Spending


(Hugo Cornejo) #1

Hi all, if you are wondering what hides under the top right button of the Mondo app, here’s a glimpse.

The question that we’re trying to answer in this section of the app is “Where is my money going?”. This is one of those questions that appears again and again in our customer interviews so we know almost everyone really cares about.

So far we are considering to answer the “Where?” question in three possible ways:

  1. Category: “My money goes to groceries”.
  2. Location: “My money goes to Shoreditch”.
  3. Merchant: “My money goes to Sainsbury’s”.

As you can see the navigation is familiar and we hope will relate nicely with how the feed works. At the same time this is the first area where we’ll get a big value out of our categories and merchant location.

The challenges ahead? A few, all of them full of interesting design trade-offs here and there (really fun stuff for designers :smile: )

  • We need to be very careful to avoid data porn just for the sake of it. Particularly when showing maps we want to be sure that we’re actually giving you valuable information and not just noise (classic PFM mistake).
  • Prediction and action. With this solution we’re not telling you if you are spending more or less than your average month, that’s a huge question and we’re working hard on it.
  • Reporting. This solution expects you to go to the app to check how things are going. That’s ok but we still need the second piece of the puzzle, something that can send you reports when Mondo realises relevant events on your spending. In other words, we want to be able to send you a feed item saying “Hey, just for you to know, you’re spending every day £30 more than this month last year”. Then, if that’s relevant to you, you can dive in and understand why is that happening.

That’s all for now, this is a big piece of the app that we’re sure we’ll iterate and improve a lot.

Any feedback is highly appreciated. What do YOU expect of this kind of tool? Are we on the right path?

Thanks!
Hugo


Forecasting your spending
This week at Mondo Towers
Breakdown
(James Allison) #2

Looks awesome - needs pie charts! Would love to select weekly/monthly expenditure in a pie chart, with the slices being categories. (or retailers & locations maybe)


(Alistair) #3

Hi Hugo,

Look forward to seeing it in action!

What I have been noticed so far from existing features is just how much I spend with TfL on a normal month. I guess what could be useful is advising how much taking a bike to work could save me / how soon I would pay it off. But then knowing about travel time departure/arrival location would be required.

Or in the above contexts how much using Uber instead of Halo would save. Similarly for Starbucks something like buying yourself a fancy coffee machine would save you x per year could be interesting to analyse.

You could get really in depth if you got more data from the merchants - e.g. GPS locations of departure & arrival or full itemised bill which would appear on the receipt. But I assume this isn’t available to you.


(Samuel Michael) #4

:wave: Hey James, we have experimented with pie charts / bar charts and found they don’t actually bring much from a usability stand point other than look pretty at this point.

Instead, we are offering a simple list breakdown of monthly category expenditure from highest > Lowest.

Interactive pie charts can get quite fiddly. In order for them to be useful you need to pack a lot of info in a relatively small area. Colour usually isn’t enough, each segment needs an icon or title and this can become problematic when segments have much smaller expenditures compared to others.

In the future, when we look at yearly expenditure overviews we may bring bars & pie’s into the kitchen again, but for monthly spending at this point they don’t seem to offer much.

Sam :slightly_smiling_face:


(Alexander Baxevanis) #5

I see 2 challenges with such ‘spending analysis’ tools:

  • Accuracy of categorisation: If I spent £200 at Sainsbury’s, I could have bought a flat screen TV or a case of wine or Christmas dinner supplies or just a normal grocery shop for a large family. Perhaps in the future there will be some AI that can understand such nuances e.g. ask me to recategorise a transaction if it doesn’t match my typical spending pattern with a retailer, or remember that if I go to M&S at lunchtime I get lunch, but if I go at weekends I do my grocery shopping. In the meanwhile, could Mondo allow you to ‘split’ a transaction and categorise different parts e.g. my £100 spending at ASDA is £50 groceries + £50 shopping for clothes. This could also work well for cash (for as long as people still use cash) as you get no indication of what happened with the money you took out of an ATM

  • Creating analytics that help people make decisions: it’s fine to show me how much I spent in one category, but what am I supposed to do next? Spend less? Save some money if I spent less than usual? If it doesn’t drive some action, people will eventually get bored looking at the raw numbers. The issue here is understanding which actions make sense depending on how people manage their money … Someone having a financial cushion of a few thousand pounds in their account will be interested in different stuff than someone anxiously waiting for their salary to land in their account


(Saveen) #6
  1. Lloyds Bank offers a free ‘Money Manager’ service within its banking app. I discovered this via a friend over the weekend.

Featuring pie charts:

  1. One of these images is from an interesting blog by Aden Davies: http://www.adendavies.com/please-give-feedback/

  2. I think it would be beneficial to create a survey (Typeform or other) of the good, the bad and the ugly features of all major banking services / apps. I have one ‘good’ suggestion right now, where shall I add it?


(Hugo Cornejo) #7

futureshape, that’s spot on, thanks! It happens with cash withdrawals too, people want to categorise that opaque spending. At the same time we want you to not need to manually add this kind of information or make any efforts. It’s a tricky problem.

Regarding your second question that’s exactly what I meant by “Prediction and action”. We are working on it :smile:

saveen, YES, please share them here or (if they have sensitive info) send them to hugo@getmondo.co.uk Thanks!!


(Saveen) #8

Actually, perhaps not just limited to banking app features.

1) Statements (NatWest / RBS)
Not a huge fan of NatWest but I do like the ability to view, save and print statements and certificates of interest up to 7 years. Not many online services go back so far but I hate having to contact support, then wait for a ‘special’ print run for older statements. I feel that paperless should always be convenient.

NatWest offer this via desktop web not their mobile app. As Mondo is mobile-first, this will require some UI considerations. PDF & CSV options via email, Dropbox etc. would be great.

  1. Last account activity (Gmail etc.)
    Device, browser, location, IP, date/time for security.

  2. Customer support (Amazon)
    Not sure how feasible this is but Amazon offers a fabulous customer support interface with “Call me now” button.


(James Billingham) #9

Also, this isn’t particularly related to recent spending, but while APIs are great, it’d be really good if you could get CSV downloads of transaction data - for people who are a little technical, but not developers.


(Hugo Cornejo) #10

Very good suggestions. Thanks!!

I think we have them all in our backlog and “badass ideas” board (the place where we store all the cool ideas that we want to consider building at some point). I hope we can get solutions for those scenarios soon :slight_smile:


(Saveen) #11

@futureshape raises a very good point. To get a truly accurate picture of “Where is my money going?” it would be useful to delve down to product-level detail.

For example, today I might buy a sandwich at Pret and that will be categorised as ‘Eating Out’ but tomorrow I might purchase another sandwich from Tesco along with a book, a bottle of wine and maybe even a TV. Would these be classed as groceries?

Perhaps the solution will be to:

  1. Extract information via the ‘Add Receipt’ feature like cloud-based receipt organisers eg. Lemon.

To upload paper receipts, people can download a mobile app and simply snap a picture of the receipt. Lemon will subsequently digitize and store the data.

Lemon squeezes out all the juicy details [receipts] contain to help you stay organized, see where your money goes and save some cash along the way. Lemon extracts data from each receipt down to the product-level detail to generate insightful reports and graphs that illustrate spending trends, make it easy to prepare for tax time or submit expense reports.

Over time, Lemon will also offer users targeted discounts and promotions from relevant brands and retailers based on their spending habits, giving users access to exclusive deals, future discounts on products they purchase often and personalized offers from their favorite brands.

  1. Allow custom Categories and Tags (similar to Gmail labels).
    Tags could provide a second layer of manual categorisation separate to categories. Like Gmail, multiple tags (labels) could be assigned to a purchase unlike categories (folders) and filters could be set up to assign them automatically.

(Hugo Cornejo) #12

Wow, that sounds like an excellent hackathon project, are you game? http://attending.io/events/mondohack-3


(Adam Horner) #13

I’m really with @billinghamj here on the CSV export thing - although on iOS a direct export to Numbers would be a good primary option in my opinion with fallbacks to export to CSV files in either iCloud Drive or Dropbox (both of which have APIs I believe).

Also some way of exporting the CSV along with linked photos of the receipts would go a long way to helping with one of the most hating tasks of executives everywhere: expenses. Direct export to Expensify? That would be literally life changing for several sales people I know.


(james_e_bell) #14

I guess there are quite a lot of reasons a user might be wanting to view their spending so its hard to design for them all.

For me personally I think there are exciting things that bank that ‘knows me’ could do.

For example what about identifying trends in my discretionary spending to help me manage it? Am I spending more or less on lunch / coffee each day than I normally do? Could I be considering going to a different merchant nearby to save money (or maybe there are more popular/highly rated ones)? I think some things here can be done without line item information e.g. if I spend each day between 11.30-2 its probably my lunch

I find the current category list a bit difficult to use for the purposes of tracking my discretionary spend as giving ‘eating out’ there are several types that mean very different things to me from a money management point of view:

    • What I spend on lunch each day
    • What I spend when going out for meals with friends/family etc (would spend a lot more but am interested if I’m spending more or less on this over a month etc!)
    • What I spend when eating out on holiday (I see this more as a ‘holiday’ cost and would be more interested in the cost of the holiday overall)

Use transaction time to split eating out into breakfast, lunch and dinner
(samdickie) #15

The uncategorised amounts are a pain! I would be interested in how mondo tags all spending? money dashboard is a nightmare for that


Uncatogorised spend
(Saveen) #16

Agreed. The challenge for team Mondo is how to achieve this without requiring a lot of manual effort from ALL users.

My feeling is that “Where is my money going?” automatic categorisation should work straight out of the box for all users albeit with these limitations. For users who require greater granularity, a second layer of manual tagging (like Gmail) could help to solve this along with receipt OCR mentioned above. Then crowdsourcing this manual effort plus a bit of simple AI could assist with automation.


(Clive Richardson) #17

On prediction and action: it’d be handy to have a feature which nudges you to rein in discretionary spending, based on known upcoming transactions. You’ll know when (large) regular direct debits or standing orders leave my account - mortgage, car finance, childcare - and when salary is paid in (or multiple salaries assuming joint accounts will be a thing)

Getting a warning a few days before a large debit is likely to leave would be useful - you’ve spent £x today, your balance is £yyy, and you have £zz leaving within the next few days - would nudge you to head to Lidl rather than Waitrose (or, conversely, maybe think about paying off more of your credit card bill).


(james_e_bell) #18

I like that idea - also getting a clear view when your salary comes in of how much is available for discretionary spend (and have that count down separately to total balance?)


(Hugo Cornejo) #19

Totally agree @cliverchrdsn Thanks for the suggestion :slight_smile:

So far (because we don’t have recurrent payments) we are not doing anything like that but at some point we will do. Probably via the pulse (that little graph on top that already estimates your spending based on previous months) + bespoke notifications we’ll try to give you a proper understanding of upcoming transactions.

I hope down the road we’ll be even able to investigate some intelligence around seasonal events (we tend to spend more on xmas, summer holidays, etc.) but I presume that must be much much harder… at least it sounds like sci-fi to me :rocket:


(Stephanie - Alpha) #20

In the past I’ve used “Spending” app to do my analysis - found it useful, except for the fact you had to manually add everything in. Mondo will completely replace that app.

One thing which would be awesome is the ability to break a particular transaction down for analysis - for example if I go to Sainsburys and buy my groceries but also pick up some beauty/health products, it would be good to be able to separate that transaction.

Also the totally lazy/forgetful person in me would love to be able to mark a transaction where I want to make a return (e.g. Topshop) and it to remind me the date I need to take it back by…

Apologies if it has already been mentioned! :smiley: