4 ms·

There is an underlying assumption that a basket of features = sum(individual features). That is true sometimes, but other times a bunch of features launched tog

There is an underlying assumption that a basket of features = sum(individual features). That is true sometimes, but other times a bunch of features launched together have a compound aggregate value - they're much more valuable launched together than individually. But for cases where that's not true this is an excellent insight. The part I didn't understand was the P50. Why do you get a demerit if you do 100% of everything, and how is discovering and solving an objective discovered along the way encouraged? I would love if the author or someone else explained it to me. Overall a great article. Bit of advice: I'd mention early in the article and/or the title that this will apply to product thinking. I was curious about the financial part and pleasantly discovered that it uses this learning in product development - though way far down the page, I almost didn't finish reading it.

andreareina 7y ago
"P50 goals address sandbagging." Hitting 100% when the expectation is ~50% means implies a really bad estimate, or that the estimates were a lie, probably in an attempt to look good.
- archgoon 7y ago
  Which on the face of it almost seems reasonable; but it seems that this would just introduce another form of sandbagging. "What are your goals for this quarter?" "Make the color of the button to be green instead of red and World Peace"
mannykannot 7y ago
> [A bunch of features] are much more valuable launched together than individually. More than that, they are often dependent on one another - maybe not so much if you are just polishing a mature product, but much more so at startup, when you are aiming for MVP. And a feature does not become an 'option' just by putting it on a list; you have to work on it to have a possibility of 'exercising' it. To be fair, the article sort-of acknowledges this, but does not recognize the extent to which this weakens the metapor. In the P50 scenario, I can only guess that hitting all the goals earns a demerit is because it indicates sandbagging. Well, there is an obvious way to avoid that outcome...

andreareina 7y ago

"P50 goals address sandbagging." Hitting 100% when the expectation is ~50% means implies a really bad estimate, or that the estimates were a lie, probably in an attempt to look good.

archgoon 7y ago
Which on the face of it almost seems reasonable; but it seems that this would just introduce another form of sandbagging. "What are your goals for this quarter?" "Make the color of the button to be green instead of red and World Peace"

mannykannot 7y ago

> [A bunch of features] are much more valuable launched together than individually. More than that, they are often dependent on one another - maybe not so much if you are just polishing a mature product, but much more so at startup, when you are aiming for MVP. And a feature does not become an 'option' just by putting it on a list; you have to work on it to have a possibility of 'exercising' it. To be fair, the article sort-of acknowledges this, but does not recognize the extent to which this weakens the metapor. In the P50 scenario, I can only guess that hitting all the goals earns a demerit is because it indicates sandbagging. Well, there is an obvious way to avoid that outcome...