Systems evaluation
What each platform can and cannot do.
This covers what Ninety's API exposes, where the gaps are, and what those gaps mean in practice. It's meant as input for your evaluation. Two areas come up, and they're independent of each other, so they can be weighed separately.
Access comes first
Ninety's public API is available only on the Thrive plan. Free, Essentials, and Accelerate have no API access.
It's also in limited beta, with general availability targeted for summer 2026. That target is now, so it may be open already or still pending for your account.
The two areas
Different problems, different systems. Neither one depends on the other.
Automation inside Ninety
The API can create To-Dos and Issues, and write numbers into the Scorecard. That covers some of the manual entry around the weekly cycle, within the limits on the next tab.
Depends on: Thrive plan and beta access.
SOPs and knowledge base
Ninety's Knowledge Portal has no API and no announced roadmap for one, so SOPs are created and maintained by hand on every plan. A general-purpose tool would change that, with its own tradeoffs.
Depends on: nothing from Ninety. Fully independent.
Why these are separate. SOPs sit outside the API on every plan, so nothing about Thrive or the beta changes how they work. The two can be weighed in any order, or independently.
What the API doesn't reach
These tools have no API coverage, so they work by hand regardless of plan.
The Level 10 meeting
The meeting itself, its agenda, and the IDS flow aren't exposed by the API. Nothing can read from or write to them.
Rocks & the Scorecard
Both are in the API. Rocks support full read and write. Scorecard values can be written but not read back out.
V/TO & Accountability Chart
No API coverage. Along with Headlines and 1-on-1s, these stay manual on every plan.
Context
How EOS works, and why the limits matter
EOS is a closed system where the constraints are the mechanism. Since the API can write to several of the objects those constraints govern, the rules are worth having in view when weighing what to automate.
The objects and their rules
Rocks
The 3–7 most important priorities for the next 90 days, each with exactly one owner.
The rule: set at the Quarterly, and they don't change mid-quarter. At quarter end a Rock is done or it isn't. No partial credit. Milestones, the sub-steps inside a Rock, are what move week to week.
Scorecard
5–15 weekly measurables, each with an owner and a numeric goal. Reviewed in under five minutes.
The rule: the Scorecard works like a tripwire. An off-track number doesn't get discussed there. It drops to the Issues List and gets solved in IDS.
Issues & IDS
The running list of what needs solving, worked through as Identify, Discuss, Solve.
The rule: prioritize the top three, then solve one at a time. Identify means getting to the root cause. Solving produces To-Dos.
To-Dos
Seven-day action items coming out of the weekly meeting, with an owner and a due date.
The rule: weekly commitments only, measured against a 90%+ completion rate. If it can't be done in seven days, it belongs somewhere else. A Rock or an Issue.
The weekly Level 10
Ninety minutes, same time, same agenda, hard-timed.
| Segment | Time | What happens |
|---|---|---|
| Segue | 5 min | Good news, personal and professional |
| Scorecard | 5 min | Scan for off-track numbers → drop to Issues |
| Rock Review | 5 min | On-track or off-track only → off-track drops to Issues |
| Headlines | 5 min | Customer and employee bulletins |
| To-Do List | 5 min | Done or not done → not-done drops to Issues |
| IDS | 60 min | Prioritize the top three, solve one at a time |
| Conclude | 5 min | Recap To-Dos, cascading messages, rate the meeting |
Why this shapes the tooling. The first five segments are reporting. 25 minutes total. Everything that needs conversation drops into the 60-minute IDS block. The meeting is built around those constraints, so tooling that makes it easier to add more Rocks or discuss more items changes how the constraint operates. Whether that tradeoff is worth making is a judgment for you and your implementer.
Capability
What Ninety's API can and cannot do
The API covers six resources. The gaps matter as much as the coverage, because they set a hard ceiling on what any automation can promise.
| Resource | Support | Notes |
|---|---|---|
| To-Dos | Full | The most automation-ready object |
| Issues | Full | Also strong |
| Rocks | Full | Available, but should be used sparingly |
| Milestones | No delete | Sub-steps within a Rock |
| Scorecard | Write only | Values cannot be read back out |
| Teams | Read | Context and lookup |
The four limitations that matter
The Scorecard can't be read
Numbers can be pushed in and never pulled out. External dashboards and reporting on Scorecard data aren't possible, so Ninety stays the only place those numbers get seen. For pushing data in, this one doesn't bite at all.
Items can't be assigned to others
Anything created through the API is owned by the account that created it. So your auto-created To-Dos land in one place and get reassigned by hand. Two minutes a week.
No event notifications
Nothing reacts the moment something changes in Ninety. Automations run on a schedule instead. For a weekly rhythm that's fine, and a weekly rhythm is most of EOS.
Several tools aren't covered
The V/TO, Accountability Chart, Headlines, meetings, 1-on-1s, and the Knowledge Portal all sit outside the API. Most of them don't need to be in it. The Knowledge Portal is the exception that matters.
What that means for SOPs. The Knowledge Portal isn't in the API, so SOPs get entered into Ninety by hand no matter which plan you're on, and there's no announced roadmap changing it. That's the whole basis for looking at a separate knowledge base, and it's why the two decisions on this page are independent.
Scope
What the API makes possible
Given the six resources and the four gaps, these are the things the API can support. Each carries the constraints from the previous tab.
Technically supported
Each one carries its own constraint. Listed without ranking.
Meeting notes into To-Dos and Issues
Commitments and problems get extracted after a meeting and written into Ninety. Subject to the assignment gap, so items need reassigning by hand or per-person tokens.
Scorecard numbers pushed automatically
Metrics flow from wherever they already live on a schedule. The read limitation doesn't apply since this only writes. Requires knowing which system holds those numbers today.
Milestone setup at quarter start
Standard sub-step structure gets created programmatically when Rocks are set. There's no milestone delete endpoint, so corrections are manual.
Issue intake from the wider team
A form or email address that creates Issues on the right team's list. Same assignment constraint as the first item.
One thing worth weighing on the EOS side
You know EOS better than I do, so take these as questions rather than claims.
- Rocks are writable through the API. Since Rocks are set quarterly and hold for 90 days, automated creation or mid-quarter edits would change how that container behaves. Whether that matters is your call and your implementer's.
- To-Do and Issue status can be set programmatically. The completion rate measures the team, so an automated signal would be measuring something different than it does today.
- Making Rock creation easier is possible. EOS caps Rocks at three to seven per person. Lowering the friction on adding them tends to push against that ceiling.
- Task management placement. EOS To-Dos are seven-day commitments from the L10. If general company work shares that list, the completion rate covers a different population than before.
Next
What would sharpen this
These are the things that would sharpen the picture. Answers here determine what's actually available and what it would be worth.
Questions
- Which Ninety plan are you on? The API requires Thrive. Free, Essentials, and Accelerate have no access.
- If not Thrive, what's the seat count and price difference? Needed to weigh a plan change against what the automation would save.
- Is there an active EOS implementer, and how involved are they? Relevant to anything that changes where work is tracked.
- Where does the SOP burden actually fall? How many, how often they change, how many read versus write. This sizes area two.
- Where do the Scorecard numbers come from today? Which system, who enters them, how long entry takes each week.
- Is general task management a real need? Or is Ninety's To-Do list working because it's only holding weekly meeting commitments?
On Notion costs. Paid Notion plans include unlimited guests as of March 2026, and guests don't consume paid seats, so read-only readers can be free. Full Notion AI requires the Business plan at $20 per user per month billed annually, and AI features don't extend to guests. A large read-only audience is inexpensive. Company-wide AI search is not.
About the source of this
- Ninety plan and whether API access is active on the account
- Current general-availability status of the beta
- Exact endpoints, field names, and payload shapes, confirmed against the live reference
- How ownership actually behaves on created items
The detail on this page comes from Ninety's published API documentation. The full technical reference sits behind authentication and needs an active Thrive account to open, so the items above should be confirmed against it before anyone commits to a build or an estimate. Notion pricing is current as of July 2026.