Discover how data science reveals
the hardware and software
your prospects need.
To identify accounts who need your product,
our data-mining platform looks for technology adoption patterns.
The scoring algorithms assigns an intensity level to every tech product record. Lower intensity levels are one of the inputs to our next technology purchase algorithms.
Our massive data coverage and product history makes the discovery of such patterns possible.
To find the technology adoption patterns, our data-mining platform
first researches a fire-hose of unstructured data. Each month, our search algorithms scour an average of 1,003,425,389 carefully chosen public pages.
On a human level, it would take 130,417 readers, working 8 hours a day, 5 days a week, reading 200 words per minute, to read this much information in a month.
Our platform finds installed technologies
in front of and behind firewalls. These are two distinct technical challenges that no one else is yet doing.
Behind the firewall. To discover what’s behind the firewall, we extrapolate information from areas like job postings and social forums.
As a quality check, we then compare the same products in front of and behind the firewall.
To interpret this unstructured data,
our platform uses big data technologies.
We date-stamp each product at each company
and track the changes each quarterWe also overlay third-party firmographics
to improve our data quality.
The DemandMatrix platform gives structure
to unstructured data.
Data Quality Processes within the DemandMatrix Platform
A series of automated processes cleans, consolidates, enhances, catalogs and scores the data. So far, we have found 190,711,608 installed instances of products.
Data quality trumps data volume.
We use several techniques to improve our technology data.
Our humans-in-the-loop investigate exceptions (518,410 web pages last year) and feed the findings back into the narrow-AI Engine to improve our big data mining algorithm.
We call this approach “Big Data, Narrow-AI.” (There really are things only humans can do).
Last year, we made 1,342,333 phone calls
to validate samples of our data. This feedback loop drives improvements in our search algorithms.
We have established a content syndication and
appointment-setting services. We use these interactions as a feedback loop to improve our search algorithms.
DemandMatrix Integrates Real Humans into its Platform.
No, we don’t see an elimination of humans on our product roadmap.
From data scientists to data verifiers, our Humans-in-the-Loop services plug and play with our platform.
Use them to build predictive models, validate your models, generate leads, test a proof of concept, or to syndicate content.