Digit - Automated Savings Assistant


As part of this post, I will try to understand how Digit is composed and organized, how the organization is reflected in development and how feature sets aggregate into solutions to key user problems.

I'll start by deconstructing the key features, understanding product developers original goals and how the success of each feature can be measured. We'll then map the features to different user segments and understand their problems and desires. Ultimately products are as successful as the product developers ability to empathize for the user of the product and understand what they are really trying to accomplish. Ok - ready? Let's dive right into this.

Breaking down key feature sets


Automated Savings 

Automated savings is at the core of Digit's features. Once the user's checking account is linked, Digit will note the balance, and then begin to observe the cash flow. There are 4 main things that influence Digit's savings algorithm: Checking account balance, Upcoming income (paychecks or predicted irregular income), Upcoming bills and Recent spending.

Once the saving amount is identified the amount is seamlessly subtracted from the user's linked checking account and deposited into Digit. Intended goal is to enable users save money without them having to think about it. Taking away the complexity of creating budget plans and manually taking steps to stash away money. 

Automated Savings is complemented by FDIC Insurance and Overdraft Protection - the amount that is kept with Digit is insured by FDIC for up to $250,000. Digit also provides overdraft protection to the user, guaranteeing to reimburse the over-draft amount shall a combination of automated savings events cause the user to go overdraft. Both of these elements aim to augment user's trust into allowing Digit to manipulate funds and store their funds. 

Key metrics to understand:

  • What's the magic moment? Understand the dollar amount for each specific user or customer segment and accelerate getting them to it seamlessly - can digit get you to that amount faster by tweaking the algorithm for the first X amount of saves? 
  • Retention: users still saving after 30, 60, 120 days - understand how that retention curve is looking. Are the users that saved for the first time still saving after 30, 60, 120 days?
  • Time to first dollar saved: While few steps must happen before Automated Savings kicks in, Digit streamlines account creation and linking to your bank account - what is the optimal time to the first amount withdrawn from your account? Can we reduce the overall time to the first dollar saved? 

    Automated Earnings

    Digit analyzes user accounts every week and looks for any savings that haven’t touched for three months. Digit then rewards 5 cents for every $100 saved. Users are notified when the deposits are made. They can choose to add the money to the Digit savings account or get it as a gift card.

    Key metrics to understand:

    • Retention: are automated earnings improving overall retention curve?
    • Usage: are automated savings driving higher savings per user ratio?

    What users are saying

    It’s a mental/self control thing but digit has really gotten me excited about saving more because I’m doing without thinking.
    — ashcash85, iTunes
    I have never really been able to successfully save money on my own. I have tried other apps etc. maybe it’s the part of ME actually having to do the leg work or ME being able to see the money store away.
    — Ms. JRabbit, Play Store


    From the dashboard the user sees their current savings balance and complete savings history, as well as some highlights about how Digit has been handling the money.


    Highlights provides basic statistical data to illustrate how digit has been saving for you. Seeing the total dollars that digit has ever saved for you really drives home how easy saving can be when put on auto-pilot.


    Left : Digit Dashboard showing user balance and highlights;  Right : Conversation between Digit and User resulting in saving $10 to savings through simple SMS commands.



    Sums up daily savings. Each row is a separate day that highlights the amount that was saved on that day. Users have the ability to view full history, each click expands the list by five additional days.

    The savings feed consists of a list of daily savings. Each list item contains a total amount saved and a date. The list is sorted by date, the most recent savings are displayed first. The list shows 5 days, events in the past can be retrieved by clicking View More History and is incremented by additional 5 days. 


    Communication is the key part of digit. Users can communicate with digit using commands across SMS, Android and iOS channels with Messenger bot being the latest addition to it's list. Some of the key commands enable users to view checking account balance, digit savings, set-up custom savings goal and manually move one-time amounts from your checking account into savings.                                                                              



    At the bottom of the dashboard, there’s an account management section. Users can change basic account information (name, email and phone number), disconnect their checking account or close the digit account entirely.



    Key metrics to look at, questions to ask:

    1. Exposure: quantifies the number/percentage of users who actually saw and used Highlights, Saving History.
    2. Engagement: commands/user how are users interacting with digit.
    3. Retention: understand the impact the dashboard might have on long-term usage of Digit

    Refer a friend - creating a viral Loop

    For every new user that signs up through a referral Digit rewards the referrer with $5 to their Digit account. The bonus is paid when the friend has their first automated savings transfer. Once the referrer's account is fully setup, Digit will send out an SMS letting the referrer know that friend has signed up successfully.  Once the friend has their first automatic Digit savings transfer, Digit will update you and let you know that your referral bonus has been paid.

    How to measure the success of the referral feature

    Central idea: How does one cohort of users lead to another cohort of users?

    Linear growth channels require incremental resources and effort for new users to be acquired. A loop enables self-fortifying growth that produces some number of incremental users without additional resources

    The loop repeats itself over and over again fueling viral growth. The viral loop is successful if growth multiplier > 1 (GM = 1 / (1-V)

    Initially, the viral loop is fed by linear channels and optimized by constant experimentation till it takes off. One of the key areas of experimentation is the point in user journey when she's prompted to send an invite to friends. (e.g. for messenger this is when I am chatting w/ a friend who has not installed messenger yet and I am prompted to send an invite to her) 

      Now let's explore what problems these features are aiming to solve, and for who?