To Module or not to Module...

The module is an engineering construct, not a business construct. While the advantages of modules among engineering types seems obvious, you should consider the business perspective.

When you design a product in the add-on module manner, each printed circuit board must be independently designed and fabricated. Often additional components are required to boost signals across the expensive interconnections between the boards. Manufacturing requires separate production runs for each module and so tend to multiply the unit production costs.

From a sales point of view, there are disadvantages too. The basic idea of modularity is that the customer buys only what he needs today and hopefully returns at a later date to buy other modules. In sales, you never depend upon future sales, you want to close the first deal for as much as possible before the customer either loses interest or finds a later available alternative. Close the immediate sale for the most you can... Modules run in conflict to this because it allows the customer to justify postponing some of his purchase.

Of course some would say modules allow the immediate purchase of specifically the configuration the customer requires and no more. While true, you must maintain all the inventory and production runs for the customers to find their selections. This is costly for the company.

Modules also compartmentalize the customer's view of features' premium costs... ie letting them see some of the pricing aspects that is of no value for the company to reveal.

The non-Module alternative is the Bundled-Features approach. In this approach, you design one system with rich features and then make marketing decisions to offer a small set of models. Here the idea is to create a minimum set to offer models for certain price levels the customers are likely to require. The objective is not necessarily to offer precisely what the customer needs, but rather to offer a model within the price they will pay that include the features they require.

If they have to pay for higher model to get their requirements, they can come to terms with that because they have no options to later upgrade as they would with the module approach. This puts the customer into a different buying position than under modules; he must consider his future requirements and buy to cover those now because he can't later upgrade. He must risk buying low and scrapping it later when his requirements grow, or else spend a little more now. When the customer convinces himself to spend a little more money with your company... that's a good thing. :)

Production costs are lower too. Only one system is designed to be used for all marketing models envisioned. Your development costs are much less because you only design and develop one model. The same board and case are used in all models and therefore you get the advantages of quantity purchasing prices. Automated printed circuit board (PCB) assembly lines easily deal with data where one model production run omits certain components in the same printed circuit board. The adjustment costs for production setup is much reduced.

By putting all the components on the same PCB, you can omit line drivers, connectors and cables between otherwise modules.

A more profitable product has the characteristic that its customer sought features are more realized in software than in hardware. Software development is a one-time cost whereas commponents cost you each time you build a unit. With features based in software, you have the profitable choice of pricing your product by the perceived utility of the customer rather than pricing as a ratio of your hardware cost to manufacture it. This advantage is most useful under a bundled-feature design than a modular add-on approach.

You also have another marketing option. You can strategically decide when to offer the various models and even redefine them with different manufacturing component insertion configurations and software.

For example... I'm releasing a new product line next year in a new market. I defined several models but will initially only reveal and release my mid range model. It is my most profitable by far and by offering only this, I'm able to quickly pay off my development costs and fund larger production runs.

When the timing is determine to be right, I can offer a lower price and lower feature model to keep people away from competitor's models. My production continues to buy components at high volume discounts and I update the pick and place machine data to insert a subset of components from the midrange product. I then disable some of the features in the software loaded into the low end product, stick on a new product label and I'm done.

With this production running, I can go back to the lab and add in the additional super-features for the high end model and use some of the existing new profits to fund its testing.

The thing to remember about offering different models is not to offer so many that your customers will spend less than they would with fewer options. More models means more production confusion and more inventory to be organized. With the bundled-features systems, you can, in a business emergency, offer an important customer who needs your model with no inventory, a higher model programmed as that lower model. Of course, I would offer him first a discount on the higher model to at least recoup some of the lost profits and it never hurts to let an important customer see a better model to intice them to later spend more.

I realize that this would not apply in all economies around the globe but it is something to think about. Remember that even though we are all engineering types, we have to think like businessmen to be sucessful.

aka j