We talk a lot around the office here at FullContact about the power and extensibility of our graph, yet up until now we’ve limited our customers’ ability to harness its full power. That acknowledgement–among some other important ones–was a driving force behind the latest release of our Enrich API. We spent a lot of time in the market listening to our customers, and we’ve matched that feedback with our own candid assessments of the platform. The result is a different kind of API, one that we’re pretty proud of.
If you haven’t been to FullContact’s website to learn about the upcoming release, check it out here. For this blog though, I’d really like to look at what drove the development process itself and why we chose the features that headline this pretty size-able release.
Multiple Input Queries
PROBLEM: There’s only so much you can do with one query key at a time
SOLUTION: Multiple Input Parameters
Our ability to make connections between individuals, those things that accurately identify them, and that which describes them is truly at the core of what we do. Our graph-centric model (something we’ve recently patented) underlies every product and service that FullContact offers. The problem with a one-dimensional query of that graph is that it can result in more limited responses.
By offering hints along with a primary key, our customers will be able to ask questions of our identity graph in ways that the graph is built to field. As an example: were you to query ‘email@example.com’ by itself, you may find a result that reflects only those digital identities that reference this email address; maybe a location, company, and title. If instead you were to query that email, along with an age, a twitter handle, and a full name, what you’re doing is giving the graph more evidence to consider in the response. The inclusion of those additional linkages has the ability to unlock entire clusters of identity. The twitter account that you link may also be tied to an instagram account as well as a pinterest account. By tying one, the others become linked as well. Thus the new response would have a great deal more to tell about me.
PROBLEM: With all of the possible data and associated use cases, there is no single recommendation for how frequently customers should poll for updates.
When spending time with our customers, we get this question a lot: “How often should I update my contact records?” The answer usually results in a conversation, and the recommendation grows from a combination of factors such as what changes are relevant to your business, the size of your customer data warehouse, whether you cache data or transmit and delete, along with many other factors. What’s more is that even when you consider the many variables that generate different recommendations, there are also universal ones, like end-user privacy, which dictate the need for recency.
When you enrich a contact with a subscription, what you’re really doing is dissolving all of those questions and complexities. You’re removing all of the business logic that prioritizes queries based on internal recency because, simply put, you’re only querying once anyway. By depositing a set of customer records in a secure directory for ongoing updates, you are enabling FullContact’s intelligence layers to constantly search for new connections, new matches, and updates to existing connections. The outcome is not only more and better data on today’s matches, but also incrementally increasing matches over time. As FullContact learns new things, so do you.
PROBLEM: Once the question of identity is answered, there is no end to the amount of data that can follow. How do our end users get access to more information about the people FullContact knows?
SOLUTION: Data Packs
FullContact serves a lot of different use cases for a wide variety of industries, verticals, and specialties. Brands use FullContact for a better customer 360, some SaaS platforms use FullContact for fraud prevention, services firms use FullContact for more personal and informed customer care. There is a core need underneath each of these use cases that lies in identity resolution. But beyond that, each one begs for different variations of the accompanying data. Psychographics have proved to impact the brands most, where email and social verification prove most useful in fraud prevention, and geo-demographic data is always needed for customer care.
Data Packs are a way of bundling different kinds of data (whether internally generated or sourced from a 3rd party) so that each one of our customers can get more of the data that they need and less of the data that they don’t.
That can’t be all there is to say about that…
Here at FullContact, we think we know identity resolution and data enrichment APIs as well as anyone, but then we’re also pretty sure that even as recently as 5 or 6 years ago that wasn’t even a thing :). We’re pioneering something here that has grown rapidly and now has a lot of different users with wildly varied needs. This major release is the encapsulation of both your feedback and our own, so we hope it starts to address your business needs even more effectively than the last version. The three “pillars” of functional change described above are only the most dramatic improvements from our current version, but there are many more nuanced differences to discover. As always, we love feedback from our users so please come to our site and reach out, or if you’re already a customer contact your handler (customer success rep or account manager).