Plaid: Introducing a new categorization taxonomy and category list for personal finance management apps and tools

SHARE

May 17, 2022

Build deeper user connections with data driven insights

Royce Kok

The financial lives of consumers are evolving at an unprecedented pace as a result of the pandemic. From the great resignation to the popularity of cryptocurrency to the adoption of digital platforms like Disney+, it’s clear that the spending behaviors and financial needs of consumers are constantly shifting. Additionally, fintech has gone mainstream which means consumers have more choices than ever in the applications they can use to manage their finances, and they have higher expectations for what these applications can deliver. 

Needless to say, the stakes are higher than ever for fintechs and financial service providers to differentiate their product offerings to acquire and retain their users.

We often hear from fintech innovators like you that in order to build engaging digital experiences, you need robust transaction data and insights into a user’s committed monthly expenses like rent and their expected recurring deposits. With this, you would be able to offer even better personalized experiences and tailored recommendations that address each user’s unique financial needs.

To equip you with the insights you need, we’re excited to introduce Recurring Transactions. This is a new endpoint that will enable you to easily access insights about the recurring inflows and outflows in a user’s credit, checking, or savings accounts. A user’s recurring transactions like subscriptions or bill payments can significantly impact their cash flow and committed spend for a given month, which is why it’s important information to reflect in their financial picture.

With readily available insights, you can free up engineering resources that would be used to maintain your own machine learning models to detect your user’s recurring transactions. You can get started in no time to build engaging features like payment reminders that drive value for new and existing users.

Create digital experiences that maximize user value

Insights into the recurring transactions for a user can enable you to build new tools that drive engagement and ultimately, revenue for your business. As users’ financial lives evolve, they need and expect more assistance from personal financial management apps or tools to meet them where they are today, such as help with managing subscriptions or loan payments. Using the Recurring Transactions endpoint, you can help users:

  • Understand their committed spend for monthly expenses like rent and insurance payments.

  • Reduce spend by canceling unwanted subscriptions.

  • Stay on track with upcoming payments with in-app or text reminders.

  • Transition their subscriptions and bills from an external account to the banking account or credit card that you’re offering.

How Recurring Transactions works

Once a user’s account is connected via Plaid Link to your app, call the /transactions/recurring/get endpoint. Plaid will then analyze the transaction history from the account and group all transactions that occur on a recurring basis, based on the description, amount, and cadence. Then we tag these as inflow and outflow streams. A matured stream is defined as having at least 3 occurrences. Finally, we provide insights into each recurring stream including:

  • category and merchant name

  • frequency (e.g. monthly, semi-monthly, biweekly, or weekly)

  • average and last amounts which can be used to determine changes to bills or subscriptions

  • whether a recurring stream is active or not

  • the individual transactions that make up each stream

Additionally, in our beta program, we learned that customers wanted to identify new recurring streams sooner. With this in mind, we developed an early detection feature for recurring transactions that have yet to mature (less than 3 occurrences). Access to timely insights about a new subscription or bill will enable you to move faster in providing the help users need to stay on budget. In the API response, look for ‘early_detection’ as part of the ‘status’ field to find these new streams.

Finally, we know that many users make frequent purchases for gas, groceries, or coffee from their favorite cafe. To help you understand the expenses and deposits that are recurring as opposed to ones that are habitual spending, we exclude these frequent purchases from our detection model.

Here is an example of the API response from the Recurring Transactions endpoint which shows an outflow and inflow stream.

1{
2 "update_datetime": "2022-05-03T10:34:50Z",
3 "inflow_streams": [
4 {
5 "account_id": "lPNjeW1nR6CDn5okmGQ6hEpMo4lLNoSrzqDje",
6 "stream_id": "no86Eox18VHMvaOVL7gPUM9ap3aR1LsAVZ5nc",
7 "category": ["Transfer", "Payroll"],
8 "category_id": "21009000",
9 "description": "Platypus Payroll",
10 "merchant_name": null,
11 "personal_finance_category": {
12 "detailed": "INCOME_WAGES",
13 "primary": "INCOME"
14 },
15 "first_date": "2022-02-28",
16 "last_date": "2022-04-30",
17 "frequency": "semi_monthly",
18 "transaction_ids": [
19 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57a1",
20 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57a2",
21 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57a3",
22 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57a4",
23 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57a5"
24 ],
25 "average_amount": -800,
26 "last_amount": -1000,
27 "is_active": true,
28 "status": "mature"
29 }
30 ],
31 "outflow_streams": [
32 {
33 "account_id": "lPNjeW1nR6CDn5okmGQ6hEpMo4lLNoSrzqDff",
34 "stream_id": "no86Eox18VHMvaOVL7gPUM9ap3aR1LsAVZ5nd",
35 "category": ["Service", "Utilities", "Electric"],
36 "category_id": "18068005",
37 "description": "ConEd Bill Payment",
38 "merchant_name": "ConEd",
39 "personal_finance_category": {
40 "detailed": "RENT_AND_UTILITIES_GAS_AND_ELECTRICITY",
41 "primary": "RENT_AND_UTILITIES"
42 },
43 "first_date": "2022-02-04",
44 "last_date": "2022-05-02",
45 "frequency": "monthly",
46 "transaction_ids": [
47 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57gO",
48 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57g1",
49 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57g2",
50 "vr5mDLjd1nhMj0V7LmEkILDnmxAoZaHmk57g3"
51 ],
52 "average_amount": 85,
53 "last_amount": 100,
54 "is_active": true,
55 "status": "mature"
56 },
57 {

Get started

We recommend reviewing our API documentation for additional details about Recurring Transactions. If you’re interested in using the new endpoint, request access by contacting your Plaid Account Manager. 

Recurring Transactions is available in the U.S. and Canada as an add-on feature to the Transactions API. Recurring Transactions will be available in the UK later this year. 

This is just the start of the data insights that we’re developing to help you build more personalized and engaging experiences that can lead to better financial outcomes for your users. We’d love to hear from you if you have feedback about Recurring Transactions or any other insights you would want Plaid to provide. Send us a note with your feedback and requests: transactions-feedback@plaid.com.