Live from Business of Software 2011 - Clay Christensen, professor at Harvard Business School (@claychristensen) and popular author, kicks off the 2011 Business of Software conference in Boston...

Disclaimer: I took these frantic notes in realtime. Apologies for mistakes, inconsistencies, and blatant errors!

First career –founded and ran a company that made materials. Always wanted to teach. As he approached 40, he became a doctoral student w/ 5 kids and started a new career teaching at Harvard.

Principles of good management that they teach at places like Harvard sow the seeds for eventual demise. Q: is the enterprise of creating successful business is truly a crapshoot?
[No… it’s not a cookbook and still working it out, but it’s not a crapshoot.]

Decentralization is disruptive.

Trajectory of improvement almost always outstrips the ability of customers to use that improvement. Some people will always be hard to satisfy no matter what you deliver; others have simple needs and you can overdeliver very easily.

Incumbents tend to dominate sustaining battles (where steps forward are sustaining innovations). The leaders win. But there’s another type of technology that kills the leaders – we call it disruptive innovation. Instead of sustaining a trajectory, it discrupts it by bringing to market a technology so affordable and accessible that it brings to market a whole new set of services.
Entrants typically win at disruption.

Right now people explain company success or failure by blaming the management team. When DEC was doing well, they attributed it to the mgmt team. Same when they failed. But that didn’t ring true… especially since all the other computer companies were collapsing too. They might collude to price fix, but probably not to collapse. :)

DEC made mini-computers – mainframes. Needed to generate gross margins of 45% on a product that sold for $250k.
People designing new DEC computers offered 60% on $500k product.
Meanwhile others were suggesting going after the personal computer market.
But to their customers – business – the Personal Computer didn’t matter – so they got no signal that this market was important.
So they chose to make better products they could sell to make better products to their best customers.
Or they could make worse products (PCs) that none of their customers would but that cost less.

They made a decision that made "sense."

The tenets provide good guidance as you move up the Sustaining trajectory. But those principles of good management then hurt – they make it almost impossible for rational people to go after.

Toyota came in from the bottom while Detroit made the big cars. But now they’re into luxury, and Hundai and Kia are taking over the low end.

2 sides of a coin:
What’s the kind of innovation that could kill your company?
And if you wanted to start a company that would kill others, what would do it?

Until mid60s, most consumer electronics were made with vacuum tubes.  TV’s cost about $2500 in today’s $.  And they were big.
Transistor was disruptive. But couldn’t handle the power.
Everyone who did vacuum tubes took transistors into their labs and tried to show that they were technically deficient.

So the first app for transistor was hearing aid (coulnd’t make w/ vacuum tubes). Then Sony introduced transistor radio - $2, fit in your pocket. Crummy product but infinitely better than nothing.
In 1959 Sony introduced portable tv. Simple product so whole new population of people could have one, because it was infinitely better than nothing. Then little by little these products get better.  When this occurs, it always has a different measure of performance. The people in the back say it’s crummy, and the people in front don’t care, because they have a different measure.

Online learning is disrupting the Harvard Bus School. How they measure – in the back – is the quality of the faculty, which is measured by their research.
Others measure the quality of teaching. The online courses are good and getting better, while HBS is off doing all of their research.
U of PHX are spending $200M/yr making software that’s better for teaching. Harvard spends $0 teaching better.

Worry about the bottom when you’re thinking about who can kill you.
And if you want to kill someone, kill them at the bottom. (Like a little boy kicking a giant – pick it where he’s motivated to flee rather than fight). When think about how big is the market, it will appear that the center is bigger… but if you make something affordable and simple, that’s probably not true.

Green energy is one are. Solar electricity is one example. Trying to make it good enough to use here. But in places like Mongolia and Africa, where they have no electricity, solar electricity is a booming market growing at 50%/yr without subsidy. Clay thinks the folks competing against nonconsumption who are slowly making their product better and better will eventually win… not big government subsidies.

Expensive failure almost always results when discretion is crammed into direct competition with established technology.

What is a business model and how is it built?
Need to give it the charge to kill the parent business model
IBM dominates computers. IBM is the only one who made it through the minicomputer disruption (out of 9). They created a new group in that did minis at 45% margin rather than 60% margin on mainframes in another city. Then for PCs set up something in FL.

Business units are not designed to evolve – they do their function and they die. Corporations evolve – they can shut down business units.

Market Understanding that Mirrors how Customer Experience Life
“ 'Always listen to your customers' is wrong.”

Rather, people bring in tools need to get their job done.

Understanding the job your customers are doing is critical to succeed with a new product.

Milkshake Story – fast food chain polled customers to see what they wanted, improved the milkshakes, had no impact on sales.
Turns out more than half of the milkshakes were sold before 8am by people who were alone and drove off with them.
Turns out the milkshakes all had the same "job" to do – they had a long and boring drive to work and they just needed to do something while they drove to keep themselves occupied. Weren't hungry yet but knew they would be by 10am.
Milkshake is really viscous so takes a nice long time to drink, gave them something to do.
Competitors are bananas, donuts, bagels, coffee, etc – not other milkshakes.

In the afternoon, it turns out the milkshake does a different job for customers – often employed by fathers who have been saying No all week long and want to be able to say yes.
BUT - it's so viscous it takes forever to drink, so parents give it up and throw it away because they don’t want to sit there. So the morning and afternoon requirements are totally different.

In the morning, you make it viscous and stir in a little fruit or chunks, not to be healthy (that’s not the job) but more interesting to drink. Let people dispense and swipe and go.
In afternoon, different needs.

The unit of analysis is the job, not the customer. It’s not what the customer thinks he needs, it’s what job needs to be done.

If you understand the job, you can understand how to improve it so it does the job even better.
Customer rarely buys what the company thinks it’s selling him.

The probability that a software product will be successful is nearly zero if the idea was developed within a company that thinks they know what the customer wants.
Figure out what they’re trying to do, and help them do it even better.
People in the labs need to be out in the world watching what people do, and do it better.

Four levels in the architecture of a job
4. Base (bottom) level – What’s the job to be done? Each job has functional, emotional and social dimensions
3. What experience in purchase and use must we provide to do the job perfectly?
2. What and how to investigate?
1. Apurpose brand (at the top)

Integrating correctly to get the job done means
-    Jobs are very stable over long periods. They are not vulnerable to product life cycles
-    Where products are easy to copy, integration around a job creates defensible differentiation. Hard for competitors to copy.
-    Customers are happy to pay a premium price, instead of a zero sum relationship

Examples: IKEA, TurboTax, QuickBooks, SAS, Zara, NxStage, OnStar

How do I kill other people and have them not kill me?
If you want to attack the incumbents , look to break in from the bottom and provide disruptive technology
If I try to listen to the customer, it will mislead me – I need to understand the job.

Sometimes you just need to change the basic business model
1.    Solution shop business – they get their money on a fee for service basis. For example, healthcare.
a.    Consulting firms, high end law firms, R&D organizations, complex medical diagnosis, special education
b.    Fee for service

2.    Value-adding process businesses – make things better
a.    Manufacturing, education, food services, medical procedures, instruction, textbooks, educational software today
b.    Fee for outcome

3.    Facilitated network business
a.    Telecom, insurance, ebay, D-Life, educational software tomorrow
b.    Fee for membership or transaction

The instructional materials business historically has been a value-adding process business.

The right product architecture depends upon the basis of competition. In some cases, don’t try to beat competitors w/ their own business model – there are a lot of failures that way. Think about a different model.

HBS is being disrupted by low-end learning experiences like you’re having today.

“Recovery” without jobs. How does this work?!
Synthetic, phantom or profits... how we account...
-    GM%
-    EVA (Economic value added)
-    RONA – return on net assets – causes you to try to get things off the balance sheet, which you do by outsourcing
-    DCF
-    IRR - Faculty measures profitability by ratios – internal rate of return. The more you focus on IRR, the shorter the time horizon for investors.
-    Marginal vs full cost
These measures are causing us not to create new jobs.

Efficiency innovations – how to make things at lower cost – take jobs out of economy.

Disruptions create jobs – make things more affordable and accessible so a larger group can have access to it. And then you have to hire more people to make it available.

There are fewer new-market disruptions over last 15 years
-    Cloud
-    Online learning
-    Initial steps in healthcare
-    Digital creation of video, audio
-   
Clay thinks there are fewer innovations now because of the way we measure things.

In net present value – you see the cash flow and it makes more sense to do nothing. By definition these measure underestimate the return of innovation.

The common methods of financial analysis systematically bias managers against innovation

Outsourcing often sets in motion disruptive business model liquidation.  (As in, liquidate the company!)

Dell has outsourced everything... shocking example...
Dell –originally did brand, product design, supply chain & logistics, computer assembly, mother boards. Was just outsourcing simple circuit boards to AsusTek.
Then they started to shift mother boards to Asus (Dell profitability (?) didn’t change, Asus went up)
Then they shifted Assembly to Asus, who could do it for lower cost (20% lower plus got manufacturing off balance sheets. Profitability improved but revenue didn’t change).
Same for supply chain – they could do it for 20% less and could get all the current assets off the balance sheet. Revenues were unaffected but profit looks great – especially return on assets bc they got no assets. AsusTek’s revenue improved.
Then same for product design. Then Dell could fire all their engineers and drive costs lower bc their brand is their core competence.
So all that was left with Dell was Brand.
So then AsusTek (now doing simple circuit board, mother board, computer assembly, all of it!) went to Best Buy and said we’ll give you our brand at 20% lower cost. So now Dell has nothing.

Intel still makes their chips in the US – everyone else gets them abroad. Right now most people admire fabless semiconductor companies for having gotten everything off the balance sheet. Today it costs about $10B to build a fab.

Americans are fixated w/ keeping assets on the balance sheet. Other countries are quite happy to have it on their balance sheets. Why? Americans are awash in capital so it makes no sense! But Americans measure profitability by ratios, so you can compare one to another, but no banks accept deposits measured in percentages. Another way to measure profit is by tons of money. Doesn’t matter how many assets are employed if you are generating tons of money.

It’s like a cocoon of finance people measuring profitability by IRR and RONA – want to get more out than they put in. Migratory capital causes us to do these things w/ short time horizon.
Business rules are right at the core of the problem.