The worst kept secret in the financial services sector is how much Google wants to get into banking.
Of course, it’s a bit of an old secret too. Google is technically already in banking. They own half of Lending Club, a P2P consumer lender.
The reason bankers are scared of Google getting into banking outright is that Google is so much better at data analysis than any bank. Their ability to collect everyday pieces of data and turn them into revenue is still amazing – even by today’s standards.
But, as I point out in my book, Seven Billion Banks, in order to compete in the future, Google will have to go mobile with their banking services.
Right now, Google’s Android devices (can) use a combination of GPS, IP addresses from wireless networks nearby, and the nearby mobile network to help them pinpoint your location. So they have the basics of a location-based platform for banking.
But what’s missing? I think 3 things they will have to nail perfectly in order to be a formidable banking foe.
1. Get the Location Right
Google does a pretty good job of knowing where you are. But they’re not perfect. If the IP address of a wifi network is incorrectly linked to a geographic location, Google will “find” you in the wrong place. Sometimes, several states away. If you’re on or near a plane or cruise ship, forget it. They will place you somewhere across the globe.
The theory should be that there is enough information for Google to get this right, but as of today, they don’t. If you’re on a cruise ship in port in the Caribbean, even though there’s a mobile network that your phone may be connected to, Google may still take the IP address of the ship’s wifi and place you in Florida.
Why does this matter?
Because location-based banking depends heavily on the context of … your location. Get that wrong and everything else is wrong too – or worse, irrelevant. That may be more than an inconvenience. If your bank thinks you’re stateside while you’re really international, that could have compliance impact depending on what you want to do – like a wire transfer.
Of course, the even just the inconvenience is everything, because context will make or break a location-based virtual bank.
Prognosis: Positive. Google will figure this out. Right now, they don’t seem to care about this problem, but they have the technology. When it starts to impact revenue for a banking app, they’ll put engineers on it and fix it.
2. Get the Context Right
More than location, context matters. Google will have to take the data it knows about your purchasing habits and turn them into real recommendations that are valuable to their users. Google does a decent job of context information today. Google’s Google Now cards try to be contextual and helpful. Sometimes they’re creepy how on target they are. But sometimes they’re irrelevant. For information, irrelevancy is just annoying. But when you try to help people make better financial decisions irrelevant can be deadly (figuratively speaking, of course.)
But a banking application needs more than information. Just look at how badly PFM applications do in the marketplace when they just provide information, but don’t add value. PFM needs to help people make decisions, and Google will have to as well.
Google is very good at giving you relevant information, but so far it hasn’t had to help you make a decision – like when to finance something or when to move money around.
They seem to have contextual triggers down; but do they have next best action? That’s a critical pillar to a solid mobile banking strategy.
That means more than just real-time information, Google will have to do some optimization – decisioning. They’ll need to know what defines value for each customer individually and then optimize around it to increase not just their objective, but their customer’s as well.
Prognosis: Tentatively Positive. This isn’t Google’s current strong suit, but it could become their strong suit. The one big area where they do optimization is with ad placement. They optimization ad placement based on revenue for them (and for the advertiser). This is a real-time optimization that hones in over hundreds and thousands of impressions. When you’re dealing with that kind of scale, you can use a more real-time learning kind of approach. That’s not true with individual customers. You test poorly on one and you better know what you’re leaving on the table.
3. Avoid the “Ikk” Factor
I have to give credit where credit is due here. I can’t remember who said it, but at a conference I attended, a speaker defined the “Ikk” factor as the use of data in a way that makes the consumer feel sick that a company has used their data that way.
Of course, the difference between “ikk” and “ahh” is one word: relevance.
Study after study tells us that customers don’t mind when companies use their data – if they use it in a way that adds value to the customer’s life. When it’s relevant, people get over the “how did they know that about me” moment and appreciate the help. When it’s not relevant, then people feel sick that they’re being monitored and tracked.
The answer lies in models that predict what people want. An unlike ads that can simply be ignored when irrelevant, banking data is highly sensitive. Ask banks that execute a direct marketing campaign poorly. They’ll tell you that customers get tired of irrelevant offers. Not only do they see low response rates, but they find that customer satisfaction drops as customers begin to wonder, “do they even know me at all?”
Google, honestly, is just figuring this out. Most ads they run are highly irrelevant. Adwords and their display network ad placement gain very low response rates. Much lower than bank direct marketing.
Imagine if your mobile banking application interrupted you with messages, ads, or recommendations that you were only likely to act on – say 0.02% of the time (2 bps)? You’d definitely uninstall it.
Prognosis: Neutral. We’ll see. I think Google hasn’t done a great job with this in the past and they don’t yet show that they “get it.” Their business model right now depends on advertisers placing all kinds of offers and self-selecting where they display. Obviously, they can’t do manual quality control with that type of scale. But a mobile banking app can’t just be another display network channel for them. I’m not sure they understand that.
Prognosis for Google Mobile Banking
Net result: Tentatively Positive. All in all, it’s just a matter of time. I think everyone knows that. The Google Wallet app already has the makings of a mobile banking application. They just need the data. Some of which they’re collecting through their investments and some of which will need to wait until they enter banking formally.
Watch for Google to make significant investments in traditional optimization algorithms above and beyond what they do with ad placement today. As they gain experience with their Google Now cards, they’ll learn what context matters and what doesn’t. Where they will need to close the loop is in understanding how people react to that information: what is the behavior?
I think if they see banking as more than just another ad channel, they’ll be a formidable opponent. To do that, they’ll need to provide assistance in making decisions in a way that avoids the “ikk” factor.
If not, it’ll just be another data collection source and will never rise above being popular among Google-philes.