I feel like you're onto something super valuable here, but I don't quite follow "Instead of seeing allocations as the input—the controlling mechanism—I see them as an output of priorities." Are you basically saying to allocate based on those qualitative measures and that will help to treat lesser priorities as lesser priorities and not over-invest even in the good stuff? Basically your critique is that the current model over-invests in the good AND the bad and if it's rethought we can do even more things at an even higher quality bar?
Let me try to rephrase it. Say we have a force ranked list of priorities that capture value/urgency (and through that risk profile, how we want to approach those efforts, the style of support needed by the team). An output of that .... assuming our org chart, etc. etc. is likely an allocation. An allocation will emerge. But say we went in there with the simplistic statement that we will allocate $ in a certain way assuming that = strategy/priorities. We might be very much deluded on how that will turn out.
To echo what John's saying another way: when you plan a trip to a distant city, you can't with any certainty say that you'll spend 80% of driving time on motorways, take breaks once an hour on the hour, and stop at your favourite roadside cafe for lunch. There are far too many choices and confounding factors along the way: roadworks, accidents, closures, interesting scenery, flat tyres, and much more. But if you set a list of priorities like "1. prefer major roads 2. take regular breaks 3. eat about halfway through the day" then you can make decisions moment to moment, as you drive and learn more about the territory (and not just the map), and you'll likely wind up with something like your original plan. [Insert obligatory mumbling about Snowden, Cynefin, and entangled trios here.]
In fact it's even worse: in most situations, you can't even control how much effort you actually put into a given bucket over a long time (a month or more). Even if you say you're going to work on something 20% of your time or allocate 20% of your team to it, you're likely to find, 10% of the way in, that the problem you're solving has a hairy bug or weird customer needs at the heart of it, and it's going to need 30% or more to get it done. Your choices are then to stop, meaning you only "spent" 10%, or continue and bust your budget. And that doesn't even cover crises and catastrophes (hello Crowdstrike!) that throw everything into a cocked hat. Micro-level priority setting like John advocates here, and over a much shorter time (today, or this week) is the only way out of this that I know.
I feel like you're onto something super valuable here, but I don't quite follow "Instead of seeing allocations as the input—the controlling mechanism—I see them as an output of priorities." Are you basically saying to allocate based on those qualitative measures and that will help to treat lesser priorities as lesser priorities and not over-invest even in the good stuff? Basically your critique is that the current model over-invests in the good AND the bad and if it's rethought we can do even more things at an even higher quality bar?
Let me try to rephrase it. Say we have a force ranked list of priorities that capture value/urgency (and through that risk profile, how we want to approach those efforts, the style of support needed by the team). An output of that .... assuming our org chart, etc. etc. is likely an allocation. An allocation will emerge. But say we went in there with the simplistic statement that we will allocate $ in a certain way assuming that = strategy/priorities. We might be very much deluded on how that will turn out.
Important nuance, thanks for clarifying!
To echo what John's saying another way: when you plan a trip to a distant city, you can't with any certainty say that you'll spend 80% of driving time on motorways, take breaks once an hour on the hour, and stop at your favourite roadside cafe for lunch. There are far too many choices and confounding factors along the way: roadworks, accidents, closures, interesting scenery, flat tyres, and much more. But if you set a list of priorities like "1. prefer major roads 2. take regular breaks 3. eat about halfway through the day" then you can make decisions moment to moment, as you drive and learn more about the territory (and not just the map), and you'll likely wind up with something like your original plan. [Insert obligatory mumbling about Snowden, Cynefin, and entangled trios here.]
Nice example!
In fact it's even worse: in most situations, you can't even control how much effort you actually put into a given bucket over a long time (a month or more). Even if you say you're going to work on something 20% of your time or allocate 20% of your team to it, you're likely to find, 10% of the way in, that the problem you're solving has a hairy bug or weird customer needs at the heart of it, and it's going to need 30% or more to get it done. Your choices are then to stop, meaning you only "spent" 10%, or continue and bust your budget. And that doesn't even cover crises and catastrophes (hello Crowdstrike!) that throw everything into a cocked hat. Micro-level priority setting like John advocates here, and over a much shorter time (today, or this week) is the only way out of this that I know.