After spending the better part of three years working in ad tech, the term “sticky network effects” is simply part of my vocabulary.  For those who aren’t as familiar, network effects occur when the value proposition of a product or service gets stronger as the number of people who use it increase.

The classic example is the telephone.  If one person has a telephone and they are the only person in the world that has one, the telephone is useless.  The value proposition of the telephone is not in the device itself, but rather it’s in the ability to connect with other people who also have a telephone.  The more people who have telephones, the more valuable the telephone becomes.  The same principle applies to all sorts of things we use every day: fax machines, the Internet, Bloomberg terminals, Facebook, real-time bidded advertising, etc.

Network effects can be a very powerful tool for building a scalable business.  Because they can be so profitable, it’s logical for companies to use incentives to build their user base and activate their network effect.  However, companies should be very careful how they use incentives to build their user base.  Using the wrong kind of incentives could alienate and infuriate the very customers they want to attract.

Let me give a real world example.

I use Bank of America for my checking account.  They’re ok.  What I like about them is that they have branches in Baltimore and New York and they have one of those mobile apps that allow you to deposit checks remotely.  I’ve used them for a long time and don’t really want to switch banks.

A few weeks back, I was in need of a simple banking service.  I needed to change five twenty-dollar bills for a crisp $100 bill (it was the annual holiday bonus for the super intendant of my apartment building; in my experience, nothing says “thank you” to a super like a crisp $100 bill).

Unfortunately, there isn’t a Bank of America close to my apartment, so I just walked into the nearest bank I could find.  The bank I visited was a brand new bank that I had never been to before called “Santander.”  When I walked in, the bank was clean and empty.  I walked up to the teller and politely asked if they could change my 20s for a 100.  To my surprise the bank teller told me that they would only change currency if I had a Santander account – and then, she asked me if I wanted to open one.

I was shocked!  I understand that Santander wants to get more people to sign up for their banking services, because the more banking customers they have, the better chances they’ll have at creating a sticky network effect for their retail bank.  I would have totally understood if Santander offered me a positive incentive to open an account – for example, a $100 account bonus, a referral program, or something else like that.  Those positive incentives are pretty standard and I often get mailed solicitations with those kinds of incentives.  However, Santader was taking a different approach.  Rather than providing a positive incentive to open a bank account, they were using negative incentives.  They were depriving me of a money-changing service unless I opened a bank account.

I was very frustrated.

Because Santander refused to change my money, I had to walk 10 blocks to the nearest Bank of America – where they changed my money without even asking if I had an account.

For me, the moral of the story is that it’s always ok to give customers a positive incentive to participate in your network effects, but it’s much riskier to use a negative incentive.  It can be very frustrating for a customer to be denied a service and, chances are, it’ll just leave them upset and heading straight into the arms of your competitor.

Sticky Network Effects and Influencing the Behavior of Your Customers
Tagged on:         
  • Minor note: I’m not really sure how you’re distinguishing a “sticky network effect” from a “network effect.”
    That said: yeah, your example is pretty bad of Santander.
    To me, it screams of the fallacy of governing-by-averages. It sounds like some MBA-type took a look at the “average cost” for, e.g. a bank teller to serve someone, and then measured the percentage of non-customer interactions, and figured they could save 5% – 10%.*
    Probably not true, in the end (also: devilishly hard to measure, so they executive probably ended up saying something like – “I saved $XXX dollars by reducing the amount of wasted time our tellers spend”, and got promoted/moved on).
    * That kind of thinking also results in things like: “Our average costs for tellers are X, the number of account-holding customers we have is Y, so therefore we can calculate the value of our teller service per account as X/Y.” Which leads to silly conclusions and mental gymnastics, like saying “We know our account holders currently place a value of $Y on these services. We’ve managed to cut our costs on delivering that service by 12%, so we are now more efficient – delivering $Y of value at 88% of the cost.” And then that number becomes inscribed in the institutional memory and thrown around to justify different decisions.
    [Damn, I’m cynical tonight]