Metaview's operating principles: the five we ship the product on
Most companies have operating principles. Most do not ship them. They live on a Notion page, get read once during onboarding, and then nobody looks at them again until the next offsite. Ours show up in the product because that is the only test that matters.
Metaview's operating principles are the five sentences that decide how we build, how we hire, and how we handle a Slack thread at 11pm on a Sunday. Each one is polarizing on purpose. If a candidate reads them and feels energized, the company is going to work for them. If they flinch, the company is going to be harder than it needs to be. That filtering is the point.
What follows is the version that ships, not the version that goes on a slide.
What an operating principle actually is
The test for a principle is not whether you can recite it. The test is whether the principle gets used to decide something hard. A principle that does not decide anything is not a principle. It is a value statement, and value statements are how companies justify the decisions they were going to make anyway.
The five below were written down because we needed them to settle disagreements that were going to keep happening. Each one carries a cost. Each one filters out a kind of person who would otherwise be a fine hire at a different company. That is the work the principle is doing, and that is the work it has to keep doing or it stops being load-bearing.
If you cannot point at a decision in the last six months that the principle decided differently than the default would have, retire the principle. It is not pulling its weight.”
- Phrased so blandly nobody could disagree (e.g. "we value excellence"). Filters out nothing.
- Lives on a Notion page. Read once during onboarding. Never referenced in a real decision.
- Justifies the decision the team was going to make anyway. Decoration, not load-bearing.
- Audit at 6 months: zero decisions can be traced back to the principle. The principle is unused.
- Phrased polarizingly. Filters out candidates and partnerships that would not work. The cost is the point.
- Referenced in PR descriptions, retros, hiring debriefs, post-mortems. The principle decides the call.
- Shows up in the product: one workflow per principle, visible to the customer, defended in design review.
- Audit at 6 months: at least one hard decision per quarter went the principle's way against the default.
Own it end-to-end
When you pick up a problem, you do not hand it off. You see it through to a customer feeling the difference. The product manager is in the support thread. The engineer is in the customer call. The CEO is on the second-tier escalation when it matters. The boundary you are tempted to draw, the one that lets you say "that is not my function," is the boundary that produces a company where nothing ships end to end.
The price is that you carry context across functions that you did not used to. The payoff is that the work does not get dropped at the seam between two people who each thought the other had it. You will not like this principle if you are more energized by mastery of a craft than by the impact of the work crossing the finish line.
This is the reason Metaview Notetaker does the scorecard, the debrief brief, and the per-candidate proof inside one workflow. The recruiter is not handing the interview off to a transcription tool, then a scorecard tool, then a debrief tool, with three seam-failures in between. One product carries the interview from the conversation to the decision.
- 1The interview lands here verbatim, with role and round on the header. No handoff to a transcription tool first.
- 2Scorecard fields populate from the Q&A template against the same rubric the recruiter briefed. The principle is the column shape, not a separate tool.
- 3The TLDR is what the hiring manager reads at debrief. One product carries the interview from conversation to decision; no seam-failures between three vendors.
Radical transparency
The default human reflex is to filter. Soften the news, hold the bad number, wait for a better moment. Filtering at scale produces a company where the leadership team thinks everything is fine and the IC layer knows it is not. You cannot fix that with a townhall. You fix it by making the un-filtered version the path of least resistance.
For us that means customers get access to the internal Slack channels where their bugs are triaged. The roadmap is in writing, with the disagreements visible alongside the consensus. The post-mortem names the decision and the person who made it. None of this is comfortable. All of it is faster than the polite alternative.
You will not like this principle if you find sharing something half-baked embarrassing, or if you spend more energy curating how things look than making them better. See good-interviewer-bad-interviewer for the version of this we apply to hiring loops specifically.
Radical transparency only works if the un-filtered version is the path of least resistance. The moment telling the truth costs a meeting or a careful Slack thread, you have built a company that filters by default.”
Work backwards from the customer
Every meaningful artifact starts with what the customer should feel, then works back to what we have to build to produce that. The PRD opens with the customer quote. The Loom video opens with the use case. The sales conversation opens with the workflow before the feature. If you cannot describe what the customer is going to say after using the thing, you do not yet know what you are building.
The honest version of this principle is harder than it reads. It means killing roadmap items that are interesting but do not change a customer's day. It means saying yes to integration work that is unglamorous but the right next thing. It is the principle that fights the loudest with engineering pride. It usually wins.
The shape it takes in the product: every Metaview feature is anchored to a specific moment in the recruiter's workflow, not to a technical capability we wanted to ship. The question that always closes the design review is "what does the recruiter feel when they open this?"
- 1The Ideal Candidate Profile is the customer-feel artifact. The recruiter wrote what good looks like; the surface treats that as the canonical instruction.
- 2Every applicant gets a Great / Good / Okay verdict against that profile, with the rationale on the row. No black-box score; the principle ships as a visible column.
- 3Progress or Reject is the recruiter's call, not the model's. The product shape always closes with "what does the recruiter feel when they open this?"
Optimize for rate of learning
The number we track in our heads is not throughput. It is loops per quarter. How many customer conversations turned into a product change? How many AB tests resolved into a permanent decision? How many things did we believe in January that we no longer believe in March, because we ran the test, not because someone changed their mind in the all-hands?
This principle costs you the certainty that a more measured pace would buy. The trade is that you compound. Two companies that start in the same place but learn at different rates end up in entirely different markets eighteen months later. We have watched that happen to peers in real time and decided the right side of it is the only one worth being on.
The practical version: when we want to discover the appetite for something new, it is rare that the first thing a customer touches will be powered by code. It is usually a deck, a report, a mechanical-turk version of the feature. Learning first, building second.
Drive to clarity
Ambiguity is a tax. Every decision that is not made today gets made later, by someone who has less context, in a worse forum, with more stakeholders. The principle is not that every meeting ends with a decision. The principle is that you do not leave the room with the same ambiguity you brought into it.
The tactical version: every meeting has a DRI named in writing. Every async thread that hits five replies gets a one-line summary at the top. Every PRD opens with the question it is trying to answer. Clarity is cheap to produce if you make it the habit. Letting ambiguity persist is expensive every single time.
You will not like this principle if you prefer consensus-driven decision-making, or if your reflex when something stalls is to think "well, at least it was not me who made the call." Driving to clarity costs accountability. Avoiding it costs progress.
Where AI gives recruiting teams use
Works backwards from the customer's intake call. The principle "work backwards from the customer" becomes the workflow shape: the conversation feeds the search.
Drive to clarity at the volume end. Every ranked candidate ships with the reasoning trail attached, not a black-box score. The principle is the column, not the slogan.
Own end-to-end. One workflow from interview to scorecard to debrief, no seam-failures. The principle ships as a single product surface instead of a handoff between tools.
Closes the rate-of-learning loop on every hire 12 to 18 months in. The team measures whether the principle decided correctly, then updates the principle if not.
The principles work the same way in the product as they do in the company. Metaview Notetaker is own-end-to-end for the recruiter: one workflow from interview to scorecard to debrief, no seam-failures. Application Review is drive-to-clarity at the volume end: every ranked candidate ships with the reasoning trail, not a black-box score. AI Sourcing works backwards from the customer's intake call. Reports closes the rate-of-learning loop on every hire, 12 to 18 months in. For the AI-augmented-recruiter framing, see claude-for-recruiters.
Numbers from the 2026 AI & Hiring Alignment Report, surveying 505 recruiting leaders and hiring managers across North America and EMEA. The 79% goal-attainment number is the most relevant to the principles thesis: when the principles do the work to keep teams aligned and the AI captures the evidence, the team exceeds its hiring goals at a rate the misaligned shop just does not.
The operating shift
Three concrete moves for any team that wants its principles to be load-bearing rather than decorative:
One: write the polarizing version, not the polished one. A principle that nobody could disagree with is not a principle. It is a value statement. Write the one that filters out the candidates and partnerships that would not work, and accept the cost of being clear about who you are.
Two: tie every principle to an artifact. Every decision references the principle that decided it. Every PR description names the principle. Every retro evaluates against them. Principles become wall art the moment the company stops doing the work to keep them load-bearing.
Three: retire any principle that has not decided anything in six months. The cost of holding onto a non-load-bearing principle is that it dilutes the load-bearing ones. Audit ruthlessly; cut decisively. Better to live with four principles that decide things than five that decorate.
The companies that internalize these three moves get principles that show up in the product. The ones that do not get principles that show up on a slide. That is the operating shift.
Bring Metaview into your hiring stack.
Live notes, structured scorecards, and ATS sync - set up in under 10 minutes.
Frequently asked questions
Are these the principles you actually hire against?
Yes. The polarizing version is the point. If a candidate flinches at radical transparency or at own end-to-end, the role is not going to work, and the kindest thing for both sides is to find out at the interview rather than at month three.
How do you stop principles from becoming wall art?
You tie them to artifacts. Every decision references the principle that decided it. Every PR description names the principle. Every retro evaluates against them. Principles become wall art the moment the company stops doing the work to keep them load-bearing.
Which principle is the hardest to live up to?
Radical transparency. The default human reflex is to filter. Filtering at scale produces a company where the leadership team thinks everything is fine and the IC layer knows it is not. We work harder on this one than on the other four combined.
Do the principles change over time?
The wording sharpens. The substance does not. If a principle was right at 20 people, it should still be right at 200. If it stops being right, the principle was decoration, not a principle.
How can a team install operating principles from scratch?
Start by listing the disagreements that keep happening across your team. The principles you need are the ones that would settle those disagreements without a meeting. Write the polarizing version. Tie each to an artifact (PR, retro, decision log). Audit at six months and retire anything that has not decided anything.
