The software estimation debate iluminates a much deeper paradigm shift in how we make business decisions.
There is much debate in the software sector about not just how to estimate work, but whether we should even try.
The right people in a room can decide what to do next, taking into account an array of considerations of strategy, market, money, politics, technology, knowledge, and skills that any numeric estimation system never captures.
Also, whoever comes are the right people. Maybe the decision on what to work on next is taken by the “wrong” set, but they’re human adults. They can process the complexity of the factors to make a decision better than an algorithm can.
The idea of making our decisions based on a numerical ratio like cost-benefit analysis is going to increasingly seem silly. If only the world were so simple that cost could be foreseen and benefit could be quantified.
Since the Second World War, the world has been sufficiently stable and calm that cost estimating had some level of validity. Costs were close enough often enough. But we all know how dodgy they are.
Benefit estimating has usually been bullshit theatre to justify getting the money.
Most of us have been in the situation where an idea was rejected because the cost benefit analysis did not stack up when it was obvious to everyone that it was a good idea. Usually that reflects more badly on those who wrote the proposal, because anybody who can’t rig a cost-benefit analysis to come out with the right results isn’t very bright.
Or other situations where the numbers don’t give us the result we want so we modify the numbers. Any estimation that involves weighting is a sitting duck for this. You can Monte Carlo your way out of just about anything.
Then we get into the troubling zone of questioning whether at least some forms of estimates are more effort and trouble that they are worth.
There are some macro numbers which are useful in guiding decisions, such as cost of delay, but using numerical ratios to make decisions will increasingly look foolish. What got us here won’t get us there. Simple models don’t work so well.
The estimation debate is a symptom of a deeper clash of cultures: between structure/rules/methods/mechanisms to decide, vs fluidity/sensemaking/exploration/adaptation.
The latter is the future. It is the new paradigm of decision-making that is less about compliance to rules, and having justification as a defence against punitive review, i.e. all the evils of bureaucracy; and more about modern adults adapting constantly to a volatile world.
That word “paradigm” doesn’t get abused as much as it did in business. Here it is apt. It’s a new world order.