How we support Get Paid Early 💸

What we didn’t expect was that a huge wave of these requests would hit us at exactly 4pm

My immediate thought: These are the people who are not very good at money management and budgeting.


Could be financial anxiety (me) and I’d like it sorted asap and where it needs to be so I can stop stressing.


You (Monzo) tell people that at 4pm you can use the feature, I’m surprised it never occurred to you that this is exactly what people would do.


GPE - forcing the playback of ‘topup.mp4’ files everywhere - best feature ever…


Very cool reading, thanks for sharing!


I consider myself the opposite! I like to get everything organised and into my pots for the next period.


Totally disagree.

Getting my funds earlier means I can get all my slicing and dicing of salary sorted a day earlier, so any bills and such just go out smoothly and all my savings are completed and done.

I do not consider myself particularly bad at money management.


I guess I meant those people who sit there counting down the seconds until 4pm, not those who just get things in order good and early.

(Sorry @Revels)

1 Like

We’re an anomaly really.

There’s been threads about people needing to get paid early in order to pay a bill/eat.

To try and get back on topic… Joe, are you able to share some numbers behind GPE? How many people use it that could? How many get it before 5pm, 5-6pm etc? How many use it frequently?


im sure if you dont collect after 4 monzo send you a ping to say you have cash waiting ( or used to do) i make frequent use of the get paid early as some one else said it lets me get all the money aside for bills ect sorted without stressing which im sure counts as good money managment rather than ‘bad’


They still do!

Thanks for sharing, Joe!

A related (I hope) question: are you seeing different scaling challenges as the US grows? Presumably the CPU usage by time graph will start to flatten as there are more US users?

Granted, there probably aren’t enough yet to make an impact, but as they start to scale…

This is the same as me - I use the feature because I like to set everything aside the moment I can.

The impact of GPE across the platform also supports that this is the most common usage - e.g. customers don’t get their money then close the app, they immediately start doing lots of things with it like sorting, settling bills, moving to pots, and even ATM withdrawals.


Agree, if you get paid a day early every month then there little or no benefit at all.

Hey, that’s a great question!

I can share with you some high level data:

  • Each day ~70% of payments that can be paid early, are.
  • More than half of those happen in the first 2 hours of GPE.
  • After the first few minutes, the traffic drops steeply, levels off and we then see a long tail that clears completely at around midnight.
  • Over the last week, the highest GPE usage was from customers who joined back in 2016, and customers who joined in 2020.
  • Since the feature began, we’ve paid over £17.5b early.
  • We’re trending towards regularly paying ~£1b early each month.
  • 99.4% of users who are shown the GPE carousel item use the feature at least once.

Is that interesting data? :sweat_smile:

(I edited this as I’d said m before not b ! :person_facepalming:)


It’s not so much that no one thought people would do it… more that no one thought as many people would do it as actually do ! :sweat_smile:


It’s great! Love me some numbers! Thank you


That’s right! We actually see a little burst of new activity at ~4:25pm when the push notification has gone out. Up until that point, the number of requests are dropping sharply from the initial burst, but they then level off as a more steady stream of customers start to request their money too !


Great question!

We’re not seeing different scaling challenges as the US grows just now, but you’re right that the graph is likely to flatten off.

I think what we’ll see (assuming a good spread of users across the USA relative to the population in the area) is a big bump when Eastern Time wakes up, then smaller bumps as Central Time and Mountain Time users wake up, then a big bump when Pacific Time wakes up.

Aside from GPE events (or similar), our autoscaling should pick this up without us really needing to do anything special to handle it.