Our focus is on data which provides 1) value to the customer, 2) quality, and 3) timeliness. We gather data ourselves as well as purchase and partner. In most cases, we use multiple sources and back-test those sources against each other. When data is missed, or late, our research team goes to work to backfill the missing data by hand. We collect a wide variety of data including County Assessor, County Recorder, County GIS, Listing, Demographic, Phone, Email, Social, Census, and More.
That turned out to be almost impossible. Data vendors wanted us to charge our customers per record, and they wanted a percentage of each record sold. It took us a year to get one vendor to agree to our new pro-small-business pricing model, and that was only for one dataset. Fortunately, we were also already gathering data ourselves that we could sell however we wanted. And thus, our original, foreclosure-only, service was born.
Since then we’ve earned the trust of many big vendors for this new pricing model. They’ve realized we aren’t after their lucrative deals with big business, allowing us to add many new data sets cost effectively for our small business customers. We’ve added more of our own data as well.
But cost-effectiveness can’t be your only criteria for data acquisition. We also look for timeliness, accuracy, coverage, and population. To that end we do bake-offs between vendors, we backtest against county records and other sources, we use multiple sources, and we have staff manually backfill as needed.
This multi-prong approach is rare in the big companies and unheard of in small business offerings. Even with all this effort, public records data will never be perfect, but through feedback from thousands of customers, we know it sets us far apart.
Our stream-based architecture, scalable event matching and criteria based targeting give us a powerful engine which processes millions of events against our customer’s unique criteria to keep their lists and segment up to date, notify them of changes and automate their marketing efforts.
When it comes to taking action, big companies have built their systems around batch processes and big campaigns. We saw a better way. Stream-based, individually matched, immediately actionable.
From the beginning, we’ve taken an event-driven streaming approach to building our systems. This gives our customers faster access to new data – the events in the life of a person or property which our customer can leverage to grow their business.
Unfortunately, a firehose of data is of very little use to most small businesses, as they have neither the time or ability to digest it all. This is where the power of targeting comes in – it allows our small business customers to focus on the right events to grow their business, using our 200 plus criteria, and easy preset Quick Lists to get them started.
That level of event-based targeting is no small feat. We process hundreds of thousands of events each day, matched to tens of thousands of customer criteria sets, leveraging hundreds of criteria.
The end result? The ability to alert our customers quickly to the best new opportunities, and to automate actions against those opportunities on their behalf. No one else in the industry offers anything close.
We’ve automated data cleansing and enhancement through rules engines, pattern matching, natural language processing, machine learning, normalization and other techniques we’ve honed over a decade. Our research team manually back-tests data, corrects data, and uses their insights to create new cleansing and enhancement methods. We also create new data from unique combinations of existing data, or from data models, like loan positions, estimated values, equity and more.
There are 3,142 counties (and equivalents) in the United States, and each has a different idea about how data should be gathered and stored.
Worse the number of errors is atrocious. Our background in data science led us to recognize early on that we could fix many issues that would drive our customers crazy, and we built our systems from the ground up with that in mind.
Our first service in 2007 launched with a rules engine that evaluated nearly 200 rules on each new record.
Since that time, we’ve grown from just simple pattern matching to natural language processing, machine learning and other modern techniques to find and correct errors, normalize data across counties and sources, flag problems for staff to research, and even create new data.
Our estimated loan positions and transfer types are just two examples of data that doesn’t exist directly in public records, but which we instead create from those records.
We also leverage the expertise of best in class vendors for certain data models like estimated property values (automated valuation models).
We leverage modern user experience design and software architecture to deliver some of the industry’s most powerful solutions in a package that is simple for anyone to use. Our small business customers are not expert data scientists, software developers or even marketers, so they rely on us to make them shine without the steep learning curve so typical of other solutions.
They are typically not expert data scientists, integrators, and marketers, let alone have the time do those things even if they were experts in them.
From day one we’ve worked not only to bring the best data to small business but to get them out of the business of importing and exporting data, integrating data and systems, building platforms, selecting vendors and all the other tasks necessary to be expert targeted marketers.
Instead, we do it for our customers, making it easy for them to leverage data to connect with customers and grow their business.
We leverage modern data models, sophisticated matching algorithms and graph-based relationships to tie records together across sources and over time to create our unique OwnerGraph™. This helps us discover relationships, trends and other new and interesting data that can’t be found using the industry’s typical “flat-file” approach.
Historically public records have been treated as flat files, that are simply overwritten with new records as they come in. While that “flat” data is amazingly useful, that approach leaves hundreds of opportunities to gain insights into the relationships between people and properties completely untouched.
Our unique OwnerGraph™ ties records together across sources and over time to discover those insights.
We started this work in 2006 with the launch of our best-in-market Transaction History, which showed the relationships between recorded deeds, loans, foreclosures, assignments and other documents over time. From there we added the relationships between those documents and properties, properties and owners, owners to each other, owners to – you get the idea.
From the beginning, we saw that there was often more value in understanding the relationships between these bits of data, then in the data itself. Something all the data companies focused on delivering batches of flat files, sold by the record, have missed completely.
Over 10 years we’ve still barely scratched the surface of what we believe is possible. Our work on the OwnerGraph™ is one of our most exciting endeavors and we can’t wait to show you more.