search

  • Moved to a more powerful server, everything works much faster now — see Pipeline

29.04.2024

  • Leads will get exported after you approve them by clicking “Approve” or “Reject” buttons, or enabling auto-approval in lead list settings. — see API access

  • New leads are marked as “New”, and then change to “Seen” when you open them for the first time. — see Anatomy of a lead

28.04.2024

  • Lead settings can be applied to new leads only, or the whole lead list can be emptied and re-run with the new settings. — see Anatomy of a lead

  • Automatic approval and rejection rules can be set in lead list settings. — see Anatomy of a lead

  • New tab “Leads” for working with leads in a mailbox-style with folders, downloading, unlocking emails, etc — see API access

26.04.2024

  • Tables got condensed to fit more data on the screen — see How it works

  • Uniqueness / deduplication policy is now a part of the Searches form — see Create a search

  • “Data sources” renamed to “Searches”, “Filters” and “Personas” to avoid confusion — see Find decision-makers

  • To use multiple keywords in one search, you can: — see Create a search

  • One job title per search. Splitting a search into multiple ones with one title each produces results much faster, and also shows which searches are yielding matches, and which are not. — see Advanced search

25.04.2024

23.04.2024

  • Each page (searches, filters, personas, export) showing running status automatically updates when running is finished — see API access

22.04.2024

  • Previous matches are temporarily unavailable — see Create a search

  • Credits are charged per run, not per match, so, hopefully, you’ll get much more matches per credit. — see Pricing

  • Different countries were separated to different searches for the same purpose – you can see first country results faster, before all countries searches have finished. — see Advanced search

  • Searches, Filters, and Personas are separated into different tabs and steps, so instead of having to wait until the whole search & processing will end, you can see intermediate results much faster. — see Find decision-makers

16.04.2024

  • If “Minimum number of valid emails per match” is left empty, emails will not be automatically unlocked, you can unlock them manually — see Create a search

  • You can now choose if you still want to filter out companies without identified decision-makers, using “Minimum number of identified decision-makers” and “Minimum number of valid emails per match” fields. — see Find decision-makers

  • Any query can be run manually. — see Create a search

  • You can now control how often runs are scheduled (and if they are scheduled at all) in each individual search settings. — see Create a search

  • “Max credits daily” setting is removed since match count do not influence cost. — see Pricing

  • The pricing model has changed to monthly run limits. A “run” is one job title, in one country, from one source, per day. To calculate your needed runs: (Number of Job Titles) x (Number of Countries) x (Number of Sources) x (Runs per Day) x (30 days) — see Advanced search

11.04.2024

10.04.2024

  • Search can be opened to see queries. Showing last run date/time, today/yesterday signal/match count, next run date/time. — see Create a search

07.04.2024

05.04.2024

  • Search performance improved significantly by using 2.5x more proxies, and caching all search results — see Pipeline

04.04.2024

  • Fixed a bug that led to “Failed” match status — see Create a search

  • Trial is either 150 leads or 7 days now, starting counting from user getting first match — see Pricing

  • Daily match limit — see Create a search

31.03.2024

28.03.2024

27.03.2024

  • Showing uploaded matches with a white “uploaded” badge in the matches list, and an “uploaded” button in place of approval buttons in match view. — see Anatomy of a lead

  • People without a verified email (only having a LinkedIn URL) can also be matched now. Use “require email” option in search to control this. — see How it works

26.03.2024

24.03.2024

  • Matches can be manually approved — see Anatomy of a lead

  • Search has a new parameter to require manual pre-approval of each lead before uploading — see Anatomy of a lead

23.03.2024

22.03.2024

20.03.2024

  • Fixed: matches randomly appearing and disappearing — see Create a search

  • Fixed: all matches are now visible, not only first 50 — see Create a search

  • Matches are downloadable as CSV, go to Matches -> Download — see API access

  • De-duplication policy in search — see Create a search

  • Matches are shown per day — see Create a search

  • Credits spending can be limited to avoid spending everything in one go – check profile -> settings — see Pricing

  • Run frequency depends on plan now – check the pricing page for details — see Pricing

  • Editing a search no longer deletes previous matches, they still can be deleted manually by deleting items in “runs” tab — see Create a search

  • “Runs” tab showing which queries are running and their run status — see Create a search

  • All users must edit and save their searches for new core to kick in — see Create a search

  • Research and matching core is rewritten, the process sped up significantly — see Pipeline

11.03.2024

  • More integrations are listed in Profile -> Integrations -> Add Integration, some are not yet available. Selecting an unavailable integration will increase priority for that one integration in the roadmap. — see Integrations

09.03.2024

  • Webhook integration. Sending Webhook on new matches. — see Integrations

08.03.2024

  • Message “Search will be automatically repeated every hour” makes it clear that search is not one-time but continuous — see Create a search

  • Free user gets a total of 150 credits — see Pricing

  • Credits now are per-month, not per-day — see Pricing

04.03.2024

  • Matches page shows stats instead of a progress bar — see Create a search

02.03.2024

01.03.2024

25.02.2024

  • Progress bar showing research progress in Matches — see Create a search

23.02.2024

  • Large frontend update: user interface simplified, Researchers + Qualifiers + Personas + Processors were removed, it’s all called “Searches” now. One large form instead of 4 small ones. — see Pipeline

19.02.2024

18.02.2024

15.02.2024

14.02.2024

  • Showing match details on click — see Create a search

  • When subscription plan changes, research starts right away to use new credits — see Pricing

12.02.2024

  • Same company can be matched multiple times within one day by different processors — see Pipeline

11.02.2024

09.02.2024

08.02.2024

  • Dashboard is showing signal count by researcher over time as a chart and in a table — see How it works

04.02.2024

  • Backend performance and stability improved — see Pipeline

  • Company filter by country has ‘exclude’ mode — see Advanced search

  • “Personas” form is accepting comma-delimited values, not only newline-delimited (foolproofing) — see Find decision-makers

29.01.2024

22.01.2024

30.03.2023