What to Choose So You Don’t Drown in Spreadsheets
TL;DR (no extra philosophy)
- North Star: one primary metric that captures customer value and business growth.
- KPI: “business health” and execution metrics. Not dreams, control.
- OKR: a change tool. Used when you need to move the system into a new state.
Problems start when:
- KPIs try to replace strategy,
- OKRs turn into a task list,
- the North Star is chosen for how “nice it sounds” instead of being tied to real value or revenue.
1) North Star Metric: one number so you don’t spread everywhere
The North Star answers:
“What measurable result shows the customer gets value and the business grows?”
It should be:
- about value, not activity,
- convertible to money (not always directly, but logically),
- stable for at least a year.
Examples (so it’s concrete):
- Marketplace: successful orders per week
- SaaS: active teams achieving an outcome (e.g., completing X meaningful actions)
- Fintech/payments: number of active users who completed a successful transaction, or processed payment volume with healthy margin (do not pick “volume” if it’s unprofitable)
Common mistake: choosing a vanity metric (signups, MAU with no meaning, pageviews).
That’s like measuring a restaurant by how many people sniffed the menu.
2) KPI: the business dashboard
KPIs are control metrics: what works and what breaks.
They answer: “Are we running the process, or is chaos running us?”
Typical KPI groups:
- Financial: revenue, margin, CAC, LTV, burn, runway
- Product: retention, conversion, activation
- Operational: support SLA, incident resolution time, quality, backlog
- Risk/compliance (especially in fintech): alerts, fraud, chargeback rate, KYC time-to-approve
Survival rule:
8–12 solid KPIs beat 80 “for show.”
3) OKR: when you need to change the system, not just “keep working”
OKRs are useful when you want to:
- improve the funnel,
- enter a new segment,
- speed up delivery,
- reduce costs,
- improve quality/reliability,
- rebuild processes.
OKR = Objective (where) + Key Results (how we know we got there).
A good OKR example:
- Objective: Make onboarding faster and cleaner
- KR1: reduce time to first successful action from 7 days to 2
- KR2: increase activation rate from 18% to 28%
- KR3: reduce “I don’t know how to start” tickets by 30%
A bad OKR looks like:
- “Build a new landing page”
- “Launch an integration”
Those are tasks, not outcomes. That’s a to-do list wearing a strategy costume.
What to choose: a practical setup that almost always works
If you’re a small team (up to 20–30 people)
- 1 North Star
- 8–12 KPIs (health dashboard)
- 1 set of quarterly OKRs (2–3 objectives, 2–4 KRs each)
That’s it. If you add more, you start serving metrics instead of the business.
If you’re very early (pre-PMF)
- North Star is often temporarily replaced by activation or retention (the most honest metric)
- KPIs minimum: runway, CAC (even rough), funnel conversion
- OKRs: max 1–2 objectives per quarter
If you’re scaling
- North Star stays stable
- KPIs split by function (product, sales, ops, risk)
- OKRs fix bottlenecks, while KPIs make sure you don’t break the base
How not to drown in spreadsheets: 7 rules
- One main metric (North Star). Not three.
- KPIs are health. OKRs are change. Don’t mix them.
- OKRs must not be tasks. Only measurable outcomes.
- Don’t set OKRs for things you must do anyway (e.g., “submit reports”). That’s KPI territory.
- Every metric needs one owner, otherwise it’s a “nobody metric”.
- Cut metrics monthly: remove everything extra.
- If a metric doesn’t influence decisions, it’s trash.
Mini conclusion
- North Star keeps focus.
- KPIs keep the business sane.
- OKRs push the business forward through change.
If your metric system doesn’t help you make decisions and say “no” to distractions, it’s not a system. It’s spreadsheet worship.
FAQ
Can we live without OKRs?
Yes, if you’re not trying to change the system and KPIs are enough. But once growth, chaos, or a new strategy shows up, OKRs help you stay coherent.
Does the North Star have to be revenue?
Not directly, but it must be logically linked to revenue and customer value. “MAU” without quality and monetization often leads nowhere.
How many KPIs is normal?
Usually 8–12 for the company. More starts to get in the way. Teams can have their own, but without fanaticism.
Why do OKRs fail for most teams?
Because they get turned into a task list instead of measurable outcomes. And because nobody wants to do the painful “stop-doing list”.
What if we’re fintech/compliance-heavy and have lots of mandatory metrics?
Mandatory metrics are KPIs. Use OKRs for improvements: faster onboarding, lower fraud, better monitoring quality, fewer manual operations. The key is not to turn compliance into theatre. It should protect the business.
