![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
Product planning games
Mar. 10th, 2011 09:50 pmJust 2 examples.
1. What if you decide when to start selling a product that should support some protocol that is planned to be ratified in 18 months from now. If you start selling when protocol is final, there will be competitors on the market for a while, though they will only support draft version. If you start earlier, there is a risk that it will not be fully compatible. Especially if protocol's logic is wired into ASIC rather then f/w or s/w. However if product is very successful or if you have big influence on standardization body you can mitigate this risk.
Bonus game: What if product should support more then one not-yet-released protocol?
2. What if you decide how to segment a product line. Let's assume it is apples (sorry Steve, let it be better pears). Let's assume you have enough money to create several product lines (green pears, small red pears, organic pears etc). Your competitor decides to design generic product (just orange pears of medium size). When you design several product families, you can add features that appeal the right audience and set different prices justified for each target market segment. On the other side designing several product lines is more expensive, the competitor with single product will fill all gaps between your products, and there is higher risk that some of the products will not be delivered in time.
Both examples are generic, I don't refer to any particular product. However they are totally realistic.
There must be formal mathematical methods that help to model the product design questions like above and find optimal solutions. I only seen how game theory can be used in similar but more trivial settings. In the cases above there are also so many technical assumptions and dependencies that I suspect no correct modeling is possible at all. So it is all intuitive.
So when I see how this kind of decisions are made in real life there is no math involved. I usually see quite smart people collecting data, reasoning together, coming up with reasonably good ideas and deciding to bet on one of those. The latest decision often depends on some internal politics.
When I am working with those people who brainstorm on the problems like above, my role is usually to provide data. Some times I see small part of the reasoning process, but usually I just get the final result. It would have been great to expose such reasoning more often - I think this is THE most fascinating part of our trade.
1. What if you decide when to start selling a product that should support some protocol that is planned to be ratified in 18 months from now. If you start selling when protocol is final, there will be competitors on the market for a while, though they will only support draft version. If you start earlier, there is a risk that it will not be fully compatible. Especially if protocol's logic is wired into ASIC rather then f/w or s/w. However if product is very successful or if you have big influence on standardization body you can mitigate this risk.
Bonus game: What if product should support more then one not-yet-released protocol?
2. What if you decide how to segment a product line. Let's assume it is apples (sorry Steve, let it be better pears). Let's assume you have enough money to create several product lines (green pears, small red pears, organic pears etc). Your competitor decides to design generic product (just orange pears of medium size). When you design several product families, you can add features that appeal the right audience and set different prices justified for each target market segment. On the other side designing several product lines is more expensive, the competitor with single product will fill all gaps between your products, and there is higher risk that some of the products will not be delivered in time.
Both examples are generic, I don't refer to any particular product. However they are totally realistic.
There must be formal mathematical methods that help to model the product design questions like above and find optimal solutions. I only seen how game theory can be used in similar but more trivial settings. In the cases above there are also so many technical assumptions and dependencies that I suspect no correct modeling is possible at all. So it is all intuitive.
So when I see how this kind of decisions are made in real life there is no math involved. I usually see quite smart people collecting data, reasoning together, coming up with reasonably good ideas and deciding to bet on one of those. The latest decision often depends on some internal politics.
When I am working with those people who brainstorm on the problems like above, my role is usually to provide data. Some times I see small part of the reasoning process, but usually I just get the final result. It would have been great to expose such reasoning more often - I think this is THE most fascinating part of our trade.
