10 OKR Pitfalls in City Government and How Software Helps

What three years of platform data from 150+ cities and counties says about why municipal OKRs stall — and the four pitfalls no software can fix.

Open the City of Bartlett, Tennessee’s performance dashboard right now and you can watch a city govern in public. Its goals sit on a live, citizen-facing page — economic vitality, infrastructure, public safety — each carrying a status anyone can check.

Bartlett built that plan, “Bartlett Vision 2030,” in 13 months, then ran it on weekly director meetings and quarterly reviews. The work wasn’t the software. “The ClearPoint Strategy team pushed us,” the city’s Chief Administration Officer, Steve Sones, told us. “They’re setting dates and goals for you, just like you are with your team, and that’s what brought it together.”

That’s what good looks like. It’s also rare.

Across the 150+ cities and counties on our platform, just 6% of objectives are given a live performance status. Only four give one to even half their goals. These aren’t cities that skipped the software — they’re running on it. The tool didn’t get them to 6%. Discipline did, and most places don’t have it yet.

One hundred ten of these governments sit under ten percent. Thirty-nine score nothing at all. Four clear half. That’s not a bell curve. It’s a cliff — and almost everyone is at the bottom of it.

So the gap isn’t really the platform. It’s ten specific places where OKRs meet the machinery of city hall and stall. Some are software problems. Some aren’t. Four of the ten below, no platform can fix for you — we’ll flag them as we go.

First, a translation — because OKRs arrive in startup language and a city doesn’t run on startup language.

• Objective = a council priority (“a safer, more connected downtown”).
• Key Result = a performance measure with a target (“cut average EMS response to 7 minutes”).
• Initiative = a project that moves the measure (the new fire substation).

Three layers, plus the five-year plan above them and the budget beside them. Any tool that manages only the middle layer leaves most cities exactly where they started — at 6%. Keep that in mind as we count down. The ten sort into the three things a real system has to restore: alignment across departments, accountability for the numbers, and focus on the few things that matter.

1. The plan resets with every election

Private companies keep their goals when the CEO changes. Cities don’t. Priorities are set by people on a two-to-four-year clock, and a new majority usually arrives with its own agenda. Some cities formally rewrite the strategic plan at the start of each term. So the OKRs tied to the last regime’s plan don’t get improved — they get orphaned. Staff quietly stop updating a dashboard that just turned political.

The career city manager lives in the middle of this. Accountable for outcomes across leadership they didn’t choose and can’t control.

A plan of record is the fix here, not a feature. When OKRs live in one director’s spreadsheet, they leave when that director does. Anchored to a council-adopted plan inside a shared system — with a revision history of who changed what, and when — the plan and its memory survive the transition. The software preserves the history; it can’t preserve the political will to honor it. But a successor who inherits four years of clean, owned history starts from a real position, not a shrug.

2. Council owns the goal, staff owns the work, no one owns the gap

In a company, one chain of command owns a goal. Government splits it by law. Elected officials set the priorities. Staff deliver them. The two answer to different audiences — voters and the city manager — so the objective has a built-in ownership gap on day one.

You can see the gap in our data, not just in theory. The objective — the council’s layer — is the single most neglected element on the platform. The measures beneath it get more attention. The council’s own goals get the least. That 6% from a moment ago isn’t a measurement problem. It’s an ownership problem, rendered in data.

Linking is what closes the gap. Every measure and project links up to the council-approved objective it serves, and every link carries a named staff owner with a defined slice. Nobody owns “the most livable city.” Someone owns the EMS response time, the paving backlog, the permit turnaround that ladder up to it. Alignment stops being a poster and becomes a structure you can click through.

3. Phantom owners

This is the quiet epidemic. A measure has a name attached. That person has never touched it. Across local governments on our platform, 76% of assigned metric owners have never updated their data — not once. It isn’t laziness. It’s sprawl meeting diffuse civil-service accountability: the median city or county carries 136 active projects across 18 separate plans. Of course half the numbers are stale. There are too many, owned too thinly, with nothing nudging them alive.

This is where a platform earns its keep. Automated reminders reach the actual owner before each cycle. A personal queue shows each person what’s theirs and what’s overdue. Missed-update warnings make a stale measure visible to leadership instead of invisible. Revision history shows who reported and who went quiet.

The candid limit: reminders reduce phantom owners. They don’t delete the human who ignores three of them. The 76% is our own number — proof the tool alone isn’t the cure. Pair it with a standing review where “no update” is a sentence someone says out loud.

4. Nobody trusts the number

Sam Edelstein, the former Chief Data Officer of Syracuse, put it plainly: “It is often hard to count things. The processes we have in government are often complicated.” He found nearly every review opened by arguing whether the data was even right — before anyone could ask what it meant. Legacy systems and a number that lives in three formats will do that.

A review that spends its first twenty minutes litigating the sewer-backup figure isn’t a strategy review. It’s a data-cleaning session with an audience.

Pull the number straight from the system that owns it. A data loader and integrations with the tools a city already runs — Tyler Munis for financials, ArcGIS for anything geographic, Power BI, SQL — push authoritative figures into the measure on a schedule. Status calculates from that one feed. Now the room argues about the decision, not the spreadsheet. One limit worth naming: if the underlying count is genuinely contested — if no source system tracks it — software can’t manufacture truth. It can only stop the copy from drifting from the original.

5. Eighteen plans that never link

Public Works, Parks, Police, and Finance run on separate budgets, separate systems, separate clocks. Information stays trapped where it was born — and we measured how much. Of the 156 local governments on our platform, 134 — 86% — have never linked one element’s progress to another’s. The median city runs its strategy across 18 separate plans; one runs 472. And the walls aren’t only cultural. They’re financial. A city’s departments are siloed by fund — restricted grants, dedicated levies, capital kept apart from operating — so even when two of them chase the same outcome, the money often can’t legally pool. The org chart is a funding chart. You don’t workshop your way out of that.

The cost stays invisible until it isn’t: two departments chasing the same outcome, blind to each other, and a grant one of them could have used sitting undiscovered in another’s inbox.

Linking is the unglamorous fix. Every department’s objectives, measures, and projects sit in one place and link to each other. A link explorer shows where the work connects and where it’s falling behind. Leadership reads the whole board; each department still owns its corner. Shared visibility turns eighteen disconnected plans back into one. It’s the change Bartlett’s Steve Sones described: “Previously, you had everybody doing all these great things, but in their own separate entities … We’re not working in silos.” Same departments. One plan.

6. Nothing on the list is optional

A company picks three goals and drops the rest. A city can’t drop the rest — the rest is the law. A state reporting mandate, a consent decree, a charter requirement, a grant condition: each lands non-negotiable, and each adds to the list. So the list only grows. The median local government on our platform runs 5.14 measures for every objective — and the handful that matter get buried under the many nobody is allowed to cut. The leadership meeting loses its center.

ICMA has said the opposite for years: less, but better — three to five core measures. Almost nobody does it.

Linking and scorecards separate the vital few from the merely tracked. Tag the council’s top objectives. Build the executive view around those alone. Let the other 400 measures live in the system for the departments that need them, off the leadership table. The platform makes the hierarchy visible — but choosing what not to put on the front page is a discipline, not a feature. This is the first one software can’t do for you. It shows you the noise. You turn it down.

7. The audience for activity is voters

In a company, “47 meetings held” dies quietly in a board review. In a city, it plays — in a council session, in a re-election mailer, in a press release. Visible motion carries a political payoff that real outcomes, which take years to show, often don’t. So departments report motion. ICMA warns against exactly this — measures “destined to sit on a shelf,” outputs that prove activity but never change. It shows up in our numbers: a city gives barely a quarter of its measures a live status, and fewer still to the objectives above them. The outcome layer — did anything change for residents? — is the most ignored of all.

A platform lets you attach an outcome measure to every objective and link the activity beneath it, so “meetings held” sits under “jobs created,” not in its place. Analysis can flag when the activity moves and the outcome doesn’t. But the platform tracks whatever you point it at, vanity included. Software can’t choose your metrics. Picking the outcome over the activity is a human call it can support, surface, and nag about — never make. That’s the second one on you.

8. Your OKR clock and your money clock don’t match

OKRs assume you can shift resources when priorities shift. Government budgets don’t work that way. Once funds are appropriated, they’re largely locked for the year — so a sharp pivot found in March may have no money behind it until the next budget, nine months out. Capital plans widen the gap. A capital improvement plan runs five to ten years; a bridge does not move 25% a quarter. Force it into a quarterly key result and it reads 0% for six straight cycles, and leadership hears “dead” when the truth is “permitting.”

Two moves help. Tie goals and projects to their budget lines, so in-year drift shows against the locked appropriation and gets flagged early. And model multi-year work as a multi-year initiative — a Gantt with real milestones — then link the milestones to your annual goals, not the decade-long project. “Design complete,” “permits secured,” “phase one funded” become the quarterly signal. The plain limit: software gives you visibility into the budget calendar. It can’t give you flexibility the law withholds, and it isn’t an appropriations system. This is the third one software can’t fix. The mismatch is structural. Seeing it early is the whole win.

9. A public dashboard nobody decides anything with

Transparency mandates push cities to publish a portal. So they do — then manage the city off a different spreadsheet. The public dashboard becomes a compliance artifact: technically open, functionally ignored. Governing called this the “digital moat” — data that’s available without being usable. Worse, when poor numbers go public with no plan attached, the temptation is to measure less, not manage better.

Wire the public view to the internal one. The same live data that publishes to residents — in plain language, ADA-compliant — generates the council briefing book and the leadership summary. The dashboard stops being a separate chore and becomes the same system of record, aimed at two audiences. Bartlett’s is live right now — the same goals staff manage, published for any resident to check. Just down the road, Germantown, Tennessee publishes its own — the same city that won a 2019 Baldrige Award for this exact kind of rigor. Most cities on our platform stand one up within a few weeks of go-live, because it’s a publish action, not a second project. The limit: software closes the loop from data to the meeting. Whether the meeting makes a decision is still on the people in the room. Call that the fourth one software can’t do for you.

10. GASB 103 turns stale data into a fire drill

This one has a date on it. GASB Statement 103 reshapes the financial reporting model, and it takes effect for fiscal years beginning after June 15, 2025 — first hitting the June 30, 2026 year-end for many governments. The new Management’s Discussion and Analysis demands detailed analyses: not just what changed, but why, in substance, with boilerplate discouraged. If your performance data lives in twelve disconnected spreadsheets, finance spends August emailing every department to reconstruct the “why” behind the variances — under deadline, at the worst possible time.

When goals, measures, and results live centralized and current all year, the operational narrative is already there. Finance pulls structured, audit-trailed context on demand instead of begging for it, and automated reporting turns weeks of assembly into days. The precise version: a strategy platform supplies the performance context for the MD&A. It is not your CAFR system and does not produce the audited financials. But the part GASB 103 newly demands — the substantive “why” — is exactly the part a living performance system already holds.

What software does, and where you’re still on the hook

If you only fix one of these, fix the linking

We went looking in the data for the pitfall that moves the others — the one whose fix drags the rest along. One stood out.

Of the cities that have never linked a single piece of their strategy to another, 73% score none of their objectives. Zero. Of the cities that do link their work, only 14% fall that far. Same platform. More than five times the gap.

That isn’t proof that linking causes accountability — the cities that link are probably more deliberate to begin with. But across 150-plus of them, linking and live objectives travel together so tightly that the recommendation is hard to argue with: if you’re choosing where to start, start where the work connects. It’s the cheap fix that makes the expensive ones — ownership, focus, honest reporting — possible at all.

Back to Bartlett

Bartlett isn’t ahead because of the software. Plenty of cities run the same platform and sit at 6%. Bartlett is ahead because someone decided the plan would be small enough to mean something, public enough to be checked, and reviewed often enough that “no update” became a question with a name attached. The tool made that durable. People made it real.

In a city, an OKR isn’t a productivity hack. It’s a promise — made in public, to people who voted, inherited by whoever’s in the room after the next election. The fix was never “run OKRs harder.” It’s to stop treating objectives, measures, projects, the five-year plan, and the budget as five separate things in five separate tools. Hold them as one. Let the structure outlast the people. Let the dashboard residents see be the one the council decides from.

The software won’t keep the promise for you. It makes sure that, four years from now, someone can still tell whether you kept it.

📘 Free ebook — The City Hall OKR Playbook: the 10 pitfalls, the fix for each, and a 12-point self-scan to find your phantom owners, unlinked plans, and unscored objectives before your next council meeting. Get the playbook →

Methodology: platform figures are drawn from ClearPoint’s live local-government plans — 150+ U.S. cities and counties, refreshed June 2026. “Live performance status” means an objective or measure has an active evaluation rule; “never updated” means an assigned owner has logged zero updates. City of Bartlett details come from ClearPoint’s published case study and the city’s public dashboard.

Frequently asked questions

Why do city departments struggle to manage OKRs effectively?

Because OKRs were built for one chain of command and a flexible budget, and a city has neither. Priorities are set by elected officials on a short clock, money is locked annually, departments run on separate systems, and accountability is bounded by civil-service rules. The data shows where it lands: across 150+ cities and counties on our platform, only 6% of objectives are given a live performance status, and 76% of metric owners have never updated their data. The framework isn’t the problem. The fit is.

How do OKR management tools help governments align department goals?

By linking every departmental measure and project up to a council-approved objective, so alignment is a structure instead of a slogan. Each link has a named owner and a defined slice. Leadership sees the rolled-up objective; each department sees its corner; the cross-department work that used to hide in silos becomes visible on one board. The honest limit: the tool can show the gap, but only a manager assigns the owner who closes it.

What causes public sector teams to lose focus on OKR priorities?

Volume and obligation. A city can’t pick three goals — it’s legally responsible for dozens at once, and the measure list swells to match (the median local government tracks over five measures for every objective). When everything is tracked, the leadership meeting loses its center. Focus returns only when someone separates the council’s vital few from the operational many — a human call software can support but not make.

How do OKR management platforms affect accountability across government departments?

They make ownership visible and updates traceable: reminders reach the real owner, a revision history records who reported and who went quiet, and missed-update warnings surface stale measures. The caveat that matters: software reduces phantom owners but can’t erase them — civil-service rules cap individual consequences, so the durable form of accountability in government is visibility (to peers, council, and the public), not punishment.

Which OKR management tools work best for government departments?

Look for four things a generic corporate tool won’t have: a council-adopted plan of record with a full audit trail, granular role-based access (department, council, and public views), citizen-facing dashboards, and reporting that supports GASB and FOIA. Federal teams may need FedRAMP; for a city or county, SOC 2, a FOIA-defensible change history, and GASB-aligned reporting matter far more. For a side-by-side of the platforms built for this, see our guide to the best OKR software for public sector departments.