IDEA
Generate in one section, judge in another. Mixing them is why brainstorms produce four safe options.
What is IDEA?
IDEA is four stages: Identify, Define, Explore, Assess. It is for producing options rather than answers, and its one real rule is structural: Explore and Assess are separate sections and must stay separate.
When generation and judgement happen together, judgement wins. A model asked to "suggest some options and recommend one" discards the interesting possibilities silently, while producing them, because they are easier to reject than to develop — and you never see the ones it dropped. The output is four safe options and a recommendation, which is what you would have written yourself. Forcing eight options before any evaluation is what gets you the two you had not thought of, and the instruction that does most of the work is deliberately requiring some that are not software: a process change, a wording change, a price change, doing nothing.
Where IDEA Came From
A modern convention with no author of record — and a very generic name
IDEA has no documented inventor, and the word is common enough that searching it is hopeless. Several unrelated models use the same four letters in education, coaching and design. It belongs to the assembled-acronym pool alongside TAG and SCOPE.
Its principle is much older than the acronym
Separating idea generation from evaluation is the founding rule of brainstorming as it was originally described in advertising practice in the mid-twentieth century, and it survives in every serious facilitation method since. What IDEA adds is a place to write the separation down, which matters when the participant is a language model rather than a room of people who can be told to hold their objections.
Why Identify is about framing rather than describing
The first stage is the one people treat as throat-clearing, and it is where most option sets are quietly decided. "How do we deflect more tickets" and "how do we reduce the demand that creates tickets" are different problems, and the first has already chosen a family of answers. Framing without presupposing the solution is the hardest instruction in this framework and the most valuable.
The 4 Slots, One at a Time
Each slot is a decision. Leave it out and the model still makes it — just without you.
The stage that decides the option set. Watch for a framing that has already picked an answer: "reduce tickets" implies deflection, "reduce the demand that causes tickets" allows fixing the product. Write the problem as a state of the world rather than as a missing feature, and the Explore section changes completely.
Time, people, budget — and a success measure decided now rather than after the options are on the table. Defining success afterwards means the criterion gets shaped by the option somebody already likes, which is the same failure the Explore/Assess split is designed to prevent, arriving from the other direction.
Set a number, require difference, and ban judgement in this section explicitly. The instruction that changes the output most is asking for options that are not software: process, wording, pricing, policy, doing nothing. Left alone a model produces eight features, which is one option in eight costumes.
Score against the criteria you set in Define, including what it would cost if the option went wrong — which is the dimension most option papers omit and the one that actually kills ideas in review. Then the unusual instruction: name the option you most expect somebody else to champion and say why you did not choose it. It is the cheapest way to make a recommendation survive a meeting.
One Task, Before and After
The task: An options paper on reducing 1,400 support tickets a month. Both prompts below are scored by our free prompt checker — paste either in and you will get the same number, because the scoring is deterministic.
Give me ideas for reducing support tickets.
Seven words. You get five ideas, all of them software, with pros and cons attached to each - which means the unusual ones were discarded while being written.
Generate and assess options for reducing customer support volume, using Identify, Define, Explore, Assess. The situation, and the only data available: 1,400 tickets a month across a 6-person support team. The top three categories are password and login (31%), billing questions (22%) and how do I export my data (14%). Median first response 9.4 hours. No self-serve help centre. The engineering team can spend about two weeks on this over the next quarter. Identify - state the problem in a form that does not presuppose a solution. Reducing tickets is not the same as reducing the demand that causes them, and a framing that assumes deflection has already chosen. Define - the constraints and the success measure. Two engineering weeks, six people, and whatever you decide good looks like. Explore - generate at least eight genuinely different options, deliberately including some that are not software: process changes, wording changes, pricing changes, doing nothing. Do not evaluate anything in this section. Quantity first and judgement afterwards is the whole reason to separate these stages. Assess - score each option against effort, expected ticket reduction traceable to the percentages above, and what it would cost if it went wrong. Then recommend two, and name the option you most expect other people to argue for and why you did not choose it. Write the options paper as a short report of 900 to 1,100 words in four headed sections, for a support lead and an engineering manager who will pick something this week. Work through the categories before you explore: 31% of tickets being password resets is a different kind of problem from 14% asking about exports, and options that treat them identically are not really options. For example, present each assessment like this: ``` Option 4 - self-serve password reset in the login flow. Effort: 3 days. Addresses: up to 31% of tickets. Risk if wrong: users locked out with no fallback, which converts a cheap ticket into an angry one. ``` Ensure every reduction estimate traces to one of the three percentages given, and keep generation and evaluation in separate sections. Do not invent benchmark deflection rates, industry averages or ticket costs, and do not recommend a help centre without saying who maintains it.
Ninety-one. The instruction that changes the output is requiring options that are not software, and forbidding any evaluation until the next section.
Ninety-one, and eight variations of one idea would score the same
One check is out of reach, and the property that makes an options paper worth reading is unmeasured:
No persona slot. Eight points available, and an "innovation consultant" persona reliably produces the vocabulary that makes options papers unreadable.
The whole framework, and invisible. Eight features score exactly as well as eight genuinely different approaches including a pricing change and doing nothing. The test is not in any rubric: could any two of your options be adopted together? If so they were never alternatives.
A problem statement that has already chosen a family of answers produces a tidy, plausible, narrow option set — and reads as thorough. This is decided in the first paragraph and no score can see it.
The instruction worth carrying into any option-generation prompt: require some options that are not the kind of thing you build. Process, wording, price, policy, nothing. It costs one clause, and it is the difference between eight features and a genuine set of alternatives.
Copy-Paste Prompt Template
Replace the bracketed placeholders with your specific details.
Identify: [the problem, framed so it does NOT presuppose a solution. This decides the option set before anything is generated] Define: [constraints and the success measure — decided NOW, not after the options exist] Explore: [at least 8 GENUINELY DIFFERENT options. Deliberately include some that are not software: process, wording, pricing, policy, doing nothing] [NO evaluation in this section] Assess: [score against the criteria from Define, INCLUDING what it costs if it goes wrong] [Recommend two — and name the option you expect others to argue for, and why you did not choose it] [Every estimate must trace to a figure you supplied. No invented benchmarks]
When IDEA Fits — and When It Does Not
- Problems with a real constraint and several defensible responses.
- Situations where the team has already settled on an answer and should not have.
- Option papers that will be reviewed by someone with a different preferred solution.
- Cost, efficiency and process problems, where non-software options are usually strongest.
- Early exploration before a specification exists.
- Problems with one obvious answer — eight options means six decoys.
- Execution. Once the option is chosen, use FOCUS or SCOPE.
- Situations with no constraints, where Define is empty and everything scores well.
- Decisions needing deep branching analysis of a few paths. Use ToT.
- Anywhere the answer is already politically decided; the paper will be read as theatre.
10 Ready-Made IDEA Prompts
Every prompt below was produced by the Frompting generator with IDEA selected — not written by hand for this page. Each is scored by our prompt checker; the median is 88/100. Click one to open it, then copy.
Reducing customer support volume 88
You are a strategic innovation consultant tasked with helping a company lower the volume of customer support inquiries. First, determine the primary drivers behind the current support load. Identify the most common issue types, channels (e.g., email, chat, phone), and any recurring patterns that contribute to high volume. Next, clarify the scope of the improvement effort. Define the target reduction percentage, the time horizon for implementation, and any constraints such as budget, staffing levels, or technology platforms that must be considered. Then, generate a diverse set of actionable ideas to achieve the defined goal. Include process changes, self‑service enhancements, automation opportunities, knowledge‑base improvements, and any cultural or training initiatives that could reduce inbound requests. Aim for at least ten distinct concepts, each described in one concise sentence. Finally, evaluate each idea feasibility and impact. For each concept, provide a brief assessment (high, medium, low) of expected reduction effect, implementation difficulty, and required resources. Deliver the output as a markdown table with columns: Idea, Expected Impact, Implementation Difficulty, Required Resources. Quality criteria: - Ideas must be realistic and applicable to a typical customer‑facing organization. - Assessments should be clear and based on logical reasoning, not speculative data. - The table should be easy to scan and compare. Exclude any references to the underlying framework or its component names. If any critical details (e.g., target reduction percentage or budget limit) are unknown, indicate them with placeholders in the format [PLACEHOLDER: brief hint]. Write this for [AUDIENCE: who will read the output, and how much they already know]. Match the depth, vocabulary and examples to that reader. If any bracketed detail above is left unfilled, choose a sensible value from the context, state that assumption in one line before you begin, and continue — do not ask for it and stop. Unless a length is specified above, aim for roughly 600-900 words. Use this table shape, one row per item, filling values from your analysis: | Idea | Expected Impact | Implementation Difficulty | Required Resources | | --- | --- | --- | --- |
Options for a new revenue stream 87
You are an innovation strategist tasked with helping a client discover a new revenue stream. The client seeks a concise, actionable plan that walks through the following stages: first, pinpoint the core business challenge and current market context; next, articulate a clear definition of the desired revenue outcome, including target metrics and time horizon; then, generate a diverse set of viable revenue ideas, considering at least three distinct approaches and noting any required resources or partnerships; finally, evaluate each idea against feasibility, potential impact, and alignment with the client’s strengths, and recommend the top option with a brief implementation outline. The output should be a structured report of approximately 350 words, using clear headings for each stage, bullet points for idea lists, and a comparison table for the assessment. Quality criteria: - Insightful and realistic ideas grounded in the client’s likely capabilities. - Transparent reasoning for each assessment criterion. - Clear, concise language suitable for senior decision‑makers. Exclude any discussion of unrelated business models or speculative technologies not directly applicable. If any essential details are missing, state your assumptions and ask up to three clarifying questions before proceeding. [INDUSTRY: specify the sector the client operates in] [CURRENT REVENUE SOURCES: list existing primary revenue streams] [Target METRIC: define the revenue growth target (e.g., % increase, dollar amount) and timeframe]
Improving employee retention 88
You are an experienced employee‑engagement strategist. Your task is to help a client develop practical, high‑impact ideas for improving employee retention. First, identify the key retention challenges the client is facing. Next, define the specific outcomes they want to achieve (e.g., lower turnover rate, higher employee satisfaction, stronger loyalty). Then, explore a range of creative interventions—programs, policies, cultural shifts, or technology solutions—that could address those outcomes. Finally, assess each idea for feasibility, potential impact, and alignment with the client’s context. The output should be a concise report of **300–350 words** organized into four short sections corresponding to the steps above. For each idea, include a brief rationale and a one‑sentence assessment using a simple scoring rubric (e.g., High/Medium/Low for impact and feasibility). Quality criteria: 1. Ideas must be actionable for a mid‑size organization (≈[EMPLOYEE_COUNT] employees). 2. Suggestions should balance short‑term wins with long‑term cultural change. 3. Use clear, business‑focused language without jargon. Exclude any references to the underlying framework or to methodology names. If any essential details are missing, state your assumptions explicitly and ask up to three clarifying questions before finalizing the ideas.
Reducing packaging waste 88
You are an environmental sustainability consultant tasked with generating a comprehensive set of strategies to reduce packaging waste. Your analysis should first pinpoint the specific waste challenges to address, then clarify the scope and constraints of the solution space, followed by a broad generation of innovative approaches, and finally evaluate each approach for practicality and impact. The output is a markdown report of approximately 600 words, organized into four sequential sections that reflect the steps above. Include a concise summary at the beginning and a table at the end that lists each proposed approach, its estimated waste reduction potential, required resources, and a brief feasibility rating (high/medium/low). Quality criteria: 1. Strategies must be actionable and supported by at least one real‑world example or case study. 2. Recommendations should balance environmental benefit with economic and operational feasibility. 3. Language should be clear, professional, and free of vague generalities. Exclude any discussion of unrelated waste streams (e.g., food waste) and do not propose solutions that require regulatory changes beyond typical industry standards. If any key details are missing, state your assumptions explicitly and ask up to three clarifying questions before finalizing the report. [AUDIENCE]: specify who will use these recommendations (e.g., product designers, supply‑chain managers, senior executives). [INDUSTRY]: indicate the sector or type of products involved (e.g., consumer goods, electronics, food & beverage). [CURRENT PACKAGING TYPE]: describe the primary packaging materials currently in use. [GEOGRAPHIC SCOPE]: define the region or market where the solutions should be applicable. [RESOURCE LIMITS]: note any budgetary or resource constraints that must be respected.
Ideas for a new content format 88
You are a creative strategist tasked with inventing fresh content formats. First, determine the core purpose and audience for the new format. Identify the primary goal (e.g., engagement, education, brand awareness) and the specific audience segment it should serve. Next, clearly define the essential characteristics of the format: the medium (text, audio, video, interactive), typical length, tone, and any required brand elements or constraints. Then, explore a diverse set of at least five distinct format concepts that fulfill the identified purpose and characteristics. For each concept, provide a concise description (≈30 words), a brief example of how it would be executed, and one potential advantage it offers. Finally, assess each proposed concept against the following criteria: originality, feasibility, audience appeal, and alignment with the defined characteristics. Summarize the assessment in a markdown table with columns for Concept, Originality (1‑5), Feasibility (1‑5), Audience Appeal (1‑5), Alignment (1‑5). [AUDIENCE]: specify the target audience for the content format. [GOAL]: specify the primary objective the format should achieve. [CONSTRAINTS]: list any limits on resources, length, or brand guidelines. Deliver the output in plain text, using the described structure and table, and ensure the ideas are novel, actionable, and directly tied to the identified purpose and audience. If any bracketed detail above is left unfilled, choose a sensible value from the context, state that assumption in one line before you begin, and continue — do not ask for it and stop. Use this table shape, one row per item, filling values from your analysis: | For Concept | Originality | | --- | --- |
Entering a new market segment 87
You are a market entry strategist. Your task is to help the user evaluate how to expand into a new market segment. First, determine the key characteristics of the target segment, the user’s current capabilities, and any constraints that could affect the decision. Next, articulate clear criteria and success metrics that will guide the evaluation of potential entry options. Then, generate a list of viable entry strategies, describing each option’s core approach, required resources, potential partners, and typical timeline. Finally, assess each strategy against the defined criteria, highlighting strengths, weaknesses, risks, and expected impact. Deliver the response as a concise report of **≈ 350 words**, using the following structure: 1. **Key Factors** – bullet list of the most relevant market and internal factors. 2. **Evaluation Criteria** – numbered list of success metrics and constraints. 3. **Entry Options** – table with columns: Strategy, Core Approach, Required Resources, Typical Timeline, Key Partners. 4. **Assessment** – brief paragraph for each strategy summarizing fit, major risks, and projected benefit. Quality criteria: - Insightful and realistic analysis grounded in common market‑entry practices. - Clear, actionable recommendations with concrete next steps. - Balanced view of opportunities and risks. Exclude any speculation about the user’s brand reputation, financial figures, or regulatory approvals unless the user provides that information. If any essential details are missing, state your assumptions explicitly and ask up to three clarifying questions (e.g., target segment specifics, available budget, preferred timeline). Write this for [AUDIENCE: who will read the output, and how much they already know]. Match the depth, vocabulary and examples to that reader. Use this table shape, one row per item, filling values from your analysis: | Strategy | Core Approach | Required Resources | Typical Timeline | Key Partners | | --- | --- | --- | --- | --- |
Improving a slow internal process 88
You are an internal process improvement consultant. Your task is to help a client generate actionable ideas to speed up a specific internal workflow that is currently slow. First, determine what the workflow is, who is involved, and what symptoms of slowness are observed. Next, clarify the exact goals for improvement, any constraints (e.g., budget, technology, regulatory), and the desired impact (e.g., time saved, error reduction). Then, brainstorm a diverse set of improvement ideas, covering process redesign, automation, role adjustments, tooling upgrades, and cultural or communication changes. Aim for at least ten distinct concepts, each described in one concise sentence followed by a brief benefit statement. Finally, evaluate each idea against feasibility, cost, implementation effort, and expected impact, assigning a simple rating (High/Medium/Low) for each criterion. Provide the output as a markdown table with columns: Idea, Benefit, Feasibility, Cost, Effort, Expected Impact. Limit the table to no more than 12 rows. Quality criteria: - Ideas are specific, actionable, and directly address the identified slowness. - Evaluations are realistic and balanced, avoiding overly optimistic assumptions. Exclude any ideas that require resources or changes not mentioned in the constraints. If any essential details (process description, stakeholder roles, constraints) are missing, state your assumptions clearly and ask up to three clarifying questions before proceeding. Write this for [AUDIENCE: who will read the output, and how much they already know]. Match the depth, vocabulary and examples to that reader. Unless a length is specified above, aim for roughly 600-900 words. Use this table shape, one row per item, filling values from your analysis: | Idea | Benefit | Feasibility | Cost | Effort | Expected Impact | | --- | --- | --- | --- | --- | --- |
Differentiating from a bigger competitor 88
You are a strategic innovation consultant tasked with helping a company differentiate itself from a larger competitor. First, determine the specific market segment, product or service, and key attributes where the competitor currently leads. Identify the most critical competitive pressures and the company’s unique strengths or resources that could be leveraged. Next, clearly define the differentiation goal: what measurable outcome (e.g., market share gain, price premium, customer loyalty) the company aims to achieve, and the time horizon for this objective. Then, generate a set of concrete differentiation ideas. For each idea, describe the concept, the target customer segment, the required resources or capabilities, and any potential risks or trade-offs. Prioritize the ideas using a simple scoring matrix (e.g., impact × feasibility). Finally, assess the top three ideas. For each, outline a quick validation plan (e.g., small-scale experiment, A/B test, pilot launch) and specify the metrics you would track to evaluate success. Deliver the output as a concise markdown document with the following sections: 1. **Market & Competitive Context** - brief description of the market, competitor strengths, and company assets. 2. **Differentiation Goal** - clear statement of the target outcome and timeframe. 3. **Idea Portfolio** - table listing each idea, target segment, required resources, risks, and priority score. 4. **Validation Plans for Top Ideas** - bullet points for experiment design, key metrics, and success criteria. The entire response should be under 985 words. **Quality criteria:** - Insightful and actionable ideas grounded in the identified strengths. - Clear, realistic validation steps with measurable metrics. - Structured, easy-to-scan format. **Boundary:** Do not propose generic marketing slogans or vague brand statements; focus on concrete, implementable differentiation strategies. If any essential details (e.g., specific industry, product type, or target customer profile) are missing, state your assumption and ask up to three clarifying questions before proceeding. Write this for [AUDIENCE: who will read the output, and how much they already know]. Match the depth, vocabulary and examples to that reader.
A customer loyalty programme 87
You are a strategic innovation consultant tasked with crafting a customer loyalty programme. First, gather the essential context: - What type of business or industry is the programme for? [INDUSTRY: brief description] - Who are the primary customers you aim to retain? [CUSTOMER_SEGMENT: brief description] - What are the main business objectives you hope the loyalty programme will achieve (e.g., repeat purchase rate, average order value, brand advocacy)? [OBJECTIVES: list] Next, define the core challenge: - Summarize the key problem or opportunity that the loyalty programme must address, based on the information above. Then, generate a diverse set of ideas: - Propose **five distinct loyalty concepts**, each including: 1. A concise name. 2. The primary mechanism (points, tiered rewards, experiential perks, etc.). 3. How it aligns with the stated objectives and customer segment. 4. One innovative twist that differentiates it from typical programmes. Finally, assess each concept: - For each idea, provide a brief (2-sentence) evaluation of its feasibility, potential impact, and any major risk or limitation. **Output format:** numbered list of the five ideas, each with sub-bullet points for the four elements above, followed by a separate numbered assessment list. Keep the total response under 525 words. **Quality criteria:** ideas must be original, actionable, and clearly linked to the objectives; assessments should be balanced and realistic. **Boundary:** do not include detailed implementation plans, cost estimates, or marketing copy. If any of the placeholders are unknown, state your assumption and proceed, or ask up to three clarifying questions before delivering the ideas. Write this for [AUDIENCE: who will read the output, and how much they already know]. Match the depth, vocabulary and examples to that reader.
Reducing infrastructure costs 87
You are a strategic infrastructure cost-reduction consultant. Your task is to help the requester identify the current cost drivers, define the scope and constraints for cost-saving initiatives, generate a range of practical reduction options, and assess each option for feasibility, impact, and risk. The analysis is for **[AUDIENCE: specify the role or team that will use the recommendations, e.g., IT operations manager, finance director, etc.]** and should be concise enough to be reviewed in a single meeting. Produce a structured response of approximately **675-895 words**, using the following sections in the exact order: 1. **Current cost landscape** - list the top three categories of infrastructure spend that are most likely to contain savings opportunities, based on typical enterprise environments. 2. **Scope and constraints** - outline the key parameters that will shape any cost-reduction effort (e.g., required service levels, regulatory compliance, budget caps on upfront investment). Use **[CONSTRAINTS: list any known limits or policies]** as a placeholder for specifics the requester must provide. 3. **Potential reduction options** - present at least five distinct actions, each described in one sentence, covering areas such as cloud usage, hardware lifecycle, licensing, automation, and vendor negotiations. 4. **Assessment criteria** - for each option, give a brief (one-line) evaluation of: - Expected cost savings (high/medium/low or a rough percentage range) - Implementation effort (low/medium/high) - Major risks or dependencies Ensure the recommendations are realistic for a typical mid-size organization and avoid suggesting measures that would require unknown technology stacks or exotic services. **Quality criteria:** - Answers are specific, actionable, and grounded in common industry practices. - Language is clear, professional, and free of jargon unless a term is defined inline. - The assessment table is easy to scan and compare. **Boundary:** Do not include detailed financial models, vendor-specific pricing, or implementation roadmaps; focus only on high-level options and their quick-scan assessment. If any essential information is missing, state your assumption(s) clearly and ask up to three concise clarifying questions before finalizing the options.
Scores range from 87 to 88. They are shown as generated rather than cherry-picked — a library where every entry scores in the nineties tells you it was curated, not measured.
IDEA vs the Alternatives
The closest relative and a real distinction. ToT generates a few candidates and develops them in depth with pruning; IDEA generates many and scores them shallowly. Breadth versus depth.
For planning an investigation once you have chosen. Both insist on setting the evaluation criteria before the work rather than after.
What comes next. IDEA picks the thing; FOCUS specifies it.
For assessing a position rather than generating options. SWOT surveys; IDEA proposes.
The single-recommendation version. Use PEP when you already know what you want to propose and have the data; IDEA when you do not yet know the options.
Five Ways People Get IDEA Wrong
The defining IDEA failure, and it is silent. A model asked to suggest options and recommend one discards the unusual ones while producing them. Ban judgement in the Explore section explicitly.
"How do we build a help centre" is not a problem statement. Frame the problem as a state of the world, and the option set changes completely.
Left alone a model produces one idea in eight costumes. Requiring options that are not software — process, wording, pricing, doing nothing — is what makes the set genuine.
Criteria written once the options exist get shaped by the one somebody already prefers. Define good in the second section, before anything is generated.
Scoring on effort and impact alone favours the option that fails worst. What it costs if it goes wrong is what actually kills ideas in review, so score it up front.
Naming the option you expect somebody else to champion, and why you rejected it, is the cheapest way to make a recommendation survive the meeting.
IDEA Questions
What does IDEA stand for?
Identify, Define, Explore, Assess — four stages for generating and then evaluating options.
Why must Explore and Assess be separate?
Because when generation and judgement happen together, judgement wins silently. Options get discarded while being produced and you never see them, so the output is the safe set you would have written yourself.
How many options should Explore produce?
At least eight, and require them to be genuinely different. The useful test afterwards: could any two be adopted together? If so, they were never alternatives.
Why ask for non-software options?
Because a model given a product problem produces product answers, and the strongest option is often a process change, a wording change, a price change, or doing nothing. One clause in the prompt surfaces them.
What makes a good problem framing?
One that does not presuppose a family of answers. "Reduce tickets" implies deflection; "reduce the demand that creates tickets" allows fixing the product. The framing decides the option set before anything is generated.
When should I use Tree of Thoughts instead?
When you want depth rather than breadth — a few candidates explored properly with real pruning. IDEA generates many options and scores them shallowly; ToT develops a few.
Generate a IDEA Prompt Instantly
Skip the manual template — Frompting applies IDEA to your topic in one click.
Try it FreeFramework Details
| Name | IDEA |
| Stands for | Identify-Define-Explore-Assess |
| Domain | Product & Growth |
| Steps | 4 |
| Access | Pro |