Shipulski On Design

All-or-nothing thinking is exciting – we’ll launch a whole new product family all at once and take the market by storm! But it’s also dangerous – if we have one small hiccup, “all” turns into “nothing” in a heartbeat. When you take an all-or-nothing approach, it’s likely you’ll have far too little “all” and far too much “nothing”.

Instead of trying to realize the perfection of “all”, it’s far better to turn nothing into something.  Here’s the math for an all-or-nothing launch of product family launch consisting of four products, where each product will create $1 million in revenue and the probability of launching each product is 0.5 (or 50%).

1 product x $1 million x 0.5 = $500K

2 products x $1 million x 0.5 x 0.5 = $500K

3 products x $1 million x 0.5 x 0.5 x 0.5 = $375K

4 products x $1 million x 0.5 x 0.5 x 0.5 x 0.5 = $250K

In the all-or-nothing scheme, the launch of each product is contingent on all the others.  And if the probability of each launch is 0.5, the launch of the whole product family is like a chain of four links, where each link has a 50% chance of breaking.  When a single link of a chain breaks, there’s no chain. And it’s the same with an all-or-nothing launch – if a single product isn’t ready for launch, there are no product launches.

But the math is worse than that. Assume there’s new technology in all the products and there are five new failure modes that must be overcome.  With all-or-nothing, if a single failure mode of a single product is a problem, there are no launches.

But the math is even more deadly than that. If there are four use models (customer segments that use the product differently) and only one of those use models creates a problem with one of the twenty failure modes (five failure modes times four products) there can be no launches. In that way, if 25% of the customers have one problem with a single failure mode, there are can be no launches.  Taken to an extreme, if one customer has one problem with one product, there can be no launches.

The problem with all-or-nothing is there’s no partial credit – you either launch four products or you launch none. Instead of all-or-nothing, think “secure the launch”. What must we do to secure the launch of a single product? And once that one’s launched, the money starts to flow.  And once we launch the first one, what must we do to secure the launch the second? (More money flows.) And, once we launch the third one…. you get the picture. Don’t try to launch four at once, launch a single product four times in a row. Instead of all-or-nothing, think one-in-a-row, where revenue is achieved after each launch of a single launch.

And there’s another benefit to launching one at a time. The second launch is informed by learning from the first launch.  And the third is informed by the first two. With one-in-a-row, the team gets smarter and each launch gets better.

Where all-or-nothing is glamorous, one-in-a-row is achievable. Where all-or-nothing is exciting, one-in-row is achievable. And where all-or-nothing is highly improbable, one-in-a-row is highly profitable.

Image credit – Mel

Comments are closed.