Hit Ctrl + R on Gov Tech Shopping | Letter to the Editor

It is not easy to index procurement specs by performance- and outcome-based attributes for many reasons: (1) Arriving at a consensus on the definition of outcome is hard (2) Even if that happens, it is impossible to ascertain that the received goods fulfill the outcome since most outcomes take time to be achieved. While the author names Geekbench and other benchmarks for outcomes, they are actually tests of output, and most performance criteria in government procurement already include such tests. (3) Even if consensus is reached at the point of procurement and delivery, the shape and nature of outcomes is unlikely to be durable and versatile enough to last through the product’s total ownwership experience of 5-7 years. Whereas a 386 processor will be a 386 processor for its entire lifetime.

I’m also reminded of the following lines in another recent op-ed entitled “End dependence days: What India should and should not build” by M Muneer that appeared in Economic Times dated 15 August 2026: “Civil servants are rewarded for avoiding mistakes, not for taking risks. CAG investigates failed purchases far more enthusiastically than missed opportunities. The safest procurement decision is almost always buying yesterday’s technology from yesterday’s vendor. Changing procurement rules does not change this psychology. It merely gives risk aversion newer forms to express itself.” This resonates strongly with my personal experience and, going by it, it’s naive to expect bureaucrats to drastically change the basic structure of specifications in their tenders.