AT&T: Lesson from the Crypt # 2: You Can’t Innovate Flawlessly

“Execute flawlessly!” was a mantra at AT&T.  This noble sentiment helped to sink the company by discouraging risk-taking and innovation.

One of the best things about AT&T was a strong commitment to quality.  This commitment was absolutely appropriate in a company which provided essential phone service to most of a nation and was a big part of what built AT&T’s powerful brand.

This commitment to quality was reinforced by rate-of-return regulation which controlled AT&T’s earnings through most of its existence as a regulated monopoly.  The more the company spent on its infrastructure, the more dollars of return on that capital it was allowed to recover in rates.  There was little inducement to reduce expenses since regulators took these into account in setting rates.

The result was that consumers of phone service got more quality than they now choose to pay for in a competitive market.  “Six nines” of reliability was the standard of AT&T engineers.  This means that only one in a million calls can fail.  Are you willing to pay significantly more for this high standard than say having one in 100,000 calls fail? One in 10,000?  Most people aren’t but they didn’t have any choice.  You got six nines or nothing.

Many years ago consultants convinced AT&T executives that cellular calling (large parts of which were invented at Bell Labs) would never be very successful because the quality was so bad.  AT&T disposed of the business for a song (later bought back in for a fortune and then exited again because it needed the cash to buy cable networks).  The six nines song deafened the company to the fact that consumers would trade quality for mobility.

The consultants (same firms, different people) used the same arguments against AT&T becoming a dialup ISP.  The engineers were particularly receptive as were the brand police.

“You can’t get six nines of reliability for Internet access,” the engineers said; “can you?”

“No,” I conceded.  “I think we can do one nine if we’re really good.”  I had been hired from Microsoft to develop an Internet strategy for AT&T and implement it, an offer I couldn’t refuse.

“The brand stands for six nines of reliability,” said the brand police.  “This is what our customers expect of us.”

The brand police deserved an answer.  The brand was very valuable.  Most of the nation was still paying a premium in a competitive market for AT&T branded calls.  In fact, using the brand was a huge part of my launch plan for AT&T WorldNet Service – “Internet Access for Everyone” meaning not just the nerds.

To answer them we had to go back in the archives to the early twentieth century when Theodore Vail was building what later became the behemoth AT&T.  Phone service didn’t work in the rain.  Rats ate the wires when the wind didn’t blow them down.  The service was plagued by power failures.  But it was the best that was available!

We claimed that the brand stood for bringing customers the best available communication solutions for the kind of communication they wanted to do.  It was becoming very clear that Internet access was something our customers wanted.   In order to preserve the brand and in order to succeed, we had to be the best ISP – certainly had to be better than AOL and MSN - but we didn’t need six nines.  We owed it to our customers and to the brand to be honest.  We wouldn’t promise six nines when we could only deliver one.  But we would deliver what we promised.  And we would always lead in quality.

We also used the horrible example of AT&T’s initial spurning of the cellular business remorselessly.  To the credit of AT&T executives, they rejected the advice of consultants and decided to fund the ISP business.

“You’ve got the money,” my boss told me.  “Now execute flawlessly!”

My delight immediately turned to dismay.  “We can’t,” I said.  “I’m going to have to give the money back.”

“What do you mean?”

“Remember the engineers said it would take seven years to build an ISP,” I said.  “That’s how long it will take if we execute flawlessly.  They’re right.  But we have to do this in seven months.”

“Maybe you should take a little longer,” he said.

“Don’t have longer,” I said “The world is changing too fast.  I’m not sure what it’ll look like in seven months.  I have no idea what it will look like in seven years.”

He was kind enough not to remind me that some people said that I shouldn’t be building a service at all if I wasn’t sure what environment I’d be launching it into.  In fact, he understood the argument and let us go ahead.

Sure enough, we didn’t execute flawlessly.  We made lots of mistakes.  But we did launch a service with one nine of reliability in seven months.

In order to succeed, we had to create a sub-culture in AT&T where mistakes were allowed and the ultimate sin was not admitting a mistake.  Banning mistakes stifles innovation.

Fear of mistakes makes people move too slowly.  Fear of mistakes creates endless meetings where people seek a meaningless consensus in order to be able to share the blame if things go wrong.  Most important, fear of being blamed for a mistake causes people to cover up – often even from themselves.

If you are innovating in a rapidly changing world, you must move fast.  If you move fast you make mistakes.  If you admit your mistakes and learn from them, you succeed.  If you don’t admit your mistakes, you don’t fix them and you fail; it’s as simple as that.

It’s like skiing.  If you get down the hill and there’s no snow on your jacket because you never fell, you weren’t trying hard enough.  You weren’t pushing yourself.  Innovation is largely a process of making mistakes and learning from them.  If you’re an innovator, you’re going where no one has gone before; it’s not surprising that you don’t know the way.

In this story from the crypt, AT&T did overcome its mantra of flawless execution and succeeded in launching an ISP which was an initial huge success in terms of signup.  See Lesson From the Crypt #1 for why the success was short-lived.

But shortly after, when VoIP loomed on the horizon, the initial low quality (and it was bad) provided an excuse for AT&T not to be the leader it could have been in this technology.  “Executing flawlessly” became one of several excuses for not “encouraging” a technology which threatened the cash cow of existing service but that’s another lesson for another day.

Lesson from the Crypt #1 is don’t manage for quarterly results.

Lesson from the Crypt #3 is vertical integration doesn’t work anymore.

Lesson from the Crypt #4 is don’t sent your losers to heaven.

Lesson from the Crypt #5 is navel gazing is a bad culture.