May 2006 – 6 am pacific

The Virtue of Simplicity

The airlines business is a complex one. The pity is that most airlines
reflect the complexity in their business onto their dealing with passengers.
Passengers like me hate it.

I am planning a trip to the East Coast (I live in the San Francisco Bay
Area). The trip is a 3 city trip over 4 days. I generally prefer to do my
own travel planning. When I was at Infosys, I would talk to our company travel
agent. Now, in a startup, I find it easier and cheaper to do my own
travel bookings on the internet.

So far, all my trips to New York have been on JetBlue. They have convenient flights and low prices. This time I
need to go to Boston and Chicago as well and so JetBlue won’t work. So
I go to Expedia.com and check out the flights. My conclusion after 45 minutes
of research and copious note-taking – if I want to minimize my travel cost, I
will have to travel on 3 different airlines and fly in and out of different New York airports.

For most airlines, pricing is a game of revenue maximization. Here are some
tricks of the trade:

1. If you book your travel early you get a cheaper fare. Everyone uses this
one, including Southwest Airlines.
2. Refundable tickets cost more than non-refundable. Again, very widely used.
3. Take a hub, dominate traffic in and out of it and charge the earth for it. New York to Boston round
trip from two different New York airports can be $200 or $600 based upon the competition on that sector.
4. Round-trip fare is heavily discounted versus point to point.
5. Saturday night stay-over reduces the fare quite a bit.

There are countless other tricks that are all designed to maximize revenue.
Optimization engines and pricing rules in the innards of airline pricing
systems are some of the most complex you’ll find in the business world.

As a passenger I hate this whole system. I hate it that it takes me 45
minutes to do my tickets. I hate it that even after that, I don’t know if I
made the right choices. I hate it that I can’t travel back on a different
airline that has more convenient flights without paying a hefty premium for it.
And I cannot develop a trusting relationship with an airline who charges my
$600 when an equally good (or equally bad, depends on your perspective) airline
is charging a $200 fare for the same itinerary.

So here’s my question to these airlines. Do their fancy price optimization
algorithms put any value on what I can only call torturing the customer? I’ll
bet they don’t because they have no way to measure it or put a value to it. The
reason simplicity in business is so rare is that there are no good ways to
measure the cost of complexity. And so your finance types in the company can’t
put it into their cost-benefit analysis spreadsheets.

Complexity costs. Customers like simple products – simple to use, simple to
understand. They like simple pricing models where the price is linked to the
value they receive. This is not just true about simple-minded consumers. Business
buyers like simplicity as well.

Southwest Airlines is a company that I truly admire. The genius of Southwest
Airlines is in how they have become the most important airlines in the US by
simplifying it for their passengers and for themselves. In the morass of
complexity that is the American airlines industry, Southwest Airlines is a
shining beacon of hope. Not only is their pricing dead simple, everything about
the airlines is that way. They fly only Boeing 737s. This simplifies, crew
scheduling, training, aircraft maintenance and spares. They have only one class
– coach class. There is no seat assignment. It’s first come first serve. And
their frequent flier program is a tribute to simplicity – 8 round trips and you
get a free roundtrip to anywhere they fly. You would need a full book to fully
document the frequent flier program of United Airlines.

No wonder Southwest Airlines has delighted customers and a growing business.
Its market cap at $12.64B is way above much larger airlines like American
Airlines and United. They understand the virtue of simplicity. They understand
that it not only makes for happier customers, it also makes operations run
cheaper and faster.

Now if they’d only fly to the airports I need them to fly to.

Garage-Sale

In big business the phrase ‘cost of doing business’ is often employed to justify a cost that is deeply embedded or can’t be justified on an Return on Investment basis. Sometimes it truly is necessary. Often it is just handy management jargon to keep something from being cut by the accountants.

In a startup, on the other hand, the cost of doing business has a very real meaning. It in fact means exactly how it reads – the cost of running the startup. In management speak this is often referred to as the ‘burn rate’. And I am here to tell you, that the cost of doing business for a startup is going down fast.

There are many trends that impact the cost (and ease) of doing business for a startup. The two that I think are particularly pertinent to the cost of technology are – open source software and usage based pricing. Sometimes both these things get combined.

Let me give you an example. As an ex-head of Sales, I believe in a company acquiring good habits on Sales process early. Ergo, we use a CRM system at Gridstone. When making the choice for a CRM system we never even considered licensed software. If we had, I suspect it would have cost us several thousand dollars per user per annum, for something with the functionality that we need. We probably would have needed outside consultants to implement it adding more to the cost and time to value. Most startups nowadays don’t even consider that option.

The next option, which is very popular today is salesforce.com which is a hosted internet application. The company is very successful and is growing like a weed. Their price for the Enterprise edition – $900 per user per annum. Simple to implement. Good service. Works like a charm.

We were almost going to go with salesforce.com when we heard about this open source CRM software called SugarCRM. SugarCRM is also a company. I think the way their business model works is that they, and outside collaborators, work on the open source version of Sugar, which of course is free to use and comes with the source code for others to tinker with. But SugarCRM the company also develops add-on modules like Sugar-Outlook integration that are not open source and are priced per user per month. Sugar also offers hosting and training services around SugarCRM. If we wanted to take the full hosted solution with add-on modules, we would get an excellent CRM system that matches salesforce.com in functionality for the price per user of $480 per annum.

Now SugarCRM is open source and anyone is free to use it themselves or host it for others for a fee. There are dozens of hosting service providers who will host SugarCRM for you. Some of them even have add-on modules. And you won’t believe the prices. The service provider we finally went with charges us $5 per month. Not $5 per user per month. $5 per month period. For the usage it allows, for us it practically means unlimited users! Plus the price includes hosting for a few other applications as well. It’s a brave new world!

Now technology costs are not the only costs for a startup. All costs related to people, office space aren’t going down. But tech costs are a major component of the costs especially for early stage startups. For two guys in a garage, the quintessential valley startup, their runway is now much longer than it used to be. In the future more tech companies like Flickr will be going straight from garage to selling the company. It is ironic that this should be happening in an environment where venture capital is so abundantly available.

Zero touch traffic enforcement

One of my co-founders at Gridstone Research was driving us back from lunch in San Mateo a month or so back. As he approaches a red light where he is supposed to make a right, he doesn’t come to a full stop. He just ‘rolls’ through the right turn. Flash, flash from the wierd looking gadget across the intersection and a month later he’s got a ticket for $371! He had been caught commiting a traffic violation by technology.

My colleague was none too happy, but I was laughing my head off. Not just because the guy hates to lose money to such things, but mainly because I had just paid off an identical ticket for an identical violation near my home in Fremont. And my wife had contributed to the same cause a month before me. It was thoroughly enjoyable to see someone else bear the same pain.

I tried to piece together how this technology works. There is a sensor in the road that senses the speed at which you are approaching the intersection when the light is red. If it is above a maximum speed it sets off the camera which takes a frontal picture of you and the car. As soon as you turn right there is another sensor that times how long you took to get from the first sensor to the second sensor and if that time is less than some minimum, it assumes (correctly) that you did not stop at the red light. The camera then flashes again, getting the back side of your car. A couple of weeks later you get a ticket in the mail with all the time-stamped pictures clearly establishing your violation.

When this happened to me (just a few weeks after I had finished making fun of my wife and her ‘rash’ driving), I was upset, but also fascinated. I thought this was just so cool that traffic could be enforced without a police officer. The technology being used is pretty reliable. It does not make mistakes (far fewer than people at any rate). It uses no discretion. Unlike a police officer, it does not let off blondes in convertibles. It is always on. And best of all, it costs far less than the police officer. This, I thought, was great for traffic enforcement. The Fremont Police Department thinks the same way.

But then I thought there’s got to be more to this. I’ve often heard anecdotally that the Highway Patrol gets very strict in its enforcement a couple of weeks before the end of their budget period ends. Why? Because they have revenue targets to meet! You might say that’s not the purpose of the traffic ticket. But in the real world local governments need money to run. I am quite sure that the Red-light Cameras are being sold to the local government by justifying them on an RoI basis. If you invest so much in the camera system, the increased revenue will return your investment so many percent a year. etc. etc.

But wait, it can get better than that. If I was the Red-light camera system company, I wouldn’t even sell it to the city. I’d install it free and take a cut of the revenues. This way everyone wins. The city makes no investment and gets revenue from where it had none. The company gets an annuity, high margin stream of revenue from an unsuspecting public. The Fremont Police Department may proclaim that they are doing this for red-light enforcement. But we know better. Not that I hold it against them. Fremont city underfunds many civic services and I’m happy if they can enforce traffic rules and raise revenues.

That leaves the question – what is this company that makes these camera systems? It turns out that the company that does these red-light camera systems is an Australian company called Redflex Group. Which is unfortunate since my broker doesn’t let me buy Australian stocks. I think they are on to something. Redflex Group has a Traffic Systems business and a Communications business. The holding company’s stock has been doing rather poorly, I don’t know why but the Traffic system business has got to be smoking hot.

And guess what else is available from Redflex? Photo speed enforcement on highways! For those of you who have forgotten what 65 miles an hour feels like, there’s a traffic ticket coming your way.

The right time for labour market reform is now

In my first job out of business school with Hindustan Lever, as Area Sales Manager I had a team of over 20 unionized salesmen. I ran a Voluntary Retirement Scheme in my first year and then again in my third year by which time we were down to half the original team size. Yet we introduced new brands in the market, grew our sales and in general did well as a team.

In my third year I took over a small new business of hot beverage vending for Lipton (at that time a division of HLL). The business was small but growing rapidly as we expanded our city coverage. In Sales, Distribution and Service we had about 50 people. Of this the number of direct Hindustan Lever employees was 2. The rest were all outsourced, contract or distributor’s employees.

I then moved to Infosys in the US and over my 11 years there, hired scores of employees onto the company’s US payrolls. I also had to let go of some people for performance or other reasons. At all times, I was acutely aware that I myself was an ‘At Will’ employee. I could be fired with two week’s wages without giving a reason. As long as the reason was not discriminatory (race, sex, religion etc.) I could not bring legal action upon the company.

My experience with the vastly different labour environment in both India and the US has driven home a very important lesson – a business exists to make money for its investors, not to provide employment. And that is, paradoxically, the best way to generate employment.

Let’s take a look at how India’s labour laws distort the business environment and harm employment and employees:

1. It discourages capital investment – particularly in service oriented industries. Investing capital means taking risks. Market risk – the risk that the business may not succeed – is a risk that ‘comes with the territory’. In most countries, investors know that if their business fails in the market, they close down the business, sell off the assets at knocked-down prices, book the loss and take the remaining capital to some other investment opportunity. However, in India, failure, or a downturn, in the market also means that you are still saddled with the payroll costs because you can’t restructure or layoff anybody. You can’t exit the business because employees will lose jobs. That’s something investors don’t have to deal with in most countries. You look at so many rusting factories in every major city in the country where the factory owner has not been able to layoff employees even when the networth of the company has gone down to zero, and you wonder – what a colossal waste of assets. You also wonder – what do future investors think when they see these rust-buckets? More likely than not it’s – ‘That could be my investment 10 years from now.’

2. It provides no incentive for raising productivity through automation. Look at all the government offices or offices that have unionized staff like banks. To the last one, they opposed computerization. Why? because it could do the job faster and so it would reduce the number of jobs.

Yes it will and that is a good thing. Doing more work with fewer people raises productivity. Productivity raises incomes. The developed world’s prosperity is entirely linked to higher productivity. Also higher productivity creates the surplus (or the profit) that can be invested to create more jobs.

3. If you want to produce a quality product or service it needs carrots AND sticks. With an employee who is not performing, you train, you mentor, you put them on Performance Improvement Plans. But in the end, the employee needs to know that if his performance does not improve he can lose his job. Without this freedom for businesses to manage for performance, it may be possible to compete against companies who are similarly hampered, but it is a clear disadvantage in the global market.

4. It pushes employment generation into the informal sector. In my second stint at Levers I would have loved to hire people directly into the company instead of outsourcing critical functions like Sales. With the Levers brand name as an employer we would have got great talent which would have been better for the company. However, Levers would not do that for a new business that could have failed leaving them with employees they wouldn’t know what to do with. So all of the hiring was done by outsourced contract firms. Did these employees get the PF and benefits they would have got at Levers? I doubt that very much. I don’t think these contract firms even paid any taxes since they were probably classified as Small Scale.

In summary, the current labour laws in India distort the business environment to where it reduces employment generation by discouraging investment, reduces income growth by discouraging productivity increases, reduces quality by taking away the freedom to manage for performance and pushes employment generation into the informal sector.

Whenever I bring this up with people in industry in India, I am given many reasons why this is not a problem. Someone says ‘Only 20% of my workforce is unionized, I just work around them’. Another one will say ‘If you really want to fire an employee for performance, it can be done.’ But most of all the reason I get is ‘But the economy is doing so well why do we need to think about redundancies and labour flexibility?’ On the contrary, it is because the economy is doing so well that this is the right time for labour reform.

I believe this is the most important reform that government must now address. However, this is also the most difficult. Dismantling industrial licensing was like a walk in the park compared to this. With the government dependent upon the CPI(M) to stay in power makes it almost impossible to do major reforms. But major one-shot reforms aren’t the right answer anyway. There should be a 10 year road map on labour reform. But starting now. Let’s begin the discussion.