February 25, 2020
On-demand updates to transactions data
Ashley Cornall & Kina Abe
Updated on February 25, 2020
We’re excited to introduce a new optional refresh feature for our Transactions product!
Fintech has made managing your money so much easier, from rounding up payments to accelerate your saving goals, to identifying opportunities for reducing your spending or automating your expense reports. At the center of all of these tools is the access to reliable and comprehensive transactions data, and for the last 7 years, Plaid has been making it possible with our Transactions product.
We know that when it comes to transactions, fresh data is always better. Having the ability to deliver updates on demand lets you delight your users with timely interactions and insights, as well as allowing them signal when they want up-to-the-minute transactions data.
Introducing Transactions Refresh
Transactions Refresh is a new endpoint, which lets you pull user transactions on demand.
Transactions automatically schedules updates and notifies you proactively when new transactions are available. However, there are some situations where you want to provide greater interactivity for your user or need to ensure transactions are current as of the specific moment. For example, a user who just booked their flights for a vacation might want to check what funds they have on hand for the trip. Or a personal financial management app user may want to find out sooner when their paycheck has deposited into their account.
We’re excited about how Transactions Refresh will enable you to improve user experiences when you have even more timely transactions data. You’ll be able to draw fresher insights based on your users’ latest transactions.
An early beta tester, SmartPath focuses on helping individuals and families achieve long term success through unbiased financial coaching. Here’s how SmartPath is using the new feature:
With this new feature from Plaid, we were able to fully implement a "pull-to-refresh" feature that our users expected us to have. I know I can get impatient waiting for my bank to pass data through, so this feature eases my mind that I also have control in seeing updates when convenient for me.
Stuart Lawder, COO & Co-founder of SmartPath
How it works
First, you’ll need a way for users to indicate that they’d like to update their transactions data. This is typically done with a refresh button or a pull-to-refresh interaction on a mobile device. Then you can call our /transactions/refresh endpoint, and we’ll kick off an on-demand update for the connected Item to see if there are any new transactions.
You'll be notified of any new transactions available through Transactions Refresh via standard Plaid webhook notifications. You can then call the /transactions/get` endpoint in order to retrieve these latest transactions and account balances for use in your app. Most existing Transactions customers have already implemented these steps, making it simple to add this feature to your integration. Check out our API docs to learn more.
Getting started
The /transactions/refresh endpoint is available in the Sandbox, Development and Production environments. Anyone with API keys can begin testing this endpoint today in our Sandbox environment for free. If you don’t have API keys, sign up for a free account. For access to our Development or Production environment to test with real accounts, reach out to Sales or your Growth Account Manager.
What else is new with Transactions?
We know that transaction data is a critical part of many of your products, and we are constantly exploring ways to improve it. Early January, we released three new data fields: store_number
, authorized_date
, and payment_channel
. Use these fields to better identify where, when, and how a user made a purchase. See here for documentation.
In the past few months, we’ve also improved the accuracy and coverage of our categorization and parsing models. For example, we added a new rules-based layer that lets us fix categorization on data that was too specific and quirky to be accurately categorized by an ML model. We’ve also added layers to our location parsing logic to increase coverage.
Stay tuned over the coming months for more features and product updates.