No provenance means no reliable bot removal.
When a graph accepts third-party links without source metadata, it cannot tell whether the join came from a person, bot, or click farm.
With John Kosturos
#SpringDB#DataLineage#AdFraud
Full episode link in the thread.
A company can reveal its next move through hiring.
Job-posting data can show where a business is investing before the impact appears in revenue or market results.
With John Kosturos
#IntentData#B2BData
Full episode link in the comments.
More records do not create more intelligence.
MixRank ties employees, technologies, apps, jobs, and signals to a company-centric graph so every object gains context.
With John Kosturos
#B2BData#DataGraph
Full episode link in the comments.
A cleaner graph requires continuous exclusion.
@BDEX evaluates 2 to 3 trillion signals a month, decides which IDs qualify, removes invalid ones, and rebuilds the graph weekly.
#SpringDB#DataQuality
Full episode link in the thread
A deterministic match can still be wrong.
A bot can be directly linked to a stolen ID. Identity quality depends on validation and disqualification, not only the matching label.
With John Kosturos
#SpringDB#IdentityGraph#AdFraud
Full episode link is in the thread.
New episode: @helloaswin , Co-founder and CRO at @MixRank , on sourcing company, employee, technology, and job posting data at web scale.
Bulk licensing, hourly delivery, owned data centers, and how to evaluate a data vendor for the long run.
Link is in thread
Ad fraud can poison identity data before an ad is served.
Bots steal IDs, mimic consumers, and attach false behavior to real profiles. That contamination spreads through the graph.
With John Kosturos
#SpringDB#AdFraud
Full episode link in the thread
The identity layer can become the product.
BDEX found that email, offline identity, and mobile IDs could not create value until they were reliably linked. The graph became the key to the exchange.
Full podcast link is in thread!
#SpringDB#DataProducts#IdentityGraph
New episode !
David Finkelstein, CEO of @BDEX , on how identity graphs get built, validated, and cleaned of bot driven fraud.
A deterministic match can still be deterministically wrong.
With John Kosturos, CEO of @springdbio#IdentityGraph
Full episode link in the thread.
A data provider without a SaaS product can be easier for platforms to trust.
PDL avoids competing with the products built on its data.
With John Kosturos
#SpringDB#DataProducts
Full episode link is in thread.