{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-08-03T04:07:39.617Z"},"content":[{"type":"documentation","id":"7b545afe-5ba9-45b2-bf7d-6defca3ff050","slug":"first-90-days-ux-researcher","title":"The First 90 Days as a UX Researcher: A Week-by-Week Onboarding Plan (2026)","url":"https://www.koji.so/docs/first-90-days-ux-researcher","summary":"A new UX researcher should spend the first 30 days learning the system rather than producing findings - auditing past research, mapping decision-making, and interviewing stakeholders as if they were a research project. Days 31 to 60 are for shipping one small study end to end to establish credibility and expose process friction. Days 61 to 90 are for changing exactly one process. This guide gives the week-by-week plan, the week-one audit checklist, a stakeholder interview script, day-90 success criteria, and how AI-native tooling lets a new hire ship real findings in week two instead of month three.","content":"# The First 90 Days as a UX Researcher: A Week-by-Week Onboarding Plan (2026)\n\n**Short answer:** Spend days 1-30 learning the system rather than producing findings; days 31-60 shipping one small study end to end; days 61-90 changing exactly one process. The instinct to prove yourself with an ambitious foundational study in month one is the single most common way new researchers spend credibility they have not yet earned - on a question nobody was blocked on.\n\nThe stakes are higher than most managers assume. Gallup finds that only **12% of employees strongly agree their organisation does a great job of onboarding new hires**, and only **29% of new hires say they feel fully prepared and supported to succeed** after onboarding. The same research finds that when managers take an active role, employees are **3.4 times as likely** to strongly agree their onboarding was exceptional. Onboarding is not an administrative phase. It is the period in which your ceiling gets set.\n\nThis guide is written for the researcher. If you are the manager, [UX Researcher Career Path](/docs/ux-researcher-career-path) covers what to expect at each level, and [UX Research Team Structure](/docs/ux-research-team-structure) covers where the role should sit.\n\n## Days 1-30: learn the system\n\nYour first month has one deliverable, and it is not a study. It is a map: **who decides what, on what evidence, and what they currently believe.**\n\n### The week-one audit\n\nBefore you talk to anyone, find out what already exists. Most organisations have more research than they think and less than they claim.\n\n| Audit item | Question to answer | Red flag |\n|---|---|---|\n| Past studies | What has been researched in the last 24 months? | Nobody can find them |\n| Repository | Where do findings live, and does anyone read them? | Findings live in individual Slack DMs |\n| Participant access | Can you legally and practically reach customers? | Every request routes through one gatekeeper |\n| Consent and legal | What consent, retention, and DPA terms apply? | No template exists |\n| Tooling | What is licensed, what is actually used? | Three overlapping tools, none adopted |\n| Prior conclusions | What does the org believe it already knows? | Beliefs with no traceable source |\n\nThat last row is the most valuable and the most skipped. Write down every confident claim you hear in your first fortnight - *\"our users don't care about pricing,\"* *\"enterprise buyers want SSO first\"* - and note whether anyone can point to where it came from. Unsourced beliefs are your best backlog. They are what the organisation is currently betting on without evidence.\n\n### Interview your stakeholders like participants\n\nNielsen Norman Group makes this point directly: approach onboarding intro calls **as a small research project**, to learn as much as possible about the people and the context. That reframing matters. It converts a fortnight of pleasant introductions into a study with a sample, a discussion guide, and findings.\n\nAim for 8-12 stakeholders across product, design, engineering, support, sales, and one executive. Use a consistent guide so the answers are comparable:\n\n1. What decision are you making in the next quarter that you are least confident about?\n2. What do you believe about our users that you cannot prove?\n3. When did research last change your mind? What happened?\n4. What has stopped you using research in the past?\n5. If you could have one question answered perfectly, what would it be?\n\nQuestion 3 is the diagnostic. If nobody can recall research changing their mind, you are not joining a research practice - you are founding one, and your first 90 days should be planned accordingly. [Stakeholder Interviews](/docs/stakeholder-interview-guide) covers the full method.\n\nThen synthesise it properly and share it back. A one-page readout of *\"here is what this organisation believes and cannot prove\"* in week four is the highest-leverage artifact a new researcher can produce. It is genuinely useful, it demonstrates your method, and it costs you nothing politically because every claim in it came from them.\n\n## Days 31-60: ship one small thing\n\nPick a first study by three criteria, in this order:\n\n1. **Someone is actively blocked on it.** Not interested - blocked. A blocked person is a person who will act on your finding and tell others they did.\n2. **It can be done in under three weeks.** Momentum beats comprehensiveness at this stage.\n3. **It has a visible decision attached.** You want to be able to say *\"we shipped X instead of Y because of this.\"*\n\nWhat to avoid: the foundational persona refresh, the full journey map, the research repository rebuild. These are the projects new researchers instinctively reach for because they look substantial and offend nobody. They also take a full quarter and produce nothing anyone can point to at your first review.\n\n**The traditional bottleneck.** The honest problem with \"ship a study by day 60\" is that a conventional study requires exactly the things a new hire does not have: a recruiting pipeline, a participant panel, calendar access to customers, and a relationship with whoever guards them. This is why the classic advice quietly assumes your first real study lands in month three or four.\n\n**The 2026 version.** That constraint has genuinely loosened. Maze's *Future of User Research 2026* study (n≈500, fielded December 2025 to January 2026) found **69% of research practitioners now use AI in their work, up 19 percentage points year over year**, and that the share saying research is essential to strategy rose from **8% to 22%**. Demand for research is up **66%**, and **39% of product managers** now conduct research themselves. You are joining an organisation that expects research faster than the old pipeline can deliver it.\n\nWith an AI-native platform, the sequence changes:\n\n- **Week 2:** launch an always-on AI-moderated study on a question from your stakeholder interviews. No calendar slots to book, no moderation to schedule, no transcription queue.\n- **Week 3:** 30 completed interviews. Koji's thematic analysis clusters the open-ended responses into named themes with the verbatim quotes attached, so you are reviewing findings rather than tagging transcripts.\n- **Week 4:** readout to the person who was blocked, with structured-question data behind every claim.\n\nKoji's six structured question types - `open_ended`, `scale`, `single_choice`, `multiple_choice`, `ranking`, `yes_no` - matter more than usual for a new hire, because you are establishing a reputation for rigour. A finding that pairs a comparable scale distribution with the reasoning behind it survives challenge from a sceptical stakeholder; a finding that is three quotes does not. See the [structured questions guide](/docs/structured-questions-guide).\n\nYou do not need a research background to run this well, which is also the point: in an organisation where PMs are already doing their own research, your value is not being the only person who can run a study. It is being the person who makes everyone else's studies trustworthy.\n\n## Days 61-90: change one process\n\nExactly one. New researchers who try to reform recruitment, the repository, intake, and consent simultaneously finish the quarter with four half-built systems and no adoption.\n\nChoose based on what actually blocked you during your first study:\n\n| What blocked you | Process to fix | Success measure |\n|---|---|---|\n| Could not find past findings | A working repository with a naming convention | Someone else finds something without asking you |\n| Could not reach customers | A standing recruitment pipeline | Time-to-first-participant under 5 days |\n| Requests arrived as vague asks | An intake form with a decision field | Requests name the decision they inform |\n| Legal review stalled you | Approved consent and DPA templates | Zero legal review for standard studies |\n\nThe success measure is the part people skip. \"We built a repository\" is not an outcome; \"someone found a study without asking me\" is.\n\n## The plan at a glance\n\n| Phase | Focus | Primary output | What to resist |\n|---|---|---|---|\n| Days 1-30 | Learn the system | A map of decisions, beliefs, and evidence gaps | Launching a big study |\n| Days 31-60 | Ship one small thing | One acted-upon finding | Scope creep into a foundational project |\n| Days 61-90 | Change one process | One improvement with a measurable outcome | Reforming everything at once |\n\n## What good looks like at day 90\n\nA checklist you can hold yourself to:\n\n- [ ] At least one study shipped **and acted on** - you can name the decision it changed\n- [ ] A written map of who decides what, on what evidence\n- [ ] A shared list of the organisation's unsourced beliefs\n- [ ] One process improved, with a measure that moved\n- [ ] Relationships with 8-12 stakeholders who know what you do\n- [ ] A repeatable study setup someone else could run\n- [ ] Awareness of your legal and consent constraints\n\nWhat is deliberately **not** on that list: a large number of studies. Anyone measuring a new researcher on study volume in the first quarter is measuring the wrong thing, and it is worth saying so early and calmly.\n\n## Five mistakes to avoid\n\n1. **Leading with methodology critique.** Pointing out that the previous survey was leading is correct, unhelpful, and expensive in your first month. Fix it silently in the next study.\n2. **Waiting for perfect access.** If customer access is genuinely blocked, run the study with prospects, churned users, or an adjacent segment and label it honestly. Something imperfect in week six beats something perfect in month five.\n3. **Over-indexing on your manager's framing.** Interview widely. Your manager's view of the research gaps is one data point, and often the one most shaped by past frustration.\n4. **Confusing being busy with being useful.** Attending every design review is not research. It is the most comfortable way to spend 90 days and produce nothing.\n5. **Not writing the 30/60/90 down.** Share it with your manager in week one and revise it in week six. It is the artifact that converts a vague review into a factual one.\n\n## Frequently asked questions\n\n**What should a UX researcher do in the first 30 days?**\nLearn the system, not the users. Audit existing research, map how decisions actually get made, and interview 8-12 stakeholders about what they believe and cannot prove.\n\n**What is a good first project?**\nSomething small, decision-linked, and already wanted - ideally answering a question a specific person is actively blocked on, completable in under three weeks.\n\n**Should you run research in your first month?**\nYes, but on the organisation. Stakeholder interviews are real research and produce the map you need before any customer study is worth running.\n\n**How do you build credibility as the first researcher at a company?**\nShip something small quickly, attach it to a decision someone cares about, and let the person who acted on it tell the story. Credibility is transferred, not asserted.\n\n**What does success look like at day 90?**\nOne study shipped and acted on, a working map of decisions and evidence, and exactly one process improved with a measure that moved.\n\n**How does AI-assisted tooling change this?**\nIt moves your first real study from month three to week two, because the blockers it removes - recruiting, scheduling, moderating, transcribing - are precisely the ones a new hire has no relationships to solve.\n\n## Related Resources\n\n- [UX Researcher Career Path](/docs/ux-researcher-career-path) - levels, competencies, and promotion criteria\n- [UX Research Team Structure](/docs/ux-research-team-structure) - centralised, embedded, and hub-and-spoke models compared\n- [How to Hire a UX Researcher](/docs/hiring-ux-researcher-guide) - the other side of this process\n- [ResearchOps: The Complete Guide](/docs/research-ops-guide) - the processes you will be improving in days 61-90\n- [Stakeholder Interviews](/docs/stakeholder-interview-guide) - the method for your first month\n- [How to Build a UX Research Repository](/docs/research-repository-guide) - the most common day-61 project\n- [User Research Maturity Model](/docs/user-research-maturity-model) - diagnosing which stage you have joined\n- [Structured Questions in AI Interviews](/docs/structured-questions-guide) - making your first findings hard to dismiss\n\n---\n\n**Sources:** Gallup, *Why the Onboarding Experience Is Key for Retention*; Nielsen Norman Group, *Successful Onboarding for New Hires in UX Roles*; Maze, *Future of User Research 2026*.","category":"Research Operations","lastModified":"2026-08-01T03:20:59.339339+00:00","metaTitle":"First 90 Days as a UX Researcher: A 30/60/90 Onboarding Plan","metaDescription":"A week-by-week onboarding plan for new UX researchers: the week-one audit, stakeholder interviews as your first study, which project to ship by day 60, and what good looks like at day 90.","keywords":["first 90 days ux researcher","ux researcher onboarding","30 60 90 day plan ux research","new researcher onboarding","first ux research hire","research onboarding plan"],"aiSummary":"A new UX researcher should spend the first 30 days learning the system rather than producing findings - auditing past research, mapping decision-making, and interviewing stakeholders as if they were a research project. Days 31 to 60 are for shipping one small study end to end to establish credibility and expose process friction. Days 61 to 90 are for changing exactly one process. This guide gives the week-by-week plan, the week-one audit checklist, a stakeholder interview script, day-90 success criteria, and how AI-native tooling lets a new hire ship real findings in week two instead of month three.","aiPrerequisites":["A new or incoming user research role","Basic familiarity with common UX research methods"],"aiLearningOutcomes":["Structure a 30/60/90 day plan for a new research role","Run a week-one audit of an existing research practice","Interview stakeholders as a research project rather than a formality","Choose a first study that builds credibility instead of consuming it","Define measurable day-90 success criteria"],"aiDifficulty":"beginner","aiEstimatedTime":"12 min"}],"pagination":{"total":1,"returned":1,"offset":0}}