Role-Based Prompting Tool
Most people use AI like a search engine. Here's how to use it like a thought partner.
Three techniques change almost everything about the answers you get: make it plan before it acts, tell it what's at stake, and give it the story — not just the task. Pick your role below and get all three as copy-ready prompts written for your job.
Plan First
Ask the AI to analyze and produce a plan before taking action. Separate the thinking from the doing.
When: Before any complex task where wrong assumptions will cost you time.
Label the Risk
Tag your request LOW or HIGH risk. The AI shifts its behavior based on the framing — more caution, more flags, fewer leaps.
When: Anytime you're touching something sensitive, client-facing, or hard to undo.
Give Context
Open the session with a narrative, not just a task. History, stakes, and what's gone wrong before change every answer that follows.
When: At the start of any session involving ongoing work, projects, or history.
Find your role
Click your role to jump to your prompts. We remember your pick for next time.
Vendor evaluations · security planning · incident response
IT Director
You live in the blast radius of other people's shortcuts — one rushed change, one missed patch, one vendor surprise. These techniques force the AI to work the way you'd force a junior tech to work: plan before touching anything, flag what's risky, and never start without the ticket history.
Plan First
When to use it: Vendor evaluations, migrations, anything with a rollback plan
I need to evaluate replacing our current backup vendor. Before you recommend anything, analyze the problem first: our environment is [size, sites, key systems], our pain points are [what's broken], and our constraints are [budget, timeline, compliance]. Give me your evaluation plan and the questions you'd need answered — do not jump to a recommendation yet.
Label the Risk
When to use it: Incident comms, security changes, anything client-facing
HIGH RISK: I'm drafting the incident notification for [audience] about [incident]. This is legally sensitive and hard to walk back. Draft carefully, flag any wording that could create liability or overpromise, and list what you're unsure about before I send anything.
Give Context
When to use it: Rollout planning, projects with organizational history
Context before we start: I run IT for a [size]-person company. Last year we did [previous project] and it went [how it went], so leadership is [current attitude toward change]. I'm now planning [new project]. With that history in mind, help me build a phased plan that accounts for how this organization actually behaves.
Budget forecasting · reconciliation · cash flow modeling
Finance Team
In finance, a confident wrong number is worse than no number. These techniques make the AI show its work before it touches your model, treat sensitive figures with the caution they deserve, and stop it from 'helpfully' assuming facts about your business it doesn't have.
Plan First
When to use it: Forecast builds, model changes, new reporting
I need to build a 12-month cash flow forecast for [company/department]. Before producing any numbers, lay out your approach: what inputs you need from me, what assumptions you'd otherwise have to make, and where forecasts like this usually go wrong. I'll confirm the plan before you build anything.
Label the Risk
When to use it: Reconciliations, restatements, anything auditors will see
HIGH RISK: I'm reconciling [account] and the books are off by [amount]. This will be reviewed by our auditors. Walk through the discrepancy methodically, flag every assumption you make, and clearly separate what you can verify from what you're inferring.
Give Context
When to use it: Ongoing budget cycles, recurring analysis
Context: I'm the [role] at a [size, industry] company. Our fiscal year ends [month], we're currently in [budget stage], and last cycle the pain point was [what went wrong]. Today I need help with [task]. Keep that history in mind — especially [the thing that burned us].
Board presentations · acquisition evaluation · strategy pivots
CEO
Your decisions are expensive to reverse and everyone downstream amplifies them. These techniques stop the AI from telling you what sounds impressive and make it tell you what's actually true — plan the analysis, name the stakes, and give it the backstory your leadership team already knows.
Plan First
When to use it: Big decisions, board prep, strategic analysis
I'm evaluating [acquisition/expansion/pivot]. Don't give me an opinion yet. First, lay out how you'd analyze this decision: the key questions, the data you'd want, the failure modes of deals like this, and the strongest case against it. I want the analysis plan before any conclusions.
Label the Risk
When to use it: Investor and board communications, public statements
HIGH RISK: I'm drafting my board update on [topic] and the news is mixed. This audience remembers every number I've ever given them. Draft it straight — no spin — flag anywhere I might be overcommitting, and tell me what questions this update will provoke so I'm ready for them.
Give Context
When to use it: Strategy sessions, anything touching company history
Context: I run a [size, industry] company. Two years ago we [key history], last year we [recent event], and the board's current mood is [sentiment]. My leadership team's blind spot is [weakness]. With all of that on the table, help me think through [decision].
Performance review design · retention strategy · org redesign
Chief People Officer
Everything you touch involves real people and real careers, and half of it is confidential. These techniques make the AI plan people-processes before prescribing them, respect the sensitivity of what you're working with, and factor in the org history that makes your company different from the generic one it learned about.
Plan First
When to use it: Process redesigns, new programs, policy rollouts
I need to redesign our performance review process for [company size/type]. Before proposing a framework, plan the work: what you need to know about our culture and current process, what typically kills review redesigns, and how we'd measure whether the new process works. Plan first, framework after I confirm.
Label the Risk
When to use it: Terminations, investigations, sensitive communications
HIGH RISK: I'm preparing communication about [sensitive change — reorg, exit, investigation outcome]. Legal exposure and employee trust are both on the line. Draft conservatively, flag anything that could be quoted against us or read as a promise, and list the questions employees will actually ask.
Give Context
When to use it: Retention work, culture initiatives, org design
Context: I lead People at a [size, industry] company. Our regretted attrition is [rate], concentrated in [team/level]. We tried [previous initiative] and it [result]. The exec team believes [their theory]; I suspect [your theory]. With that context, help me build [deliverable].
Process documentation · capacity planning · vendor review
Operations Manager
Your job is making things run the same way twice, and the details the AI skips are exactly the details that break a process. These techniques get you documentation planned around how work actually flows, caution on the contracts and commitments, and answers grounded in your operation instead of a textbook one.
Plan First
When to use it: Process documentation, workflow changes
I need to document our [process] end to end. Before writing anything, give me your plan: who you'd need input from, what questions expose the undocumented exceptions, and how you'd structure the doc so people actually use it. I'll answer your questions, then we write.
Label the Risk
When to use it: Vendor contracts, SLAs, capacity commitments
HIGH RISK: I'm reviewing our [vendor] contract renewal before signing. Getting this wrong locks us in for [term]. Go through it clause by clause: flag auto-renewals, termination terms, liability caps, and anything unusual, and separate 'standard language' from 'negotiate this.'
Give Context
When to use it: Capacity planning, recurring operational analysis
Context: I manage operations for [what you run]. Our volume is [baseline] with peaks at [when]. Last [period] we got burned by [what happened]. Staffing is [current state]. With that as the baseline, help me plan capacity for [upcoming period/change].
Scope definition · risk registers · stakeholder communications
Project Manager
Every project fails at the beginning — it just gets discovered at the end. These techniques front-load the discovery: scope planned before it's promised, risks named while they're still cheap, and status told with the history that makes stakeholders trust it.
Plan First
When to use it: Kickoffs, scope definition, estimation
I'm scoping [project] for [stakeholder]. Before drafting the scope, plan the scoping itself: what you'd need to know, which requirements sound simple but usually aren't, and where projects like this historically blow up. Questions first — the scope document comes after I answer them.
Label the Risk
When to use it: Risk registers, schedule commitments, escalations
HIGH RISK: I'm about to commit a delivery date for [milestone] to [audience]. Build the case before I do: the assumptions baked into the date, the top risks to it, and what buffer is defensible. If the honest answer is 'don't commit yet,' say so and tell me what would change that.
Give Context
When to use it: Status reporting, stakeholder management
Context: I'm running [project], currently [status] against a [date] deadline. [Stakeholder] is anxious because [history — slipped date, past project, budget]. The team's real blocker is [issue]. Help me write this week's status update so it's honest about [problem] without triggering [overreaction].
Pipeline analysis · territory planning · win/loss reviews
Sales Leader
Optimism is your team's fuel and your forecast's poison. These techniques separate the two: analysis planned before conclusions get drawn, honest labels on the numbers that go up the chain, and deal reviews that account for what actually happened last quarter — not the version told at kickoff.
Plan First
When to use it: Pipeline reviews, territory design, comp planning
I need to analyze why [segment/rep/region] is underperforming. Before offering theories, plan the analysis: what data you'd want, what the common causes look like in the numbers, and how to tell a coaching problem from a territory problem from a product problem. Plan first, then we dig in.
Label the Risk
When to use it: Forecasts, quota setting, exec commitments
HIGH RISK: I'm submitting my [quarter] forecast to the CEO and I'll be held to it. Here's my pipeline: [summary]. Stress-test it — flag deals that look like happy ears, tell me what number is defensible versus aspirational, and give me the risk-adjusted version you'd actually commit.
Give Context
When to use it: Win/loss reviews, strategy sessions
Context: I lead sales at a [size, industry] company. We sell [product] at [price point] against [competitors]. Last quarter we [result], mostly because [honest reason]. The board's expectation is [target]. With that reality, help me build [plan/review/strategy].
Campaign planning · content strategy · audience research
Marketing Leader
The AI will happily generate ten campaigns for an audience neither of you understands yet. These techniques fix the order of operations: research planned before content gets written, brand-sensitive work flagged as such, and strategy built on what your audience has actually responded to — not what averages say.
Plan First
When to use it: Campaign planning, new channel launches
I'm planning a campaign for [product/launch] targeting [audience]. Don't write anything yet. First give me the plan: what we need to know about this audience, how you'd position against [alternatives], what channels fit and why, and how we'd measure success. Creative comes after the plan holds up.
Label the Risk
When to use it: Public-facing copy, pricing announcements, crisis comms
HIGH RISK: This copy is going public under our brand — [piece and placement]. It touches [sensitive area: pricing, competitor comparison, claims]. Draft it, then audit your own draft: flag every claim that needs substantiation and anything that could age badly or invite a screenshot.
Give Context
When to use it: Content strategy, ongoing campaign optimization
Context: I run marketing for [company]. Our audience is [who], and historically they respond to [what works] and ignore [what doesn't]. Our last campaign [result and lesson]. Brand voice is [description] — we never [taboo]. With that, help me build [deliverable].
Contract review prep · policy drafting · audit readiness
Legal / Compliance
You already treat every document as discoverable and every ambiguity as a future dispute — the AI doesn't, unless you make it. These techniques put the caution in explicitly: review plans before reviews, risk labels the AI takes literally, and the regulatory context that changes which answer is correct.
Plan First
When to use it: Contract reviews, policy projects, audits
I'm preparing to review [contract type] with [counterparty]. Before looking at any language, plan the review: the clauses that matter most for our side, the market-standard positions, and the traps common in this contract type. I want your review checklist first, then we go through the document.
Label the Risk
When to use it: Anything that creates or waives obligations
HIGH RISK: I'm drafting [policy/clause/response] regarding [matter]. This creates binding obligations and may be reviewed by [regulator/opposing counsel]. Draft conservatively, flag every position that's arguable rather than settled, and note where I need a licensed attorney's judgment rather than a starting draft.
Give Context
When to use it: Ongoing matters, audit preparation
Context: I handle [legal/compliance] for a [size, industry, jurisdictions] company. We're subject to [regulations]. Our last audit flagged [findings], and we've since [remediation]. Institutional sensitivity: [what leadership worries about]. With that, help me prepare [deliverable].
Board management · grant writing · program strategy
Nonprofit Executive Director
You do five jobs with the budget of one, and every stakeholder — board, funders, staff, community — hears everything you write. These techniques stretch your capacity without risking your credibility: plans before proposals, care on anything a funder will read, and the mission context that generic advice always misses.
Plan First
When to use it: Grant applications, program design, strategic planning
I'm pursuing a [funder] grant for [program]. Before drafting anything, plan the application: what makes proposals to funders like this succeed or fail, what evidence and budget detail we'll need, and the questions I should answer before we invest hours in writing. Plan first — the narrative comes after.
Label the Risk
When to use it: Board communications, funder reports, public statements
HIGH RISK: I'm reporting to [board/funder] on [program that underdelivered]. They can pull funding, and this document sticks around. Help me tell it straight — real numbers, real reasons — while showing what we learned and changed. Flag anything that reads as excuse-making or spin.
Give Context
When to use it: Strategy, fundraising, anything mission-adjacent
Context: I run a [mission] nonprofit with [budget] and [staff size]. Our funding is [mix], our board is [composition and temperament], and the tension right now is [growth vs. stability / program vs. overhead / etc.]. With that reality — not a Fortune 500 playbook — help me with [task].
Want the backstory on why these three techniques work? Read the full story on the AdapToIT blog.
New roles land regularly.
Get an email when we add prompts for a new role. No newsletter, no noise.