Please stop calling them dumb pipes

Posted by Aden Davies on March 21, 2012 | 14 Comments

Lots of people recently seem to be warning about banks becoming dumb pipes. They say banks are destined to just become the wires. The hearts and minds of customers will be won by the masters of the web. The Googles and Amazons and Apples and Paypals of the web 2.0 world. I agree they probably will but is it really a problem?

Those web 2.0 darlings are not going to make themselves into bank. The majority of them are just interested in the transaction data they don’t want the hassle of running a bank. Basel 3, MiFID and other impenetrable forms of regulation might not be too appealing.

Some may say (not me of course as I work for one) Banks have proved they don’t really get this web thing and especially not this web 2.0 thing with its rounded corners and nice fonts and helpful intuitive interfaces. Why not let the experts have a go at that bit while banks stick to what they are good at.

The banks operate a huge complex global network that moves trillions of dollars per day, usually without much issue. Complex fraud and anti-tax evasion systems operate silently. Audit requirements, data protection standards and a myriad of regulations make this system the powerful beast it is and also a potentially irreplaceable one.

No one in Silicon Valley or any other entrepreneur saturated dreamland is going to want to recreate the whole bank system (I have visions of mad stock sale billionaires from Facebook sitting in their volcano housed lairs thinking ‘we should do that’). That bank system may be a bit long in the tooth and may need some updates here and there but could we give it a chance to catch up by laying down some of these so called dumb pipes and bringing it closer to that other huge complex global network called the Internet? If we give a few more people access to the system in a web friendly way will it be of benefit to all? Will people realise the power of this network and what it allows us to do today?

Liz Lumley wrote a great post about SXSWi and how all the cool companies trying to disrupt banking are massively reliant on that network. ‘What struck me was the juxtaposition of the bravado coupled with the fairly shocking display of ignorance on how international banking and payments happen.’  These pipes are never going to be simple or dumb.

That being said I am interested in the most simple of these dumb pipes. I want an automated data feed from a current account. Every time a new transaction occurs I want a data feed I can subscribe to, just like RSS, to update. That feels very simple and you could say dumb but to make that happen is going to take some damn smart coding and some bravery.

The big problem is authentication. How do I prove I am who I say am? How do I prove that I am allowed to subscribe to that data feed? How does that authentication model satisfy banks security and fraud departments? How does it satisfy the regulators?  What would happen if someone had access to all the data behind that API? What if the Daily Mail had access?

The most simple implementation of the so called dumb pipe was planted in my head by Dave Birch. He posted the following tweet.

Setup a private twitter account. Plug it into your bank account (this dumb pipe of course has OAuth/XAuth like qualities). Follow it to catch your transactions as they happen. Now a bank would never build this. No revenue at all. It only presents risk but a customer has asked for it (albeit a quite forward thinking one who would be a guinea pig to embed payments chip into his body) but a customer need is a customer need and we know they are always right.

A few examples i have seen during my time at HSBC where customers are trying to circumvent this lack of a subscription data feed. Designer Aral Balkan was none too happy that he could only manually access transaction data from two months in the past. So he built a tool (in under 4 days) to scrape the data and save it in a format for him to upload to Freeagent. Another person, a smart gentleman going by the name Jay Fresh, went a step further. He reverse engineered online banking to produce a command line interface. I spoke with him and asked him why, his reply was that he had simply wanted to build his own iPhone app. I can understand his frustrations. Should customers have to work so hard to do this? Should they have to risk their own logon data and potentially break terms and conditions to try and get to the data? Banks spend pots of cash each year trying to figure out what customers want, why not give them the tools to build what they want. A so called dumb pipe would be a very powerful tool in the right hands.

Mr Bank 2.0 (soon to be 2.1 and available in all good book stores), Brett King, also wrote a great post on this topic (and has been talking about it for years) arguing that if the banks do become merely an infrastructure layer then they will miss out on the value built on top of it and that we may need fewer banks/infrastructure providers. I agree they might and there could be less banks but do we need that infrastructure layer to be created to allow new value chains (ugh) and innovations to truly flourish? Where would we be if we still had a fragmented electricity system? Or you could only call someone on the same telephone network as you? We need to create these commodotised infrastructural layers and allow them to weave into the wider world (web?). The innovation S-Curves of many technologies have shown this pattern. Banks may resist as the wireless telcos are doing now, except for the smart ones such as Telefonica, but I believe there is an inevitability and the banks that embrace this will be the ones that exist…but I digress.

So what would a banks dumb pipe look like? What are the technologies required to keep this mother of all honey pots safe and secure so it does not spring a sticky leak. What would be needed to build the simple sounding dumb pipe detailed above? Yes there is inherent risk on freeing customer transaction data but I think the potential benefits outweigh the risks (I may be alone on this). We are starting to see some things in the French banking market that might answer these questions SDK’s have recently been released by Crédit Agricole and this week has also seen the launch of an API by Banque AXA. The future looks French.

I look forward to the arrival of the dumb pipe. It will bring together the banking system and the web. I have high hopes for this dumb pipe. People need to realise that the pipe is not so dumb.

Posted in APIs, Banking, Ideas, Ignorance, Innovation, Ramblings |

  • http://www.redmonk.com/jgovernor monkchips

    solid post, though i feel you could be a little more hard on banking. you talk as if it wasn’t a bit of a mess, but it surely is.

    what would a dumb pipe look like – a good bank, with no investment banking arm perhaps? Where did those regulations actually get? Some organisations complied, while others didn’t – but they all still ran up debts they couldn’t pay. 

  • http://twitter.com/theodemontford Theogenes Montford

    I rather think a smart pipe (ie, dumb pipe + API + strong authentication) is actually quite an appealing model for retail banks: let someone else build the tools that consumers want while the bank focuses on core business, which is savings and loans, not payments!

  • http://twitter.com/Setok Kristoffer Lawson

    I think the term ‘dumb pipe’ can be misleading in the sense that ‘dumb’ sounds negative. Actually the fact that banks have their existing networks is great — but what most of them will be less able to do is to build up the innovative services on top. Much like you wouldn’t expect operators to build all those Internet services out there (they did try, though).

    The way to look at it is that those banks who are best at becoming those ‘dumb pipes’, with good APIs and a willingness to partner with startups, are the banks that will benefit most from these changes.

    The next question is then what to call these new services and startups. For many users we at Holvi are essentially a bank: we store their money, they can pay bills, they can move money. Just with extra functionality and UI. We utilise existing banking partnerships to make the job of this easier for us (although are regulated), but we are also something new on top. Maybe Banking-as-a-Service?

  • http://twitter.com/bernardlunn Bernard Lunn

    Interesting post. I agree that the tech ventures will not want to become banks, nor will they be able to. But I do think that a lot of what currently look like the cash cows of banking could be taken over by automated solutions that dramatically undercut the high cost banking equivalent eg in payments, in index funds, in P2P lending.

  • Pingback: » Please stop calling them dumb pipes Aden Davies « smartfinserv

  • http://www.adendavies.com Aden Davies

    I totally agree but is there a mindset that you are giving over control? Precious data that you are sure there is value in? 

  • http://www.adendavies.com Aden Davies

    There are certainly some low hanging fruit that lower margin and smarter data use models, surely exist for. 

  • http://www.adendavies.com Aden Davies

    My point exactly around the term dumb pipe. they are not dumb. Like you say The bank with the best pipes, the one who most embraces the web and the innovative potential will surely be successful. 

    As for what to call these new breed of ‘banks’ like yourself and BankSimple I actually had a few tweets with Ron Shevlin last month about this. I jokingly said Bank Front End 2.0 https://twitter.com/#!/aden_76/statuses/174247320102248449 But I think you are so much more. Banks need you guys to show how the interface should work. I think BaaS is what you guys build on top of. What would be interesting is to hear from you guys and othes about what you need from these BaaS providers. What do you want from utility banks? 

  • http://www.adendavies.com Aden Davies

    Well obviously I work, and would like to continue working for, a bank. Yes the systems are creaking in places, yes they need some major updates for the brave new world but they are massive and complex and that won’t be cheap or easy. I think you can extend the life of these systems by connecting them to the web. This not only allows for further innovation but also highlights the weaknesses in the system and should allow the major pain points to be targeted. 

  • http://people.scred.com/~setok/ Setok

    For us the requirements are fairly straightforwards:

    * We want to own the customer relationship. We currently do our own KYC for example, but it is crucial to provide a seamless signup and user experience.

    * Thus we want the ability to open up new accounts transparently, and have API access to those accounts to manage them.

    * SEPA payments (eurozone) and pound sterling.

    * If KYC can easily be offloaded to utility banks in a transparent manner, via API, that is also an option. As long as the experience is seamless and they are acting within the Holvi environment.

    I fully agree that we are much more than a frontend. In our case we e.g. add a rich set of metadata to transactions, bring together a team around an account, do full double-entry bookkeeping in the background, and offer merchant abilities (phew). It’s value-added banking.

    The name ‘Holvi’ actually means a ‘vault’, so maybe Simple and us are vaults…

  • http://www.adendavies.com Aden Davies

    An API/seemless process for KYC and account opening is going to be a tough one to crack. ID requirements alone make that really tricky. It will be great if a bank does make it happen though…then we will see a whole new breed of banks. So do you offer an API out from your system? If you are adding data can I get that out and pass it into another layer? 

  • http://people.scred.com/~setok/ Setok

    I agree, it requires a bit more effort, but there’s no technical reason why it cannot be done. We are still happy if we do the KYC (as we’re doing it now anyway), as long as we can create those accounts on utility banks easily — that is important.

    We are actually going to be in London next week for Seedhack, where we’ll showing off some early versions of an API. So yes, the idea is that data can be pulled out, and further things can be built on top.

  • http://twitter.com/brettking Brett King

    The pipes might be dumb, but the intelligence will be utilizing the pipes and that will require two willing players committed to some pretty smart integration, process and business rules.

    Mr. Bank 2.x

  • Pingback: Disruptive start-ups and banks – putting the ‘me’ in disintermediate* - NixonMcInnes