Highlights
One run history for browser and cloud work
Runs started in the browser used to be visible mainly through the active task panel, while cloud runs had their own history. Enrichment and data-source runs now report into the same collection history with their progress, outcome, and execution origin.
Start a 5,000-row enrichment, close the task panel, and return to the collection later: the history still shows how many rows were processed, skipped, or failed. A browser icon identifies runs that need the originating tab to remain open; cloud runs continue independently.
Improvements
A clearer path from duplicate matches to the right cleanup action
Duplicates Finder used to make every cleanup look like a variation of merge settings. The results workspace now starts with the outcome you want: Remove duplicates, Merge and preserve data, or Custom with AI. It also keeps a bounded set of scored matches, so you can adjust the similarity threshold without rerunning detection.
For example, if two company records share a domain but contain different phone numbers, Remove duplicates warns that data would be lost and offers a switch to Merge and preserve data. Lower-confidence groups stay in Needs review, while safe groups remain ready for batch processing.
Auto-Run only when enrichment inputs change
Auto-Run could already process new rows or follow a schedule, but editing an enrichment input on an existing row still required a manual rerun. The new Run when inputs change mode watches the columns actually mapped into the enrichment—including variables used in prompts—and reruns only the affected row.
For example, if Company Name and Website feed Company Enrichment, correcting either value can refresh that company's outputs without rerunning the rest of the collection. You can also choose whether existing output values are skipped, only empty outputs are filled, or all outputs are overwritten.
Better fallback coverage in Company Enrichment
Company Enrichment now continues through more fallback paths when the first lookup cannot return a complete company record. The workflow is unchanged whether you start with a company domain or a LinkedIn company URL, but difficult records have another chance to return useful data.
For example, a domain that does not immediately produce a company size or location can continue to another matching path instead of ending the row early. Successful rows still return the familiar name, website, LinkedIn URL, industry, location, employee range, and founded year fields.
Dates stay correctly sorted after import and cloud sync
Datetime columns could behave differently depending on whether their values came from a local import or were downloaded from the cloud. Datablist now normalizes valid dates in the browser, so sorting, filtering, and processing use the same date values in both cases.
For example, sorting a Last contacted at column from newest to oldest now gives the same order after a CSV import, a page reload, or a cloud synchronization. Existing cached collections are corrected as they are read; invalid source values are preserved rather than silently discarded.
Delete several columns in one operation
Removing several unwanted columns used to trigger a separate collection update for each one. Datablist now deletes the selected columns as one operation, which is faster and gives the action one consistent success or failure outcome.
After an enrichment test, for example, you can select five temporary output columns and remove them together instead of waiting for the table to update five times.
Duplicates Finder recognizes common versions of the same email
Smart email matching used to miss contacts when one import contained a plain address and another contained an email link or display-name format. It now extracts addresses from values such as mailto:jane@example.com and Jane Doe <jane@example.com>, while keeping the stored cells unchanged.
It also treats jane.doe@gmail.com, janedoe@gmail.com, and janedoe+event@googlemail.com as the same Gmail mailbox. Those dot and domain-alias rules are not applied to business domains, where dots may identify genuinely different mailboxes.
Email waterfalls now recover from temporary rate limits
A temporary rate limit during Waterfall Email Finder or Waterfall Advanced Email Address Verification used to make that lookup step fail immediately. Each waterfall step can now retry short-lived throttling before moving on, while an overall time limit prevents a row from waiting indefinitely.
On a large lead list, a busy verification service can pause and retry instead of turning a potentially valid address into an avoidable missing result. If it still cannot answer, the waterfall continues to the next configured check and preserves a clear row-level outcome.
Enrichment search puts title matches first
Search used to treat a word in an enrichment title much like the same word buried in a description or output label. Datablist now gives title matches priority while keeping favorites, plan availability, and catalog ranking as tie-breakers.
For example, searching LinkedIn surfaces enrichments and data sources with LinkedIn in their public name before tools that only mention a LinkedIn URL as one possible input.
Export rows in the order you are reviewing them
A sorted Datablist view could previously become a differently ordered CSV or Excel file after export. The export dialog can now preserve the active sort, including for large exports prepared in the background.
If a lead list is sorted by Lead score from highest to lowest, check Order By Lead score and the downloaded file keeps that priority order. You can also add enrichment status and credit-cost columns to an export when you need to audit which rows ran successfully.
Ignore URL paths in Duplicates Finder
Duplicates Finder previously treated each URL path as a different value, even when the goal was to identify records that belong to the same website. The URL processor can now ignore paths and compare the host instead.
Enable Ignore paths to match values such as www.datablist.com, https://datablist.com/fr, and datablist.com/pricing. Leave it disabled when individual pages—such as product or profile URLs—must remain distinct.
Save an enrichment without starting a run
Configuring a new enrichment previously led directly into its first execution. A separate Save action now stores the settings, input mappings, and output mappings without processing any rows.
You can prepare Company Enrichment on Friday, have a teammate review the selected outputs, then launch it manually or enable Auto-Run later. Continue with Instant Run remains available when you do want to configure and execute in one session.
Select and delete several sidebar collections at once
Cleaning up a workspace used to mean deleting collections and folders one by one. The Data sidebar now has a bulk-selection mode where you can select several entries, review what will be deleted, and confirm once.
For example, after a campaign you can select its three import collections and their folder in one pass. The confirmation distinguishes data that remains recoverable from permanent deletion, so the time-saving action does not hide its consequences.
The export documentation explains the available formats and export options.