Faraday predictions API

Easy customer behavior predictions

The Faraday prediction API allows you to easily make predictions about your customers' behavior using a combination of your data and Faraday-provided, responsibly sourced third-party data.

APIWhat it's for
StripeReceiving payments
TwilioSending SMS
PlaidBank account data
SendgridSending email
Faraday Customer behavior predictions

What do I need to use the Faraday API?

If you can get a CSV with any combination of physical addresses, emails, hashed emails, and/or phone numbers of US residents who have transacted with you, signed up for a mailing list, or come into your system as leads, you can probably use Faraday.

Any programming language will work, and we provide API examples using curl so that you can try things out without any scripting at all.

Who is the Faraday API For?

  • Consumer-facing brands and other teams that want to learn more about their customers and anticipate customer behavior will get a lot of value by using Faraday, including its API, directly.
  • Platforms that serve these companies as clients can add value to their own products by using the Faraday API to build in predictive capabilities.

The API is specifically for engineers who want the simplest way to make predictions about consumer behavior. If you don't have an engineering team, our enterprise offering has you covered including intuitive configuration for non-data scientists and end-to-end delivery to a wide variety of tools.

What do I get back from the Faraday API?

We can make 3 kinds of predictions for you:

  • Propensity predictions about specific actions to support goals like lead scoring and likelihood to buy again.
  • Persona predictions that organize your customers into bespoke persona systems to support goals like personalization.
  • Potential predictions that forecast expected customer activity, including LTV and churn.

We deliver predictions back to you as a CSV and/or replicate them into your data warehouse. You can join them back to your original data because we return the PII that you sent us.

| first_name | last_name | primary_email        | fdy_persona          | fdy_propensity_percentile |
| NOVA       | SMITH     | [email protected]     |  Relaxed Retiree     |                        88 |
| JOHN       | SMITH     | [email protected]     |  Bargain Hunter      |                        62 |
| MARGARET   | SMITH     | [email protected] |  Bargain Hunter      |                        69 |
| BECKY      | SMITH     | [email protected]    |  Late Career         |                        55 |
| SERENITY   | SMITH     | [email protected] |  Young Urban Pro     |                        92 |
| ANAIS      | SMITH     | [email protected]    |  Young Urban Pro     |                        52 |
| MIQAELLA   | SMITH     | [email protected] |  Relaxed Retiree     |                        83 |
| SIMONE     | SMITH     | [email protected]   |  Suburban Family     |                        62 |
| KYRA       | SMITH     | [email protected]     |  Suburban Family     |                        37 |
| JANET      | SMITH     | [email protected]    |  Active Lifestyle    |                        80 |

Is it secure?

Security is Faraday's top priority because we can't operate without your data. We have been in business since 2012 and have handled hundreds of brands' PII data.

How does the Faraday API work?

  1. Upload your customer data through the API. Typically these are orders, leads, or contact lists of some kind. We enhance this data with hundreds of responsibly sourced consumer datapoints from the Faraday Identity Graph.
  2. Choose what prediction you want to make using a simple, declarative syntax that does not require data science experience. Faraday automatically builds, validates, and continuously updates the corresponding AI models behind the scenes.
  3. Indicate what group of people you want predictions about. Faraday automatically applies your AI models to this population, producing and continuously updating the predictions.
  4. Retrieve your predictions on individual people in real time, referencing them by name/address, email, and/or phone.
  5. Retrieve predictions in bulk by securely pulling from a CSV that we host, or having us replicate them out to your data warehouse.

We recommend you try one of our quickstarts - Personalizing your outreach or Finding repeat purchasers. You can get sample data easily.

What's next

Get started with a quickstart tutorial