How should Monzo get paid for marketplace products?


(Geoff Pascoe) #22

Plans change. Working in silicon valley I’ve seen people going from being totally anti-advertising to thinking auto-playing video ads with sound are a good idea. In Monzo’s case specifically, in 2016 they said:

over the last few years many banks seem to have concluded that their customers don’t understand percentages, and so instead they charge 50p or £1 per day for going overdrawn. You do the maths! I reckon that’s more like 200-400% in terms of interest. And I thought payday lending was supposed to be expensive…

Then less than a year later:

For the preview, the overdraft will cost 50p for each day your account is overdrawn.

Yes, it’s still a preview, but clearly their thinking has changed. This is why it’s important for the incentives of the company to align with those of users - that way, if people change their minds, it still makes sense to do what’s best for the users.

I guess the problem is that, as many sites that tried to do minimal, targeted advertising will attest (eg. Facebook), you tend to get the best returns simply by showing a ton of ads. Maybe this will be different with something like banking - I guess we’ll find out. Once again though, the issue is with alignment of incentives - if Monzo’s primary income is from advertising/partnerships, and they find out that it’s not bringing in enough income as they expected, then the incentive is to push it harder, and help the advertisers more.

In the end, this is all hypothetical though. If Monzo end up making a ton of money providing a small number of highly tailored offers, with the option of a global opt-out, then it’ll all probably turn out ok. If they don’t, then they’re going to look for ways to increase income, and that’s where a misalignment of incentives worries me.


(Alex Sherwood) #23

As you say, I’ll worry about this when it happens :slight_smile: of course Monzo’s plans will change but there’s no reason why that should be negative. It’s a possible outcome but I think it’s a waste of energy to consider hypotheticals, which are likely to be different to what actually happens anyway, now.

I don’t believe there is a misalignment if incentives -

Do you?


(Geoff Pascoe) #24

Honestly, I don’t know. In the case of many websites, yes. They’ve found that spamming people with ads barely decreases engagement (to a point), whilst significantly increasing revenue. Maybe with a more ‘serious’/higher value service like banking, peoples’ behaviour will be different, and the revenue-maximising option will actually be showing very few but very useful ads. Many companies thought this would be the case, but have been proven wrong. None of them were banks though.

It really comes down to how people respond and how they behave. If spammy/low quality ads turn them away in droves, then those ads will affect revenue. Hopefully with banking that will be the case.

Trying to predict the outcome of what is essentially a large social science experiment seems like a bad idea, but if people behave like they do on other services, then there could be an issue.

TL;DR Is there a misalignment of incentives? I don’t know. Is there a potential for one? Very much yes.


(Duncan) #25

You have interpreted what I said completely literally and as a result are somewhat missing the forest for the trees. The Google / Facebook case is analogous not because of the privacy aspect, but because a non-user entity is where they generate most of their revenue. Do you know the phrase “He who pays the piper calls the tune” ? That’s what we’re talking about here.

You may not be aware, but this is essentially a Loaded question. You’re basically saying If you disagree with me then you hate poor people. Well, no, I don’t. But I also think that Monzo having a viable and conflict-of-interest-free business model is of more importance than them trying to tackle systemic poverty. You have to fix the first thing first. Additionally, if the fee for using Monzo is reasonable, then there’s nothing to say that it would be unaffordable to even the poorest. But more importantly than that, your statement is a false dichotomy. It assumes that the only reason that some people go unbanked is for financial reasons. If you read your own link there you’ll see that much of the discussion is about people who are unable to prove their identity.

Taking a cut of a partner’s revenue as a percentage for the entirety of a user’s time with that partner is parasitic.


(Alex Sherwood) #26

It looks like we’re past the point where we’re going to have a rational discussion about the pros & cons of the various options. Hopefully other users will understand the points I’ve raised earlier so I’ll leave this here :slight_smile:


(Duncan) #27

It’s a shame that you cannot see the conflict of interest here. I worry that you are blinded by your love of Monzo. But it’s very simple, really. The interests of the user and the supplier don’t 100% overlap, and as the supplier will be the source of the revenue, their interests will come first.


(Phil Hewinson) #28

Thank you all for this great, thoughtful feedback! I can’t pretend to have the perfect model in mind and having considered all of your comments, there will likely be some compromises in any way we structure this. A purist approach would be to do this all for free, but as this is the foundation for our long term business model, we’ll have to make money with our marketplace and build a sustainable model.

We’ll try and choose the best approach to maximise user value and we’ll commit to being transparent throughout. Fundamentally we want to help people manage their finances by building a financial control centre. We hope we can eventually save the average Monzo customer in the order of £1,000 per year and if we take around 10% of the value we create in commissions somehow, that feels fair and should lead to a sustainable model - so directionally that’s where we want to get to.


(MikeF) #29

@philhewinson I’ve thought a bit about this so if you’ll allow me potifications…

As others have stated, conceptually you have a problem with both models presented here such that there’s always going to be something for a distrustful customer to point at.

Flate Rate Commission

To take the ‘worst case’ analysis and looking in the energy market for ease of presentation:

  1. Flat Commission incentivises the bank to switch every customer to any plan every year to maximise revenue regardless of customer benefit. (No incentive to allow customers to remain on existing plan if cheaper.)

  2. Lifetime revenue share allows the bank to switch every customer once then leave them to their own devices and keep taking a cut. (Reduced incentive to introduce customers to new plans.)

(please remember this is all theoretical at this point! I’m not making any accusations…)

Of the two, I can see a better justification for the second since there is nothing actively preventing the bank from recommending a new plan since revenue will be accrued from either remaining or switching under these circumstances. Given the normal practice of energy firms being a fixed price for a fixed period, however, that does leave Monzo getting paid even once the customer has reverted to a more expensive ‘standard tariff’.

The proposal for ‘limited time revenue share’ addresses the problem with option 2 but pushes us straight back into the problems associated with option 1 where it’s more profitable to make sure everyone switches every year.

Percentage Commission

Once we start taking about ‘percentage saving’ based commission figures, things start to get more complicated.

  1. Flat commission incentivises the bank to switch every customer to the lowest new plan every year to maximise revenue. (No incentive to allow customers to remain on existing plan if cheaper.)

  2. Lifetime revenue share incentivises the bank to keep every customer on cheapest plan for them and keep taking a cut. (Positive incentive to introduce customers to new plans.)

Clearly option 4 looks best here with the only possible issue being that if I decide there’s something I don’t like about the new recommendation and I decide not to switch, the bank still gets money even though I’ve taken my own decision and they’ve done nothing for me. Some may not be happy with that but I admit that I’m straining for a problem here.

The question of reversion to the ‘standard tariff’ at the end of a fixed term still remains, however. Why should the bank still be paid once the deal they recommended to me has expired?

The main problem I see here is one for the bank rather than the customer and it relates to how you measure a customer’s saving in order to base a revenue stream from it. I’ve never yet had an energy supplier correctly estimate my usage so, at the end of a fixed price term, they owe me in the region of £80 to £100. As a result, the amount I pay for energy each month is not a genuine measure of my costs. Clearly, it would be possible for me to tell you what tariff I’m on, but I don’t have to.

Presumably, for recommendations from the second year onwards, you’ll need to measure savings against the SVT I would have reverted to on the first year long contract you introduced me to?

Conclusions

I agree that compromises will be required, if only because I don’t think there is going to be a perfect model for this sort of thing. In hindsight, the energy market wasn’t the simplest thing for me to try anyway because of the time limited nature of the deals on offer.

I may have a rethink around annual car insurance, for instance, which would provide a simpler model to start from.


(Phil Hewinson) #30

Thanks for your thoughful comments @Feathers! I think aiming to make a % on the savings a customer makes is a useful directional goal, but not necessarily something that will land cleanly into how we structure these deals. My impression is that we will probably structure things differently based on the product as each area is different, so we’ll probably approach this case-by-case, but with the above directional goal in mind. Even within a product area, we could structure it differently too - for example, we might setup a model closer to a life-time rev-share for a single tariff energy supplier, but have a fixed-fee or one-year rev-share deal for a supplier that reverts to an SVR after a year, which would align us better to the savings a customer makes.


(Alex Hansford) #31

Really interesting thoughts from @Feathers:

I personally like the limited time revenue share - limited to 12-18 months.
If we’re purists about it, we would say that this encourages plan changes after this point - however, it’s reasonable to suggest that if the deal they are currently on remains good - there’s no need to change. Monzo aren’t judging (or chasing revenue) here, just providing options.

The other issue is that most marketing effectiveness is either:

  • acquisition focused (new customers, clearly the primary benefit of the marketplace)
  • loyalty (keeping existing customers, they don’t need the marketplace for this at all)
  • recovery/re-engagement (turning stale customers into active ones - not really relevant to the marketplace, but possibly the feed?)

Clearly the marketplace is about surfacing new integrations that offer value to customers - so it’s about acquisition ultimately.

@philhewinson - I’m interested to know how marketplace will work for services that are less focused on generating revenue? i.e. banking information and PSD2 data freedom/management requests?

Clearly if we’re not too careful, the marketplace will only focus on profitable acquisition verticals such as insurance, etc. Although perhaps that’s the aim?

Could other worthy services can just use the API for that purpose fee-free, would that work?
Looking forward to your Open Office Q&A at the end of the month to hear more! :raising_hand_man:


(Justin) #32

The ideal model would be for Monzo to take a percentage commission for a year of the money saved over what the customer would have paid had they stayed with their previous supplier. This would incentivise Monzo to continually shop around for the best deals for its customers, and also give it the opportunity to make some revenue. The percentage should be a decent chunk - e.g. 20-25%. However, a significant challenge to this would be that you’d need to find a good transparent mathematical model and some good data scientists to make this effective.


(Hedley) #33

I’ve been working on a startup that helps people find better deals for their household services on an ongoing basis, and as a very happy long-time Monzo user I thought I’d add a few comments to this very interesting discussion:

  • There are quite a few changes coming up in the future which will likely change the competitive landscape. Namely smart meters giving comparison services instant-access to users usage information, the centralised switching service (CSS) enabling next-day switches and much more reliable switching, and also the number of startups entering the market that will be helping people switch to cheaper deals for energy & other household services.
  • These changes may mean that finding a model which prioritises maximising the savings and number of suppliers available to a customer becomes even more important, due to competition. Currently what’s happening with price comparison websites is that some of the very best deals are not available due to high commission costs.

I think a percentage based model could definitely be a step forward in general, but I would also add that making sure whatever deals you strike don’t exclude too many suppliers is important for the customer.

A few other approaches I’ve seen tried out include:

  • Not accepting any commission fees from suppliers (in order to remain entirely impartial), but taking ‘a small percentage of the savings we make you’ from the customer
  • Charging £25 per year, but guaranteeing a saving of at least £50

Great to see Monzo (as alway) being so transparent about this!


#34

This makes sense. No lump-sum commission. Instead a percentage based on saving. The greater the saving for the customer then the greater the percentage for Monzo for the recommendation.

On the face of it, that appears to be the best way to do it. Is there a catch?


(Hedley) #35

Yeah I think it is a great approach in theory, two downsides / challenges might be:

  1. Marketing - I haven’t had direct experience with this proposition, however I do know that money is a big driver and it’ll be hard to avoid some people viewing the % that Monzo keeps being ‘taken away’ from them. Also this could create competition with competitors, where they drop the % they ‘charge’. If the revenue is ‘behind the scenes’ then this issue goes away - a major reason why comparison sites do this now, rather than ‘charge’ the customer in some way and do whole-of-market, even though it would be in the customer’s best interests - people are irrational and it’s ultimately more profitable that way.
  2. Economics - the long-term sustainability of this approach is a tricky one. Year 1 is fairly simple, you’ll probably be saving the customer £200-£300 for energy, take 10% and that’s fine. Now they’re on a good deal though it’s unlikely they’ll save more than £100 year 2, if prices are rising they might even pay more. Taking a % of the savings based on their original deal forever would get around this, however it might be a hard sell.

And a smaller issue is that it could incentivise Monzo to switch customers to the very cheapest supplier to maximise their revenue - even if that supplier is rubbish. For example, right now IRESA energy will save you about £350 for an average user, but they’re terrible. Bulb might save you £250 but they’re way better and worth it though. Although I think competition in the utility-switching space should prevent this - given that Monzo are a decent company that wouldn’t want this to happen to their customers of course :slight_smile:


#36

I used to work at an online lender which acquire their customers via an online marketplace (e.g. money super market etc), so maybe I can give you an insight into how those ‘price comparison’ website works and how do companies ‘buy’ customer from price comparison websites, perhaps that can give you slightly different perspective?

There are 2 types of market places that my previous companies used to acquire customers from, and their business models are quite different, but I’ll briefly summarise how they work and how my previous company acquire customers via the 2 different types of market places.

Type 1: ‘Lead providers’.
Customers would go on to these Lead Providers to apply for a loan, then these lead providers will ‘auction’ customers to all their clients. For example, John Smith goes on to leadprovider.co.uk (fake website) and applies for a £200 loan, so based on the ‘credit score’ of the customer (which determines how risky the customer is), the lead provider will then auction this customer to all its client and sell to the client that offers the most money.

So for example, my previous company would see ‘John Smith’ and being quoted for £50. Then based on the customer’s credit score + our own internal statistical models, we would predict the lifetime value of the customer (e.g. likelihood of repaying, likelihood to re-borrowing etc), then decide if £50 was worth it. If we determine that in 12 months time, the customer will only make a revenue of only £30, then we would say no to £50. The lead provider will then ask other clients if they would like to pay for £50, and if everyone says no, then the lead provider will drop the customer to £40 etc, and keep dropping the price until someone’s statistical model says that the customer will be profitable etc.

If we first encountered John Smith via this lead provider, then in 6 months time approached by the same lead provider for the same customer, then we would always say no, and email John Smith to encourage the customer to apply directly via our website (to avoid paying for the same customer twice).

Pros and Cons for this model for Monzo:
Pros: Monzo could shortlist a select companies that Monzo deems ethical/ good enough to sell customers to, to protect the well-being of Monzo’s customer, and would only approach clients only if the customer explicitly consents (customer says they would like to apply for a loan or switch energy provider or credit card etc etc), and with an appropriate pricing model, Monzo would almost always be able to pair a customer with 1 shortlisted product/ company for the ‘highest price’ possible. This would ensure Customer always walk away with a product they were looking for.

Cons: Customer has no say over which product to pick - which product customer ends up walking away with depends on which company is willing to pay at the price Monzo is offering at. Of course customer can always say no to the company they are paired with and ‘re-apply’ via Monzo again hoping to get a ‘preferred product’, but this doesn’t provide the best customer experience and choices. --> Not the best option for customers.

From Monzo’s profitability perspective, Monzo cannot monetise from the same customer once per company as no company would want to pay to acquire the same customer twice, and this would restrict the future revenue stream for Monzo.

Disclaimer : my previous company was a payday lender and this model actually worked quite well in payday world because many of payday customers cannot get credit anywhere else and choice isn’t necessarily the most important thing to them. These customers are desperate for loans and they prize ‘ability to get a loan’ far higher than the cost of loan/ customer service etc. Despite being a very efficient model and ensuring customers looking for a product will almost always walk away with a product, I don’t think this model in my opinion will not work well for Monzo due to a very different customer demographic._

Type 2: ‘Price comparison websites’

This is perhaps a channel I think many people are more familiar with. Examples are Clearscore, Noddle, MoneySuperMarket etc. They essentially work like a search engine for loans. Users will provide their details, then these loan matchers will pull their credit scores, then match them with the providers with a ‘likelihood of being accepted’. Customers then have the ability to choose which product to apply for as all the information is being presented to them (likelihood of being accepted, APR charged etc). The power of choice now lies with the consumer.

My previous company would pay these companies a ‘sunk cost’ which is often quite high for a specific spot in their rankings. For example, if John Smith goes on to PriceComparison.com (fictional) to apply for a loan, PriceComparison will rank 10 loan providers, sorted by either ‘cost’ or ‘likelihood of being accepted’, depending on the customer’s preference.

My previous company (again hypothetical case) pays PriceComparison £50K a month in order to appear in rank #3 for example (third link from the top), regardless of how many customers gets converted. The way we calculate the ‘acquisition cost per customer’ is then £50K divided by number of customers acquired a month via PriceComparison.com. For example, if we manage to get 1000 customers via PriceComparison.com, then our acquisition cost would be £50 (before factoring credit losses + servicing costs). We would then have a statistical model to then calculate what is the optimum price to pay to appear in a given ranking. For example, we would have calculated that it costs £100K to appear in the top spot, but that would only generate 1500 loans (meaning acquisition cost per customer is 100K / 1500 = 66.6), then it’s actually cheaper per loan if we pay £50K to appear on the third link than £100K to appear on the top link.

Again, pros and cons for Monzo:
Pros: Customers will have the power to make informative choice. Instead of being ‘incentivised to convert’, Monzo will receive a fixed income for having a product listed via it’s platform, much like how Google search engine makes money.

Furthermore, many websites like ClearScore/ MoneySuperMarket etc price customers solely via their credit bureau data (and sometimes also their website cookies). However what I think Monzo has to their advantage is the customers’ Payment behaviour and data. I believe that customer’s spending behaviour is very strongly correlated by their credit score and posses hidden data not available in credit scores, and this data would allow Monzo to build a much stronger credit score to determine the credit-worthy versus less credit-worthy customers, then match them with appropriate products based on their risk.

For example, RateSetter would advertise their product on Monzo and MoneySuperMarket for example. But because Monzo was able to screen customers better than MoneySuperMarket, Ratesetter would find that despite paying £66 per customer via Monzo vs £50 per customer via MoneySuperMarket, Monzo customers have 5% default rate vs 20% default rate from MoneySuperMarket, and they actually make more profit per customer from Monzo customers. Monzo therefore would also have the power to charge higher listing fees than its competitors etc and lenders would still choose to list their products on Monzo etc.

Cons: : Ability for Monzo to monetise their customers these way really depends on how likely Monzo customers actually want to purchase a product etc. Also, customers may learn of a product via Monzo Marketplace, but eventually apply not via Monzo marketplace but via the lender’s website on their desktop directly. So from lender’s perspective, they will not be able to see that the customer originated via Monzo and would say to Monzo - why am I paying so much to list my products on your website when I’m not even generating any sales via you?

For a traditional price comparison websites (which most traffic predominately happens via desktop/laptop), everything is tracked via cookies so it’s easy to identify when the customer first sees the product. However, it’s quite easy for customers to view the product via Monzo, then switch to their laptop to actually apply for something and someone else basically takes the credit for Monzo’s advertising effort.

Monzo will have to incentivise customers from switching via Monzo rather than other platform in order to boost ‘sales/ volumes’, in order to persuade third-party providers to continue paying Monzo to list their products on Monzo.

And this brings us to a chicken-and-egg problem - third-party service providers will only want to pay Monzo if it’s worth the money, but with no third-party service providers listing their products on Monzo, Monzo consumers will not choose Monzo as their market place as there are already existing websites to get what they are looking for. Monzo will have to put a lot of effort to encourage third-party service providers and consumers to use Monzo marketplace to build the business case.

I think from a customer’s perspective, I would prefer the ‘price-comparison’ website model, which has worked very well for ClearScore and ClearScore’s customers. Personally I would very much prefer Monzo to shortlist approved companies, then present all the information in a clear and transparent manner to me so that I can decide which product best suits my needs, and apply for it via Monzo with only a few clicks. Monzo also is not incentivised to ‘sell’ me because they receive a sunk cost for each month. Monzo can then increase the revenue they earn by providing the platform with the highest conversion rate to increase their ‘listing fees’ etc.

Edit: I’m not sure if Clearscore uses a bidding system actually. The PCW (price-comparison website) my previous company used has a bidding system, and we basically pay them a ‘sunk-cost’ every month regardless of number of converted customers. I think some PCWs charge about £50 per conversion…


(Mikey Deegan) #37

Just an idea, but what if Monzo charged the company a flat ongoing fee to be included in the Monzo marketplace. That way there is no incentive for Monzo to recommend anything other than the best option for the customer.

It would also mean that only companies that were confident in their offering and really added value or saved the consumer money would sign up to the marketplace and be wiling to pay to be in it.

With this model, Monzo is totally objective, assuming all partner companies pay the same flat fee.


#38

I suppose what I was trying to get at in my previous post was that Monzo shouldn’t be incentivised to ‘convert’ or ‘sell’ the customer, but should do it ‘altruistically’ by shortlist trusted companies that they think would benefit Monzo customers the most, then present them to the Monzo customers when Monzo customers ask to be presented.

Flat fee would work, but I don’t think that would generate the most profit for Monzo - links that are presented right at the top often has the best ‘conversion’ rate because people tend to click on the first few links they see. To get the most profit out for Monzo, it would make sense for Monzo to auction the slots to the third-service providers - auction is the most efficient automated pricing mechanism. The cost is is borne by third-party provider as an ‘advertising’ cost and they will compete amongst each other blindly, and will be allocated a spot where they are most willing and able to pay for, much like search-engine optimisation advertising. All this will come at 0 cost to consumer.

But of course to present Monzo customers the fairest product, ranking should be done for equivalent products only and Monzo shouldn’t allow unsuitable products to appear right at the top. For example, if I specify I want low interest purchase credit cards, Monzo should then shortlist all the low-interest credit cards and order them by APR cost for me with the lowest at the top.

If the 3 shortlisted cards are Virgin Money, Lloyds and Halifax, and all 3 cards have the exact terms and conditions, and these 3 companies will fight amongst themselves - the company offers the highest advertising fee will appear right at the top. For example, Virgin offered Monzo 50K, Lloyds 30K and Halifax 25K a month, then from my end I’ll see Virgin at the first link and Halifax at the third link.

Barclays could offer me a 5% APR card and even if Barclays offered Monzo 100K, Barclays should still appear after Halifax, because I wanted Monzo to show me the cheapest cards first (something I tell Monzo).

However if I cannot be offered 0% interest and the best card I can get is 5%, then if Barclays pays the highest fees to Monzo then Barclays will of course appear on the top link.

The reason I think this works better for Monzo than a flat fee is because the order a link is present matters greatly in conversion, and by having appropriate mechanism that prices in the conversion rate automatically, Monzo can maximise the profit it generates. And as a consumer, I would love Monzo to be profitable or at least breakeven because this is important if Monzo wants to continue operating for a very long time. And if Monzo succeeds at proving profitability, then I as a consumer also benefit from having a wonderful product!


(Benjamin Paul Heap) #39

I’m not sure if you’re still working on this, but I submit my meter reading data to Bulb every month. It would be great if I could record it in Monzo, who could match it with what I’m paying and then see if the market could offer me anything cheaper.

i’m sure you’d be able to make it into a nice graph that I could show to the other half to convince her to turn the heating down too.

Thanks

Ben