Early access is open to founding teams. Free for 60 days.Apply →

OKRs and Accountability: Beyond Goal Tracking

Lead

September 10, 2026
LMBy Lazaro Medina··5 min read
OKRs and Accountability: Beyond Goal Tracking

OKRs fail less often because of the framework and more because nobody manages the work behind them. Key results describe outcomes. They do not say who will do what by when, and that missing layer is where most OKR programs quietly stall.

TL;DR: Give every key result one owner, link it to the initiatives meant to move it, review it weekly on evidence instead of confidence scores, and keep the list short. That turns OKRs from a quarterly ritual into a plan you can manage.

Why do OKR programs stall?

The arc is familiar: excitement in Q1, compliance in Q2, quiet abandonment by Q4. Five patterns cause most of it.

1. Objectives without owners

Shared ownership means nobody feels on the hook. When a key result slips, the conversation is about the team instead of a decision by a person.

2. Quarter-end scoring

Set in week 1, scored in week 13, a black box in between. By the time you see 20%, there is no time left.

3. Key results without initiatives

"Reduce churn to 4%" says nothing about how. Without links to initiatives, you cannot tell whether the plan is working or the metric moved for other reasons.

4. Confidence scores based on feelings

An owner who is 70% confident on Monday is often 70% confident every Monday, until the quarter ends at 30%.

5. Too many OKRs

Five objectives with four key results each, per team, means hundreds of goals. Focus, the whole point, disappears.

What is the missing layer?

Commitments. Think of execution in three layers:

  • Strategy: where we are going and why.
  • Goals (OKRs): how we will know we got there.
  • Commitments: who will do what, by when, to move each goal.

Most OKR tools cover the first two. The third is where execution lives, and where the strategy execution gap opens.

How do you make OKRs stick?

  • One owner per key result. They do not do all the work, but they own the outcome and raise the flag.
  • Link key results to initiatives. Then the review asks the right question: are the initiatives on track, and are they moving the metric?
  • Review weekly, on evidence. Milestones hit, dependencies cleared, early metric movement. Fold it into your existing operating cadence instead of adding a meeting.
  • Keep the list short. Three objectives, three key results each.
  • Check what worked. At quarter end, ask which initiatives moved results and which only consumed effort.

What does it look like in practice?

  • Objective: help new customers reach value sooner.
  • Key result: raise 14-day activation from 45% to 60% this quarter.
  • Initiative: simpler onboarding plus a guided first session.
  • Owner: the customer-success lead.
  • Commitment: pilot the new onboarding with the next 20 customers by month end.

The guide ships on time, but only half the pilot customers attend their session. The owner flags the initiative as at risk, finds the attendance barrier and proposes a fix. The weekly review assigns a new invitation process and a date to check it. Shipping the guide proved a deliverable exists. It did not prove customers activate faster.

What should you ask in the weekly review?

  • What evidence changed since last week?
  • Is that evidence about activity, outcomes or both?
  • Which dependency threatens the target?
  • Who owns the next step, and when do we check its effect?

Where does stratbox fit?

OKR software is good at recording goals. stratbox adds the layer underneath: each goal tied to an owner and the procedures that deliver it, checked against evidence from the tools your teams already use, with a flag when something will slip. Use it alongside OKRs or your own planning framework. Read What Is Accountability Intelligence? for the model.

Frequently asked questions

Why do OKRs fail?

Usually not because of the framework. Key results lack a single owner, they are not linked to the initiatives that should move them, and progress is self-reported and only scored at quarter end.

Who should own a key result?

One person. A team can share an objective, but each key result needs a single name who raises the flag when it is at risk.

How often should OKRs be reviewed?

Weekly, on evidence: milestones hit, dependencies cleared, early metric movement. End-of-quarter scoring alone finds problems too late.

How many OKRs should a team have?

Three objectives with about three key results each is a healthy ceiling. More than that and focus disappears.

Free scorecard
Where is your strategy leaking?

Take the Strategy Execution Gap Scorecard to see where your strategy leaks and what to fix first.

Get your score →
Ask AI about this article:ChatGPTClaudePerplexity
LM

Lazaro Medina

Co-founder and CEO, stratbox

Lazaro started his career as an enterprise systems consultant, building integrations that connect HR and finance platforms to the rest of the business. Today he runs stratbox and also helps run Shopferro, a small food brand selling iron-pressed Spanish cookies. He has lived and worked in Germany and the US, and is based in Miami.

Follow Lazaro on LinkedIn ↗

Keep reading

What Happens When Projects Go Off Track?
Browse guides

What Happens When Projects Go Off Track?

What really happens when projects go off track, from 40+ enterprise implementations: how the slip starts, what it costs, and what to do in week one.

By Lazaro Medina

5 min read

I've worked on somewhere between 40 and 50 Workday implementations over the last nine or ten years. I can count on one hand the ones that finished on time and on budget.

That isn't because the people were bad at their jobs. Most of them were smart, committed and genuinely wanted the project to work. Projects go off track for quieter reasons, and almost always earlier than anyone admits.

Here's what it actually looks like, using one project I'll never forget.

It starts with a great kickoff

We were implementing Workday at a large university on the West Coast. The CIO had brought Workday into his previous organization, so he wasn't just supportive, he was a champion. The kickoff was energizing. We walked through the timeline, set the rules of play and agreed on a vision: whenever there was a conflict or a question about which approach to take, we'd go back to that vision and the priorities everyone signed up for.

Everything looked right.

The first warning sign showed up in week two

The week after kickoff, I could already see it wasn't going to be smooth.

Nobody had addressed backfill.

The university had bought a platform worth millions a year, and the implementation itself was worth millions more. But the most critical piece, their people, hadn't been planned for. The same staff who already had full-time day jobs were now expected to work with consultants almost full time on mission-critical tasks: gathering documents, filling data templates, answering configuration questions and shaping the system around how the institution really works.

No one had asked the obvious question: when are they supposed to do all of this?

How the slip actually happens

The first missed deadlines were small. A template a day or two late. A document that didn't arrive. But in an implementation, a day or two late on an input is never just a day or two late. Consultants can't move forward without it, and every downstream milestone, smoke testing, end-to-end testing, payroll parallel, starts to move with it.

Leadership found out quickly, which surprised no one. We ran status calls with the project manager and every division, walking through what was essentially a souped-up spreadsheet: the item, the owner, the due date. Most items were either not done or stuck on an open question that had gotten lost somewhere across email and a dozen other tools. Maybe the message really was missed. Maybe the person just never had time to get to it. Either way, it was late.

The response was the one I've heard on almost every project: "Okay, just get it to us by this new date and we'll catch up."

That's exactly where the real damage begins. The person now spends their limited time catching up on the old task, which means the next task, the one that should have had their attention, loses priority too. One late item quietly steals time from the next. It becomes a chain reaction, and by week two we were already behind in a way that's very hard to recover from.

The cost: three delayed go-lives

Payroll made it worse. You can't go live on payroll whenever you like. Most organizations aim for specific cycles, often the end of December so the first January paycheck runs on the new system. Miss the window and you wait for the next one.

This university pushed its go-live three times.

Every delay meant running two systems at once: keeping an old, heavily customized PeopleSoft environment alive while still building Workday. Double the effort, from the same exhausted people.

And there was a subtler cost. Payroll needs a lot of configuration and a lot of decisions, and the processes underneath weren't clearly defined. Without a clear picture of the goal and the procedures that support it, every configuration session turned into a menu of possibilities. When nobody can point to "this is how we want it to work," teams default to how they've always worked. Over-customization creeps in, and the new platform slowly gets bent back into the shape of the old one. The whole point of moving to a new system is to leave old habits behind, not rebuild them somewhere more expensive.

The human cost nobody puts in the budget

This is the part that rarely shows up in a project post-mortem. I've seen people cry on these projects. I've seen people shout, and I've seen people quit between phases. When you ask someone to do two full-time jobs for months, morale doesn't just dip. It breaks.

And backfill, the thing that could have prevented a lot of it, often never even appears on the risk register. Consulting firms should be explaining this upfront. Some do. Many don't, maybe because it adds cost to the deal, maybe because it's an uncomfortable conversation. But staying quiet about it doesn't make the cost go away. It just moves it onto your people.

The fix everyone reaches for, and why it doesn't work

When a project slips, the instinct is to add more check-ins. More status calls, more touchpoints, sometimes several in the same day.

In my experience this mostly creates anxiety. The honest answer people give on those calls is: "Sorry, between this meeting and the last one I didn't have time to actually do anything." More meetings don't create progress. They consume the very time people needed to make progress.

What I'd do differently in week one

If I could go back to the start of that project, here's what I'd put in place.

  • Define what "finished" really looks like. Not "Workday is live," but what life looks like once it's working. For a finance team, maybe that means closing the books in an hour instead of one or two weeks. A concrete finish line gives every later decision something to be measured against.
  • Break the finish line into components, and put a name on each one. What has to be true to get there? Who owns each piece? Granular ownership is what turns a vision into work people can actually do.
  • Document the new way of working from day one. People will leave, change roles or move to the next project. If the new methodology is written down as it's built, it doesn't walk out the door with the people who designed it.
  • Actively hunt for what you don't know. Most project risk isn't hidden because people are careless. It's hidden because nobody thought to ask. Backfill is the perfect example. Ask early and often: "Given what we're trying to do, what are we missing that could hurt us?"
  • Keep status and risk in one place. When the goal, the owners, the procedures and the current status live in one hub instead of across emails, spreadsheets and chat threads, problems surface on day two instead of week two. People with day jobs can see exactly what's needed from them and get back to work.

The bottom line

Projects rarely blow up in one dramatic moment. They drift, one late template and one "we'll catch up" at a time, while everyone is working hard. By the time the slip is obvious, you're already paying for it in delayed go-lives, duplicate systems and burnt-out people.

The projects that stay on track aren't the ones with the most meetings. They're the ones that know exactly what finished looks like, know who owns each piece, write down how the work gets done, and look for risk before it finds them.

That gap between the plan and what's actually happening is the problem we started stratbox to solve. But whatever tools you use, start with the questions above. Your people will thank you.

What Is Accountability Intelligence?
Browse guides

What Is Accountability Intelligence?

Accountability intelligence ties goals to owners, checks progress against real evidence and recommends the next move. What it is and how it works.

By Lazaro Medina

5 min read

Accountability intelligence is the ability to connect strategic priorities to the commitments that deliver them, check those commitments against real evidence, and recommend what to do when something drifts. It answers three questions dashboards cannot: who committed to what, is it on track, and what should we do now?

TL;DR: BI tells you what happened. Task tools tell you what got done. OKR tools tell you what people intended. Accountability intelligence connects all three to a named owner and real evidence, so you see risk early and know who acts.

What is accountability intelligence?

It sits between strategy and execution, exactly where most companies lose sight of things. Every commitment traces back to the priority it serves, progress is judged on evidence instead of a status field, and each risk comes with an owner and a next step.

How is it different from the tools you already have?

Business intelligence

BI is descriptive. It tells you revenue missed target. It does not tell you which commitments made in January would have prevented it.

Project and work management tools

Task tools track activity. They know a ticket closed, not whether closing it moved a priority. Thousands of finished tasks can sit next to a failing strategy.

OKR and strategy software

Goal tools record intent. They rely on owners to self-report, which makes status late and optimistic. More on that in OKRs and Accountability.

Accountability intelligence

It links strategy to owned commitments, reads the evidence from the work, spots risk early and tells leaders where to step in.

What are the four capabilities?

  • Traceability. Every commitment links to the priority it serves. "What are we doing about enterprise expansion?" gets a live answer, not a week of slide-building.
  • Evidence-based status. Progress comes from milestones, dependencies, metrics and activity, not a status someone updated two weeks ago. This is how watermelon reports end.
  • Early risk detection. Slipping dependencies and stalled owners show up on a Tuesday, not at the monthly review.
  • Recommended actions. Each risk comes with a next step: reassign, escalate, re-scope or move resources.

How does it work, step by step?

Here is how it works on one goal:

  1. Set the outcome. More customers finish onboarding within 14 days. Name one owner and record the target.
  2. Connect the work. Link the onboarding guide, training sessions and sales handoff to that outcome, each with an owner and a date.
  3. Check the evidence. Compare finished deliverables with activation data. A finished guide can sit next to weak activation.
  4. Find the risk. Training happens after customers first try the product. The problem is timing, not a missing document.
  5. Act. Move the session earlier, assign the customer-success lead and set a date to check the next cohort.
  6. Confirm the effect. Compare outcomes, rule out other causes, then decide whether the change worked.

A good system also separates missing evidence from a confirmed failure, so leaders can ask questions without treating every gap as poor performance.

Why does it matter now?

Boards increasingly expect proof that strategy is being executed, not just a strategy deck. And AI can now do what used to take a team of analysts: read signals across tools, spot patterns and suggest the next move. Pointed at the strategy execution gap, that is accountability intelligence.

How does stratbox do it?

stratbox turns your vision into goals, ties each goal to a named owner and the procedures (SOPs) that deliver it, and turns those into short recurring assessments. It reads the evidence from the tools your teams already use and flags what will slip while there is still time. If you are building the review rhythm from scratch, start with the operating cadence guide.

stratbox strategy tree connecting a goal to the work and assessments behind it

Frequently asked questions

What is accountability intelligence?

It is the ability to connect strategic priorities to the commitments that deliver them, check those commitments against real evidence, and recommend the next action when something drifts.

How is accountability intelligence different from business intelligence?

Business intelligence describes results after they happen. Accountability intelligence follows the commitments that produce the results, so you can act before the quarter is lost.

Is accountability intelligence the same as OKR software?

No. OKR tools record goals and self-reported confidence. Accountability intelligence links goals to owners and procedures and checks them against evidence from the work itself.

Who uses accountability intelligence?

Chiefs of Staff, Strategy Ops leaders, PMO Directors and COOs who run the operating rhythm and need to know what is slipping before the CEO asks.

OKRs and Accountability: Beyond Goal Tracking
Browse guides

OKRs and Accountability: Beyond Goal Tracking

Why OKR programs stall by mid-year, the missing accountability layer, and how weekly evidence-based reviews keep key results moving. With a worked example.

By Lazaro Medina

5 min read

OKRs fail less often because of the framework and more because nobody manages the work behind them. Key results describe outcomes. They do not say who will do what by when, and that missing layer is where most OKR programs quietly stall.

TL;DR: Give every key result one owner, link it to the initiatives meant to move it, review it weekly on evidence instead of confidence scores, and keep the list short. That turns OKRs from a quarterly ritual into a plan you can manage.

Why do OKR programs stall?

The arc is familiar: excitement in Q1, compliance in Q2, quiet abandonment by Q4. Five patterns cause most of it.

1. Objectives without owners

Shared ownership means nobody feels on the hook. When a key result slips, the conversation is about the team instead of a decision by a person.

2. Quarter-end scoring

Set in week 1, scored in week 13, a black box in between. By the time you see 20%, there is no time left.

3. Key results without initiatives

"Reduce churn to 4%" says nothing about how. Without links to initiatives, you cannot tell whether the plan is working or the metric moved for other reasons.

4. Confidence scores based on feelings

An owner who is 70% confident on Monday is often 70% confident every Monday, until the quarter ends at 30%.

5. Too many OKRs

Five objectives with four key results each, per team, means hundreds of goals. Focus, the whole point, disappears.

What is the missing layer?

Commitments. Think of execution in three layers:

  • Strategy: where we are going and why.
  • Goals (OKRs): how we will know we got there.
  • Commitments: who will do what, by when, to move each goal.

Most OKR tools cover the first two. The third is where execution lives, and where the strategy execution gap opens.

How do you make OKRs stick?

  • One owner per key result. They do not do all the work, but they own the outcome and raise the flag.
  • Link key results to initiatives. Then the review asks the right question: are the initiatives on track, and are they moving the metric?
  • Review weekly, on evidence. Milestones hit, dependencies cleared, early metric movement. Fold it into your existing operating cadence instead of adding a meeting.
  • Keep the list short. Three objectives, three key results each.
  • Check what worked. At quarter end, ask which initiatives moved results and which only consumed effort.

What does it look like in practice?

  • Objective: help new customers reach value sooner.
  • Key result: raise 14-day activation from 45% to 60% this quarter.
  • Initiative: simpler onboarding plus a guided first session.
  • Owner: the customer-success lead.
  • Commitment: pilot the new onboarding with the next 20 customers by month end.

The guide ships on time, but only half the pilot customers attend their session. The owner flags the initiative as at risk, finds the attendance barrier and proposes a fix. The weekly review assigns a new invitation process and a date to check it. Shipping the guide proved a deliverable exists. It did not prove customers activate faster.

What should you ask in the weekly review?

  • What evidence changed since last week?
  • Is that evidence about activity, outcomes or both?
  • Which dependency threatens the target?
  • Who owns the next step, and when do we check its effect?

Where does stratbox fit?

OKR software is good at recording goals. stratbox adds the layer underneath: each goal tied to an owner and the procedures that deliver it, checked against evidence from the tools your teams already use, with a flag when something will slip. Use it alongside OKRs or your own planning framework. Read What Is Accountability Intelligence? for the model.

Frequently asked questions

Why do OKRs fail?

Usually not because of the framework. Key results lack a single owner, they are not linked to the initiatives that should move them, and progress is self-reported and only scored at quarter end.

Who should own a key result?

One person. A team can share an objective, but each key result needs a single name who raises the flag when it is at risk.

How often should OKRs be reviewed?

Weekly, on evidence: milestones hit, dependencies cleared, early metric movement. End-of-quarter scoring alone finds problems too late.

How many OKRs should a team have?

Three objectives with about three key results each is a healthy ceiling. More than that and focus disappears.

Operating Cadence: A Practical Guide
Browse guides

Operating Cadence: A Practical Guide

Build an operating cadence that ends in decisions: weekly, monthly and quarterly reviews, sample agendas and a copyable commitment template.

By Lazaro Medina

5 min read

An operating cadence is the set of weekly, monthly and quarterly reviews a company uses to turn strategy into decisions. A good one ends every meeting with a decision. A bad one fills the calendar while execution drifts.

TL;DR: Run three layers: a 30-minute weekly execution review, a monthly operating review and a quarterly strategy review. Give each one a decision to produce, move status to pre-reads, keep one source of truth, log decisions and audit the cadence every quarter.

What is an operating cadence?

It is a decision system, not a meeting schedule. Each review answers a specific question at a specific altitude, and its output feeds the layer above or below. If a meeting does not change what anyone does next, it is not part of your cadence. It is just a meeting.

What are the three layers?

Weekly: execution

Question: are this week's commitments on track, and what is blocked? Keep it to 30 to 45 minutes, review only exceptions, and send every blocker out with an owner and a date.

Monthly: operating review

Question: are our initiatives producing the outcomes we expected? Review results against targets, look for repeated slips, and decide to double down, re-scope or stop.

Quarterly: strategy review

Question: is the strategy still right, and what do we commit to next? Check last quarter's outcomes, adjust priorities on evidence, set owners, and retire work that no longer matters.

What makes a cadence hold?

  • Every meeting ends in a decision. Continue, adjust, escalate or stop. A meeting that never decides gets merged or cut.
  • Pre-reads replace presentations. Status is available before the meeting, so the time goes to the hard conversations.
  • One source of truth. If the weekly uses a spreadsheet, the monthly uses slides and the board pack starts from scratch, the cadence breaks at every handoff.
  • Log decisions, not just actions. A decision log stops the same debate from reopening every month.
  • Audit the cadence quarterly. Meetings pile up like technical debt. Ask of each one: does it still produce decisions?

What does a sample cadence look like?

A starting point for a leadership team. Adjust timings to your size.

Weekly execution review, 30 minutes

Owner: Chief of Staff or operations lead. Pre-read: commitments due, overdue dependencies, changes in outcome metrics.

  • 5 minutes: what changed since last week.
  • 15 minutes: blocked or at-risk commitments only.
  • 10 minutes: agree actions, one owner each, with a date.

Monthly operating review, 60 minutes

Owner: COO or business lead. Compare results with targets, dig into repeated blockers, and decide which initiatives get more resources, a new scope or a stop.

Quarterly strategy review, 90 minutes

Owner: executive sponsor. Review which initiatives changed outcomes, revisit assumptions, and leave with owners, targets and an explicit stop list.

Is there a template I can copy?

Yes. Use one record per commitment, in whatever tool you already use:

  • Strategic priority: the outcome this work supports
  • Commitment and owner: the deliverable, and one accountable person
  • Due date and evidence: the date, and a link to visible progress
  • Dependency and risk: what is needed, from whom, by when
  • Decision: continue, adjust, escalate or stop, and why
  • Next step: owner, date, expected effect

Who owns the cadence?

Usually a Chief of Staff or Strategy Ops leader. They are the architect, not the note-taker: they design the system, protect it, notice drift, and make sure each layer changes the next. A good cadence keeps running when its owner is on vacation.

Where do cadences usually break?

Between meetings. Commitments made on Monday are forgotten by Wednesday and rediscovered as surprises at the monthly review. stratbox keeps those commitments tied to the goal they serve, checks them against evidence from the tools your teams already use, and flags risk before the next review, so meetings are about decisions, not discovery. See OKRs and Accountability for how goals connect to commitments.

Frequently asked questions

What is an operating cadence?

The recurring set of reviews a company uses to run itself, weekly, monthly and quarterly, designed so each one produces a decision that feeds the next.

How long should a weekly execution review be?

About 30 minutes for a leadership team, covering only exceptions: what is blocked, at risk or newly done.

Who owns the operating cadence?

Usually a Chief of Staff or Strategy Ops leader. Their job is to design the system and protect it, not to take notes.

How do you stop meetings from multiplying?

Give every meeting a decision it must produce, move status updates to pre-reads, and audit every recurring meeting once a quarter.

Bring us the hard part

Tell us what keeps slipping.

A launch date that keeps moving. A status report nobody believes. A strategy that never reached the teams doing the work. Bring us the real problem and we will build the setup around it with you.

Start a conversation

hello@stratbox.ai