## Yet another webapp that is going to change the way we do …

Let us say we have this problem at hand:

There are seven stacks of 100 coins each. Each coin weighs exactly 10g. But we think in one of the stacks the coins (all 100 of them) weigh 11g each. We have here a very accurate and expensive electronic balance. How will you find the stack of abnormal coins?

Let us see how the three different generations of software engineers will solve this.

c/c++ Engineer:
Look at 7 stacks as 7 bits. Let us say, if the bit is 1 then it has 11g coins else it is 10g coins. So we simply need a weight to turn the respective bit to 1 if the stack it represents has 11g coins.

You pick 20, 21,22,…,26 coins respectively from each stack, weigh them once. A moment’s reflection will convince you this works not just for one abnormal stack but any number of them. All this with just 1 weighing.

Agile Java Engineer:
Let us separate the interface from the implementation. Let us also recognize that we only need to solve today’s problem of one coin stack being abnormal and not tomorrow’s problem of any number of stacks being abnormal.

The simple implementation we will provide will pick 1,2,…,7 coins respectively from each stack and weigh them once. If a stack has 11g coins, then you can find it easily from the weight difference. One weighing will still do.
Say, tomorrow the product manager changes the requirements to find more than one stack? We simply provide a new implementation for the interface. Codes using the same interface need not change one bit.

Webapp Ninja:
I will have this scale hosted as a web service with REST API. It will take one coin at a time and will return the weight. You implement the client code to pass coins from stacks 1 through 7 to find the abnormal stack.

This works not just for one stack but for any number of stacks. Yes it takes 7 weighing, but we are in the age of utility computing. Computing cycles and bandwidth are cheap.

Why bother with complex algorithm when resources are in abundance? Besides, what will those two guys do when there are not 100 coins per stack?

Oh, we will provide our own browser based client as well, complete with Fail Whale page and ShareThis buttons.
Actually, we were coding while you were talking and lready have this up and running at coin.ly (or weighify). This is a ridiculously easy tool to weigh coins.

We will price this using Freemium model,

Free: 1000 coins per month
Pro: 10,000 coins per month
Unlimited: Unlimited coins per month
Now go away, I am busy designing the pricing page.

No wait, we are building a platform. We are only beginning to scratch the surface.

Note1: Yes, I understand it is not possible to pass a coin through an API or enter it in a website. But let us not let that interfere with the story.
Note2: At least the webapp Ninja is thinking about monetization.
Note3: Surprised by this post? I have software in my upbringing, having written c/c++/java code for embedded real-time systems and as senior software manager I led a team that developed CDMA Radio Basestations.

(Photo courtesy of Flickr user Rodrigo Amorim )

## Multi Price Point Strategy From Nestle’s Playbook

Update: I learned today from a boom about Cadbury’s that Nestlé sells 1 billion products worldwide. That is at least one billion prices. That is mind boggling. But the number sounds suspect to me considering their total revenue. Can someone fact check?

In the Nestle 2009 roadshow presentation (PDF) there are three slides that read like a lesson on multi-version pricing. One of the slides is shown in this post (advance thanks to Nestle). The numbers are not exact currency numbers but rather a price index showing relative price position. Nestle says in the transcript of the presentation,

You see here an example of PetCare which has shown and proven also to be a very strong resilient and defensive sector in bad times. And you see how we through the brands are allowing the consumer to have different price points. These are all Purina products. You see it in dog food; you see it in cat food how we have spread over different price points the product portfolio and yet using the Purina brands there rightfully.

Nestle is one of the few companies to report profits this quarter. As I wrote in my previous post, one reason is their price increase. Nestle’s multi-price point strategy seems almost so simple and easy that it will not be a surprise if anyone reading this raised the following questions:

1. If it is this simple why not everyone do just this?
2. If this works for Nestle why would it disclose this secret to the rest of the world?
3. If this is so effective why not have more products at different price points, in fact take it to the extreme and have one price per person?

The answer to the first two question is the the same,

• This works for Nestle only because of the brand equity and the strong brand loyalty of its customers. Neither of these can be build overnight.
• Pricing and multi version products are aligned perfectly with the core strategy. In the case of Nestle its competitive advantage comes from a tight integration of its R&D, product innovations, comprehensive geographic presence and from its people, culture and values.

Multi-price point may seem easy to copy but without getting every strategic component right simply copying it is not going to help a competitor.

So why not take multi-price point to the extreme? This requires much longer discussion and I will write on it in a later post. But if you have your thoughts on this please share.