Smart Programme Management — Is It a Thing to Be Adopted?

Every few years the tools of programme management acquire a new adjective. We have had integrated, agile, adaptive, data-driven, and now smart: dashboards fed by real-time MIS, geotagged field verification, AI-assisted risk flags, command-and-control centres with very large screens. Vendors sell it as a category. Conferences have made it a track. The question this post takes seriously is whether "smart programme management" is something an organisation should adopt as such.

My answer, argued below, is mostly no. Not because the tools are bad, but because the category is a wrapper, and wrappers are where budgets go to be forgiven.

What the smartness actually consists of

Unbundle the package and it contains, roughly, four claims.

First, that data should arrive continuously rather than through quarterly progress reports. Second, that field reality should be verifiable, through geotagging, timestamps, and photographic evidence, rather than asserted. Third, that deviation should be flagged by the system rather than discovered by a reviewing officer's patience. Fourth, that decisions should be taken against this evidence at a cadence the evidence makes possible.

Stated this way, the first two claims are simply correct and no longer novel. Continuous reporting and geotagged verification have earned their place; several national missions demonstrated years ago that a scheme can know, daily, what was true monthly before. Where a programme still runs on typed monthly progress reports compiled from block-level registers, digitising that pipeline is not "smart", it is overdue hygiene, and it needs no philosophy to justify it.

The interesting failures live in claims three and four.

The dashboard that nobody feared

Consider a composite, drawn from patterns anyone in this field will recognise. A state-level programme installs a handsome monitoring centre. Every district reports daily. Red, amber, green. Within six months, a strange equilibrium: the dashboard is almost entirely green, field outcomes are unchanged, and the review meetings have become shorter, because the screen answers the questions before anyone asks them.

What happened is not mysterious. The system measured what could be reported, the field learned what the system rewarded, and the reporting layer adapted faster than the delivery layer. A junior engineer who can close a work order with a photograph will produce photographs. Indicators that become targets stop measuring; this is the oldest law in public administration, and no amount of processing power repeals it. The smart system did not remove the gaming of the old progress report. It automated it, and lent it a credibility the typed report never had.

This is the core of my scepticism. The technology layer inherits the incentive layer beneath it. Where supervision was already honest, digitisation makes it faster and cheaper, a genuine gain. Where supervision was theatre, digitisation produces faster, cheaper, better-looking theatre. The tool amplifies the institution. It does not correct it.

There is a second, quieter cost. Continuous visibility invites continuous interference. Programme staff who once had a quarter to try, fail, and adjust now explain every amber cell weekly to someone two levels up. Some kinds of delivery work, community mobilisation above all, are made measurably worse by being watched at this frequency. The old reporting rhythm was slow, but its slowness sheltered judgement. We should at least notice what we are trading away.

So: adopt, or not?

Mostly no, if "adopt" means purchasing smart programme management as an integrated ambition, with the platform procured first and the questions decided later. That sequence produces the green dashboard above, plus an annual maintenance contract.

But the conditional matters, so here it is plainly. Adoption is justified where three conditions already hold. The programme knows the five or six decisions its data must serve, and can name who takes each one. The verification design assumes gaming and is built by people who know how the field routes around measurement, which usually means involving the block-level cadre in designing it rather than presenting it to them finished. And there exists at least one senior officer whose behaviour will visibly change because of what the system shows, including when it shows failure. Where those hold, the tools repay themselves quickly, and refusing them is nostalgia.

Where they do not hold, the honest recommendation, and the one a consultant selling implementation services will rarely give, is to spend the first year on the boring substrate: clean beneficiary registers, reconciled financial data, a review meeting that reads what it already receives. A programme that cannot use slow data will not be rescued by fast data.

Smartness, in other words, is not a thing to be adopted. It is a property that emerges when honest measurement meets someone willing to act on it. Buy the second ingredient first.