Farewell, Pay By Touch, Farewell

Pay By Touch has come to the end of the road. In a press release issued today, the company said it was ceasing all remaining biometric operations:

Solidus Networks, Inc., DBA Pay By Touch, regretfully announced today that it will no longer process biometric transactions on behalf of its merchant customers and consumer membership base, as of 11:59:59 pm March 19, 2008.

Alas, this is no surprise. The company was overextended, and bankruptcy is a terrible position for a start-up trying to right the ship. The employees and new management brought in gave it a valiant effort.

A few post-mortem observations about the company are in order.

Consumer Adoption of Biometrics

It’s tempting to consider biometrics a loser in the consumer space. Ben Worthen at the Wall Street Journal does a nice job talking about biometrics’ value as well as its shortcomings. Biometrics is very sci-fi. And yes, it is too spooky for a large percentage of people. Inside Pay By Touch, we found there were two types of people who wouldn’t sign up: those who were concerned about privacy, and those who were against it on religious grounds (mark of the beast). If you follow Crossing the Chasm theory, there was also the mainstream part of population that would wait to adopt.

But there were early adopters. Pay By Touch processed several hundred thousand biometric transactions per month. Now in the grand scheme of things, those numbers pale in comparison to the volumes processed by Visa, MasterCard, American Express and the overall number of transactions occurring at grocery stores. But processing several million transactions over the course of the year is nothing to sneeze at. It shows that consumers were indeed willing to use biometrics for adoption.

I do think the adoption curve for biometrics is longer than for, say Twitter or RFID-enabled credit cards.

Economics of Biometrics

Biometrics requires both hardware and software. The hardware is pricey, and it does need to be replaced periodically. One thing about the field is that vendors continually innovate. Hardware gets more durable and reliable, and prices do come down. I had the chance to test different vendors’ new biometric readers. A lot of effort is being put into biometrics out there. I remember one of my favorites during testing was actually from Casio. It performed well, and brought back childhood memories of my wristwatches.

But the cost of installing and replacing the biometric readers is an issue.

Installing in the grocery lanes also took money. Each lane in each store had to be outfitted. Pay By Touch had to be ready for the different POS versions the stores were running: IBM, NCR or Retalix. Always with some different configuration or customization.

Pay By Touch’s biggest play for grocers was shifting consumers to ACH. ACH is much cheaper than PIN debit, signature debit or credit cards. So, of course, Pay By Touch couldn’t charge too much per transaction. This meant that operational costs needed to be kept low, the company kept small and lean. This didn’t happen, of course.

Level of Certainty in Authentication

Pay By Touch was very focused on a high degree of confidence in its authentication of each consumer, each transaction. There are two measures for how the biometric authentication performed: false rejects and false accepts. False rejects were cases where the consumer was legitimately trying to authenticate, but the system rejected him. False accepts were cases where the system wrongly identified a consumer as someone else.

The company’s bias was to avoid false accepts. One false accept could become a major news story undermining confidence of consumers in using Pay By Touch. And the system was quite good at avoiding false accepts.

The focus on avoiding misidentification meant that false rejects ran higher. False rejects were more acceptable - no one’s checking account would be wrongly debited. Early on, the company had a bad problem with false rejects. But a lot of work reduced that number significantly to a barely noticeable level. And that was important. Too many false rejects also undermine consumers’ confidence in the system: “Oh that fingerprint payment system never works.”

A lot of time was spent improving biometric authentication at the in-store hardware level, in-store software level and the hosted server level.

What Would Have Been Better for Pay By Touch

Hindsight is always 20/20. But it’s clear the gobs of money raised and multiple business lines hurt the company. Pay By Touch really needed to be run as a small company for a while. Keep the focus simple - ACH payments in multi-lane grocery stores. The company would have run in the red for a while, and needed infusions of funding. But it would have time to work through the operational levers and to bring costs down. Work in concert with the grocers to shift consumers toward ACH usage (e.g. better product discounts for ACH users; higher discounts funded through CPGs’ trade promotion dollars).

Expansion into other areas would come after prudent consideration of what was needed to succeed. One good, but admittedly tough example: provide grocers with lower credit card interchange for biometrics than for mag strip cards due to lower loss from fraud.

Also, this is a company that really could have used the guidance of a traditional VC firm. The hedge fund investors ultimately were little more than silent money. Still not sure why hedge funds parked money in an illiquid, high risk start-up.

It’s Over

So now the remaining Pay By Touch employees are left to find new jobs. Creditors won’t see much of what they’re owed. Merchants have to tear all that hardware and software out of their stores.

But Pay By Touch’s acquired division S&H Solutions will carry on (independently or as part of some other company). It’s got a strong loyalty marketing business and some momentum from SmartShop implementations. And the biometric check cashing business, formerly Biopay, lives on as Phoenix Check Cashing. Good luck Jon Dorsey.

And someday, some start-up might try to do mainstream consumer biometrics again.