Why Objective-Setting Fails in Public Affairs (And What to Do Instead)

Why Objective-Setting Fails — Carousel.pdf

Download PDF • 1.51MB

By Paul Shotton, Advocacy Strategy

Public affairs functions are under more pressure than ever to show their value. Budgets are scrutinised, functions are professionalised, and dashboards, reporting templates and AI tools have made it easier than ever to generate numbers about the work. So it is worth asking a simple question before producing any of them: what is measurement actually for?

Because the standard answer to why objective-setting goes badly — "people don't like KPIs" — is the lazy one. It collapses at least five genuinely distinct problems into a single grievance, and because we treat them as one complaint, we never solve any of them. The trigger is familiar: someone asks "how will you measure success?" and what follows is discomfort, a little defensiveness, and a hastily assembled list of meetings held and papers written.

Here is the insight up front, because the rest of the article is an argument for it: the purpose of objectives and KPIs is not to prove performance — it is to improve judgement. Hold the system that way and most of the "resistance" dissolves. To get there, the five problems hiding inside that one complaint need separating out:

1. People were never trained in objective-setting frameworks.

2. People fear measurement — they expect it to be used against them.

3. The environment is complex and causality is genuinely unclear.

4. Objectives become outdated as reality shifts underneath them.

5. Organisations confuse consistency with rigidity, and adaptation with drift.

Each is real. Each has a different fix.

The training gap nobody names

Start with the most overlooked problem, because it explains a lot of the rest. Public affairs professionals are experts — in policy, regulation, advocacy, communications, law, or the science behind their sector. That is what they were hired for. What almost none of them were ever taught is the machinery of management: SMART objectives, OKRs, KPIs, logic models, Theory of Change, project management.

So there is a mismatch built in from day one. We expect finance professionals to understand accounting frameworks and lawyers to understand legal ones, but we routinely ask advocacy professionals to measure impact without ever teaching them measurement — then act surprised when the conversation goes sideways.

This is not a competence problem; it is a training problem, and training problems are the most fixable on the list. A team that has never shared a vocabulary for objectives can acquire one in a single workshop. In practice, much of the "resistance to KPIs" leaders complain about simply evaporates once people understand what the tools are for.

Fear: the difference between a learning tool and a verdict

If the training gap is the widest problem, fear is the deepest. Most people, handed a KPI, hear: "this is how management will decide whether I am good or bad." What they should be hearing is: "this is our best attempt to understand whether our strategy is working." Those are opposite purposes, and people work out very quickly which one they are living under.

The distinction maps onto organisational maturity. In mature functions, KPIs are learning tools; in immature ones, they become judgement tools. The moment a missed KPI reliably triggers blame, honest engagement dies. People set targets they know they can hit, they sandbag, and they stop surfacing the assumptions that turned out to be wrong — which are exactly the assumptions everyone most needs to see.

This is the settled view in the OKR literature, not a soft observation: practitioners are near-unanimous that objectives should be kept separate from performance evaluation and pay, because linking scores to compensation drives gaming and sandbagging and destroys the ambition the system was meant to create. When targets are aspirational, missing them is information. When they decide someone's bonus, missing them is a threat — and people manage threats by lowering their exposure to them.

The fix is cultural, and it shows up in how you run the review. If a missed number leads to "why did you fail," people disengage. If it leads to "what did this tell us about our assumptions, what changed in the environment, and what should we do differently," people lean in. Same number. Completely different function.

Causality: why public affairs is genuinely different

This is where public affairs genuinely differs from most other corporate functions. A salesperson closes a deal; a factory reduces defects. In each case the chain between activity and result is short and attribution is clean.

Public affairs looks more like this:

Activity → Engagement → Stakeholder shift → Political momentum → Policy outcome → Business impact

At every link, dozens of other actors are pushing and pulling. The further you travel along the chain — away from your own activities, towards the outcomes you actually care about — the weaker the direct causal claim you can honestly make. Evaluation specialists describe advocacy as defying traditional measurement precisely because influence is relational, informal and political, and because the target keeps moving as other actors respond.

Two moves follow, and the sophisticated teams have absorbed both.

Separate attribution from contribution. Attribution asks "how much of this change did we cause?" — usually unanswerable in a crowded political arena. Contribution asks "did we plausibly help bring this about, and can we tell a credible, evidence-based story about how?" That shift, developed most fully in John Mayne's contribution analysis, is the single most important conceptual move in advocacy measurement. It lets you make honest causal claims without pretending one meeting passed a law.

Measure in layers. Because you cannot hang everything on a single outcome metric, mature functions track three:

- Activities — meetings held, briefings delivered, coalitions built. Fully within your control, easy to measure, weak evidence of impact on their own.

- Intermediate outcomes — stakeholder understanding, alignment, political support, amendments adopted. Partly within your influence, harder to measure, much stronger evidence that something is working.

- Strategic outcomes — legislative change, regulatory decisions, business impact. Largely outside your control, hardest to attribute, but the only things that ultimately matter.

Nobody serious claims a single briefing caused a regulatory decision. The honest claim is that your activities plausibly contributed to observable shifts — and the layered system lets you trace that contribution rather than assert it. The art is connecting process to outcome with a credible theory, not a spurious ratio of office visits to favourable votes.

The objective paradox

Even once training, culture and causality are sorted, a structural tension remains — the one that quietly breaks most systems. Objectives must be stable enough to drive coordinated action but flexible enough to reflect a changing reality. Too rigid, and the team keeps marching towards objectives that no longer make sense, which everyone quietly ignores. Too flexible, and direction evaporates and nothing compounds.

The mistake almost everyone makes is treating "stability versus adaptation" as a single dial to set somewhere in the middle. It isn't — it is better understood as different things changing at different speeds.

Different layers, different clocks

The view across strategy and project management is surprisingly consistent: strategy should be relatively stable, objectives should be periodically reviewed, and tactics should be continuously adapted. Think in three layers, each with its own clock:

Layer	How often it should change
Vision	Rarely
Objectives / Milestones	3-6 months
Tactics / Activities	Monthly

The classic error is changing all three at once. When the environment shifts — a new government, a stalled file, an unexpected coalition — the instinct is to declare that "the strategy has changed." Usually it hasn't. What has changed is what you need to do: different tactics, not a different strategy. Conflate the two and you get whiplash, teams lurching between grand redirections that never give any single approach long enough to work.

John Boyd's OODA loop — Observe, Orient, Decide, Act — explains why the layered approach works. Functions don't set objectives once and then execute blindly; they run continuous cycles of observing, orienting, deciding and acting, then observing again. Seen this way, objectives are not predictions. They are hypotheses about how the world works. "If we build this coalition and shift these stakeholders, political momentum will move in our favour" is a testable claim, not a guarantee — and measurement is simply how you test it.

That reframes the whole exercise. Objective-setting stops being about forecasting the future accurately — an impossible standard in public affairs — and becomes structured adaptation: making your assumptions explicit, then updating them as evidence arrives. Most mature functions converge on a rhythm: set strategic objectives annually; quarterly, review the assumptions underneath them and ask whether it is still the right objective; monthly, adjust tactics. The objective is not rewritten every month — the assumptions beneath it are what get tested. That cadence buys enough stability for cumulative effort while staying responsive to a moving environment.

What changes when you hold it this way

Pull the five threads together and they point back to where we started: objectives and KPIs exist to improve judgement, not to prove performance. Objectives are explicit hypotheses about how influence works in your world; measurement is how you test them against reality. Hold the system that way and the everyday mechanics change: KPIs become learning tools rather than instruments of blame, reviews become conversations about assumptions rather than tribunals about who missed a number, and a revised objective reads as maturity rather than failure.

This isn't a public affairs idiosyncrasy — it is where strategy, military planning and performance management have all independently landed under uncertainty. But it speaks with unusual force to a function where influence is real, the stakes are high, and causality is almost never clean. And it matters more now that dashboards and AI make numbers trivially easy to produce — and a tidy report trivially easy to mistake for a working strategy.

So the question to take back to your own function is not whether your dashboard is clean. It is this: when you last missed a target, did the conversation that followed change anyone's mind about what to do next? If it did, measurement is doing its real job. If it didn't, no number of new metrics will fix what is actually a question of how you use them.