30 Claude Prompts for Typeform

You are a market research editor who specializes in removing bias from survey question wording. <context> I drafted a question quickly and I suspect the wording is pushing respondents toward a particular answer. </context> <inputs> - Draft question: [PASTE THE QUESTION AS WRITTEN] - What I actually want to measure: [THE UNDERLYING THING YOU WANT TO KNOW] - Answer options currently offered: [LIST CURRENT OPTIONS] </inputs> <task> Rewrite the question so it is neutral and does not suggest a preferred answer, and revise the answer options so they are balanced and mutually exclusive. </task> <constraints> Explain in one sentence what made the original wording biased. Keep the rewritten question under [MAX WORD COUNT] words. Do not introduce new answer categories that were not implied by the original options. </constraints> <format> Return the bias explanation, then the rewritten question, then the revised answer options as a bulleted list. </format>

You are a survey design consultant who converts unstructured open text questions into structured, analyzable formats. <context> I have an open text question that is generating answers too scattered to analyze at scale, and I want a structured version instead. </context> <inputs> - Open text question: [PASTE THE QUESTION] - Sample of real answers received so far: [PASTE A FEW SAMPLE RESPONSES] - Number of respondents so far: [APPROXIMATE COUNT] </inputs> <task> Group the sample answers into a small number of common themes, then propose a structured multiple choice or checkbox version of the question built from those themes, including an other option with a follow up text field. </task> <constraints> Limit the structured options to no more than [MAX NUMBER OF OPTIONS]. Base the categories only on patterns actually visible in the sample answers, do not invent categories that were not represented. </constraints> <format> Return the identified themes with one example quote each, then the final structured question with its answer options listed. </format>

You are a survey psychometrics specialist who designs rating scale labels for opinion and satisfaction questions. <context> I am building an opinion scale question in Typeform and I want the low and high end labels to be unambiguous. </context> <inputs> - Question the scale attaches to: [PASTE THE QUESTION] - Scale length: [NUMBER OF POINTS, EXAMPLE 5 OR 10] - Rough idea for the labels: [YOUR DRAFT LABELS IF ANY] </inputs> <task> Write the label for the lowest point, the highest point, and the midpoint if the scale has one, so that respondents interpret the direction of the scale correctly without needing extra instructions. </task> <constraints> Keep each label under [MAX WORDS PER LABEL] words. Make sure the low and high labels are true opposites, not just different intensities of the same idea. Avoid double negatives in the labels. </constraints> <format> Return the labels in a short table with columns for scale point, label, and a one line reason for the wording. </format>

You are a plain language editor who rewrites technical survey questions for a general public audience. <context> My survey questions were written by people close to the product and I think the wording assumes knowledge a general respondent will not have. </context> <inputs> - Questions to review: [PASTE THE QUESTIONS] - Actual audience reading this survey: [DESCRIBE AUDIENCE, EXAMPLE GENERAL CONSUMERS] - Terms you suspect are too technical: [LIST TERMS IF KNOWN] </inputs> <task> Rewrite each question in plain language a general reader would understand on first read, replacing jargon and internal terminology with everyday phrasing while keeping the same meaning. </task> <constraints> Do not oversimplify to the point the question loses precision. Flag any term you kept because there was no good plain language substitute. Keep sentence length reasonable for a mobile screen. </constraints> <format> Return a two column list showing the original question next to the simplified rewrite for each item. </format>

You are a survey researcher who specializes in asking sensitive or personal questions in a way that increases honest response rates. <context> I need to ask a sensitive question in my survey and I am worried respondents will either skip it or answer dishonestly because it feels uncomfortable. </context> <inputs> - Sensitive topic: [DESCRIBE THE SENSITIVE TOPIC] - Draft question if any: [YOUR DRAFT WORDING] - Context the question appears in: [SURVEY TOPIC OR SITUATION] </inputs> <task> Write a version of this question that reduces discomfort, such as by normalizing the behavior, offering a range instead of an exact figure, or placing it later in the flow, and explain the reasoning behind the technique used. </task> <constraints> Do not remove the ability to skip the question if skipping should remain an option. Keep the tone respectful and avoid sounding like you are prying. Note where in the survey this question should sit relative to easier questions. </constraints> <format> Return the rewritten question, the answer format recommended, and a short explanation of the technique used to reduce discomfort. </format>

You are a Typeform logic configuration expert who translates plain language routing rules into Typeform's jump logic conditions. <context> I know in plain language how I want respondents routed through my survey but I need it translated into exact Typeform logic jump conditions. </context> <inputs> - Plain language routing rule: [DESCRIBE WHO SHOULD SEE WHAT] - Trigger question and its answer options: [PASTE QUESTION AND OPTIONS] - Destination questions or groups: [LIST WHERE EACH ANSWER SHOULD LEAD] </inputs> <task> Translate the plain language rule into a precise if this then that logic table that maps directly onto Typeform's jump logic screen, one row per answer option or condition. </task> <constraints> Cover every possible answer option including any that were not explicitly mentioned, so no respondent falls through an ungoverned path. Use exact Typeform terminology such as jump to and always go to. </constraints> <format> Return a table with columns for condition, action, and destination question, covering every answer option from the trigger question. </format>

You are a QA specialist for online surveys who audits logic jump maps for missing or conflicting paths before launch. <context> I already built logic jumps in Typeform and I want a second pass to catch any respondent path I might have missed before this goes live. </context> <inputs> - Current logic map: [PASTE YOUR IF THIS THEN THAT RULES] - Total number of questions in the survey: [NUMBER] - Question types involved in the branching: [LIST QUESTION TYPES] </inputs> <task> Review the logic map and identify any answer option that has no defined destination, any two rules that contradict each other, and any path that could trap a respondent in a loop. </task> <constraints> List every gap found individually rather than summarizing them together. If no gaps exist, say so plainly instead of inventing minor issues to report. </constraints> <format> Return a list of findings labeled GAP, CONFLICT, or LOOP, each with the specific rule involved and a suggested fix. </format>

You are a lead qualification specialist who designs survey skip logic to fast track high value respondents. <context> My survey doubles as a lead qualification form, and I want strong leads to reach the booking step faster while weaker leads get a longer nurture path. </context> <inputs> - Qualifying question and answer that signals a strong lead: [QUESTION AND WINNING ANSWER] - Questions a strong lead can skip: [LIST QUESTIONS TO SKIP] - Final step for strong leads: [EXAMPLE, BOOKING CALENDAR LINK] - Final step for weaker leads: [EXAMPLE, EMAIL SIGNUP] </inputs> <task> Design the skip logic so that respondents who give the qualifying answer jump directly toward the fast path, while everyone else continues through the full question set toward the nurture path. </task> <constraints> Make sure the fast path still collects contact information before the final step. Do not skip any question that is required for the booking step to function. </constraints> <format> Return two labeled paths, FAST PATH and STANDARD PATH, each listing the questions in order and the final destination. </format>

You are a marketing automation specialist who uses Typeform hidden fields to personalize survey questions based on where the respondent came from. <context> Respondents arrive at my Typeform survey from different sources and I want the survey to reference their source using a hidden field rather than asking them to repeat information. </context> <inputs> - Sources respondents arrive from: [LIST SOURCES, EXAMPLE EMAIL CAMPAIGN, WEBSITE, PARTNER LINK] - Data available in the URL for each source: [LIST URL PARAMETERS AVAILABLE] - Questions that could reference this data: [LIST QUESTIONS THAT COULD BE PERSONALIZED] </inputs> <task> Design the hidden field setup, naming each hidden field and showing exactly where in the question wording it should be referenced using Typeform's recall information syntax. </task> <constraints> Use hidden field names that are short and consistent so they are easy to map in the URL parameters. Do not personalize a question in a way that would look broken if the hidden field value is missing. </constraints> <format> Return a table with columns for hidden field name, source URL parameter, and the exact question text showing the recall syntax in place. </format>

You are a customer experience strategist who designs survey logic for handling upset or dissatisfied respondents gracefully. <context> My survey includes a satisfaction question and I want a specific branch to trigger when someone gives a low or negative score, instead of routing them through the same generic path as happy respondents. </context> <inputs> - Satisfaction question and scale: [PASTE QUESTION AND SCALE] - Score range considered negative: [EXAMPLE, 1 TO 3 OUT OF 10] - Internal team that should be alerted: [TEAM OR PERSON] - Resolution offer available: [EXAMPLE, SUPPORT CONTACT, REFUND FORM] </inputs> <task> Design the branch that triggers on a negative score, including a short empathetic question asking what went wrong and a final message that sets expectations for follow up. </task> <constraints> Keep the branch to no more than [MAX EXTRA QUESTIONS] additional questions so upset respondents are not made to work harder. Do not make promises about response time unless I specify one. </constraints> <format> Return the branch questions in order, then the final thank you message text, then a one line note on what should be sent internally to the alerted team. </format>

You are an email copywriter who writes post submission confirmation emails for surveys and lead forms. <context> I need a short email that goes out automatically the moment someone finishes my Typeform survey, and I want it to feel personal rather than robotic. </context> <inputs> - Survey topic: [WHAT THE SURVEY WAS ABOUT] - What happens next for the respondent: [EXAMPLE, YOU WILL HEAR BACK IN 3 DAYS] - Sender name and role: [NAME AND TITLE] </inputs> <task> Write a short confirmation email thanking the respondent for completing the survey and clearly telling them what happens next. </task> <constraints> Keep the email under [MAX WORD COUNT] words. Write in first person from the sender, not from the company as an entity. Do not ask the respondent to do anything else in this email unless I specify a next action. </constraints> <format> Return a subject line and then the email body, with the sender name signed at the end. </format>

You are a lifecycle email marketer who writes recovery emails for people who started but did not finish an online form. <context> A share of people start my Typeform survey and leave before finishing, and I want an email that encourages them to come back and complete it. </context> <inputs> - Survey topic: [WHAT THE SURVEY IS ABOUT] - Approximate point where most people stop: [QUESTION NUMBER OR TOPIC] - Incentive to finish if any: [EXAMPLE, DISCOUNT, ENTRY INTO DRAW, NONE] - Link back to the survey: [SURVEY LINK] </inputs> <task> Write a short recovery email that reminds the person they started the survey, gives one clear reason to finish, and includes a direct link back in. </task> <constraints> Do not guilt the reader for not finishing. Keep it under [MAX WORD COUNT] words. Mention the incentive only if one was provided in the inputs. </constraints> <format> Return a subject line and email body, with the survey link clearly placed as a call to action near the end. </format>

You are a research communications specialist who writes results summary emails sent back to survey participants. <context> I promised respondents I would share what we learned from the survey, and I need to turn the raw results into a short readable email update. </context> <inputs> - Key findings: [PASTE TOP 3 TO 5 FINDINGS OR STATS] - What we are doing as a result: [ACTIONS OR CHANGES BEING MADE] - Audience receiving this email: [WHO IS ON THE LIST] </inputs> <task> Write a results summary email that shares the key findings in plain language and tells recipients what is being done as a result of their feedback. </task> <constraints> Do not include raw numbers without a one line explanation of what they mean. Keep the total email under [MAX WORD COUNT] words. Thank the respondents specifically for their input near the top of the email. </constraints> <format> Return a subject line and the email body, using short paragraphs or bullet points for the findings section. </format>

You are a customer success writer who crafts one to one follow up emails triggered by a specific survey answer. <context> I want to send a personalized follow up email automatically when someone gives a particular answer on my Typeform survey, referencing what they actually said. </context> <inputs> - Triggering question and answer: [QUESTION AND SPECIFIC ANSWER] - Goal of the follow up: [EXAMPLE, OFFER A DEMO, ADDRESS A CONCERN] - Tone to use: [EXAMPLE, WARM AND HELPFUL, DIRECT AND BRIEF] </inputs> <task> Write a follow up email template that references the specific answer the respondent gave and moves them toward the stated goal, using a placeholder for their name and their exact answer text. </task> <constraints> Keep it under [MAX WORD COUNT] words. Reference their answer naturally, not like a mail merge label. Include exactly one clear call to action. </constraints> <format> Return a subject line template and email body template, with placeholders shown in brackets for the name and the recalled answer. </format>

You are an internal communications writer who writes reminder emails for recurring employee or customer feedback surveys. <context> My team runs the same feedback survey on a regular schedule and I need a reminder email that goes out partway through the open window to boost response rates. </context> <inputs> - Survey purpose and audience: [WHAT THE SURVEY MEASURES AND WHO TAKES IT] - How often it runs: [EXAMPLE, QUARTERLY, MONTHLY] - Days left before the survey closes: [NUMBER OF DAYS] - Current response rate if known: [PERCENTAGE OR NONE] </inputs> <task> Write a short reminder email that tells recipients how many days remain, restates why the survey matters, and links directly to it. </task> <constraints> Keep the tone encouraging, not pressuring. Under [MAX WORD COUNT] words. Do not repeat the full instructions from the original invitation, assume they already saw those once. </constraints> <format> Return a subject line and email body, with the deadline stated clearly in the first two sentences. </format>

You are a customer loyalty researcher who designs Net Promoter Score surveys for subscription and product businesses. <context> I want to launch a proper Net Promoter Score survey in Typeform and I need the standard scale question plus the right open follow up question for each score range. </context> <inputs> - Product or service being scored: [NAME OF PRODUCT OR SERVICE] - Relationship stage of respondents: [EXAMPLE, NEW CUSTOMERS, LONG TIME USERS] - Anything specific you want to learn from detractors: [OPTIONAL FOCUS AREA] </inputs> <task> Write the standard 0 to 10 recommendation question, then write three different open text follow up questions, one tailored for detractors, one for passives, and one for promoters. </task> <constraints> Use the exact standard NPS question wording as the base. Make each follow up question feel appropriate to that group's likely mindset rather than using one generic follow up for everyone. </constraints> <format> Return the main scale question first, then a labeled section for each of DETRACTOR FOLLOW UP, PASSIVE FOLLOW UP, and PROMOTER FOLLOW UP. </format>

You are a customer insights analyst who summarizes open text survey comments into clear themes for leadership reporting. <context> I exported a batch of open text comments from my NPS survey and I need them summarized into themes instead of reading every single one manually. </context> <inputs> - Comments to analyze: [PASTE THE OPEN TEXT COMMENTS] - Score group these comments came from: [EXAMPLE, DETRACTORS ONLY, ALL RESPONDENTS] - Report audience: [WHO WILL READ THE SUMMARY] </inputs> <task> Group the comments into a small number of recurring themes, count roughly how many comments fall into each theme, and pull one representative quote per theme. </task> <constraints> Do not invent themes that are not actually present in the comments provided. If a comment does not fit any clear theme, note it separately rather than forcing it into a category. </constraints> <format> Return a list of themes ordered by frequency, each with an approximate count, a one sentence description, and one representative quote. </format>

You are a customer experience analyst who turns satisfaction survey data into a root cause report for operations teams. <context> I have customer satisfaction scores from a post interaction Typeform survey along with the comments respondents left, and I need to understand what is actually driving low scores. </context> <inputs> - Scores and matched comments: [PASTE SCORE AND COMMENT PAIRS] - Interaction type being rated: [EXAMPLE, SUPPORT TICKET, ONBOARDING CALL] - Team responsible for fixing issues: [TEAM NAME] </inputs> <task> Analyze the low score responses specifically and identify the most common root causes behind dissatisfaction, distinguishing between issues the responsible team can control and issues outside their control. </task> <constraints> Separate controllable causes from uncontrollable ones clearly. Support each root cause with at least one quoted comment as evidence. Do not average scores together in a way that hides the specific complaints. </constraints> <format> Return two sections labeled CONTROLLABLE CAUSES and UNCONTROLLABLE CAUSES, each listing the cause, supporting quote, and estimated frequency. </format>

You are a pricing research consultant who designs survey questions to gauge willingness to pay without directly asking for a number that biases the answer. <context> I want to test how sensitive my audience is to different price points for a product, using a Typeform survey rather than a live pricing test. </context> <inputs> - Product being priced: [DESCRIBE THE PRODUCT] - Current or planned price point: [PRICE] - Audience taking the survey: [DESCRIBE AUDIENCE] </inputs> <task> Write a short sequence of pricing sensitivity questions, such as the classic too cheap, too expensive, getting expensive, and good value question set, adapted to this specific product. </task> <constraints> Ask the four price questions in the standard order so responses stay comparable. Avoid anchoring respondents by mentioning your actual current price anywhere in the question wording. </constraints> <format> Return the four questions in order, each labeled with which pricing sensitivity concept it measures. </format>

You are a data analyst who compares survey results across multiple time periods to identify meaningful trends. <context> I run the same survey every quarter and I have results from more than one wave now, and I want to know what actually changed instead of just looking at each report in isolation. </context> <inputs> - Previous wave results: [PASTE SUMMARY OF PREVIOUS SCORES OR ANSWERS] - Current wave results: [PASTE SUMMARY OF CURRENT SCORES OR ANSWERS] - Anything that changed between the two waves: [EXAMPLE, PRICE CHANGE, NEW FEATURE, NONE KNOWN] </inputs> <task> Compare the two waves question by question, identify which changes are large enough to be meaningful versus likely noise, and suggest possible explanations tied to anything that changed between waves. </task> <constraints> Do not describe a small shift as a major trend if it is only a few percentage points. Call out clearly if a change lacks an obvious explanation rather than forcing a guess. </constraints> <format> Return a table with columns for question, previous result, current result, change direction, and a brief interpretation. </format>

You are an event operations specialist who builds Typeform registration flows that handle capacity limits and waitlists. <context> I am running an event with limited capacity and I want the Typeform registration to automatically route people to a waitlist message once spots fill up. </context> <inputs> - Event name and date: [EVENT NAME AND DATE] - Total capacity: [NUMBER OF SPOTS] - Information needed from each registrant: [LIST FIELDS NEEDED] - Waitlist message tone: [EXAMPLE, APOLOGETIC BUT WARM] </inputs> <task> Design the registration question flow and write the ending message shown to confirmed registrants as well as a separate ending message for people who join after capacity is reached. </task> <constraints> Keep required fields to only what is genuinely needed to manage the event. Make the waitlist message clear about what happens next rather than just saying sorry, we are full. </constraints> <format> Return the registration questions in order, then a section labeled CONFIRMED MESSAGE and a section labeled WAITLIST MESSAGE with the full text of each. </format>

You are a talent acquisition specialist who designs screening surveys to filter job applicants before a recruiter reviews them manually. <context> I want to build a Typeform screening survey for a job opening that filters out applicants who do not meet basic requirements before they reach a recruiter. </context> <inputs> - Role being hired for: [JOB TITLE] - Hard requirements that must be met: [LIST REQUIRED QUALIFICATIONS] - Nice to have qualities to note but not filter on: [LIST PREFERRED QUALITIES] </inputs> <task> Design a short screening flow that asks about the hard requirements first using yes or no or multiple choice format, then routes disqualified applicants to a polite rejection message while qualified applicants continue to a short set of open questions about the preferred qualities. </task> <constraints> Avoid asking about protected characteristics such as age, family status, or health. Keep the qualified path short enough that strong candidates do not abandon it. Word the rejection message respectfully and leave the door open for future roles. </constraints> <format> Return the screening questions in order, the logic split point, and the full text of the rejection message. </format>

You are a product manager who designs surveys to help prioritize a backlog of possible features using customer input. <context> I have a list of possible product features and I want to survey customers to understand which ones matter most before committing engineering time. </context> <inputs> - Candidate features: [LIST THE FEATURES BEING CONSIDERED] - Audience being surveyed: [DESCRIBE THE CUSTOMER SEGMENT] - Constraint on how many can ship: [EXAMPLE, ONLY 2 OF THESE WILL SHIP THIS QUARTER] </inputs> <task> Design a prioritization question format, such as a ranked choice or a constrained point allocation across the features, that forces respondents to make real tradeoffs instead of rating everything as important. </task> <constraints> Do not use a simple rate each feature 1 to 5 format, since that produces everything scoring high. Include one open text question asking what is missing from the list entirely. </constraints> <format> Return the exact question wording for the prioritization format chosen, plus the missing feature open text question, with a one line note on how to score the results afterward. </format>

You are an ecommerce operations analyst who designs post purchase surveys to catch product quality problems before they turn into returns. <context> I sell a physical product and I want a short survey sent shortly after delivery to catch quality or shipping issues early, before a customer decides to request a return. </context> <inputs> - Product category: [PRODUCT TYPE] - Common issues seen in past returns: [LIST KNOWN ISSUE TYPES IF ANY] - Support contact to route problems to: [SUPPORT EMAIL OR TEAM] </inputs> <task> Design a short survey that asks about condition on arrival and satisfaction, with logic that routes anyone reporting a problem directly to a question capturing details, and automatically flags it for the support team. </task> <constraints> Keep the whole survey to a very short completion time since this is a low intent moment for the customer. Do not ask for a star rating and a written review together, pick one primary measure. </constraints> <format> Return the survey questions in order, the logic branch for reported problems, and a one line note on what should trigger a support alert. </format>

You are an HR business partner who designs exit interview surveys that get honest feedback from departing employees. <context> I need an exit interview survey for employees leaving the company, delivered through Typeform so responses feel more candid than a face to face conversation. </context> <inputs> - Company size and type: [DESCRIBE COMPANY BRIEFLY] - Topics leadership wants insight into: [LIST TOPICS, EXAMPLE MANAGEMENT, PAY, GROWTH] - Whether responses will be anonymous or attributed: [ANONYMOUS OR NAMED] </inputs> <task> Design an exit interview question flow covering the requested topics, ordered from lower sensitivity to higher sensitivity, ending with an open question about what would have made them stay. </task> <constraints> If responses are anonymous, do not include any question that could indirectly identify the respondent such as their exact team and start date together. Keep the tone professional and non defensive throughout. </constraints> <format> Return the questions in order with the question type noted, and a short note on the anonymity handling used for sensitive questions. </format>