How to Use AI to Analyze Form Submissions & Automate Follow-Up Work

Follow these steps to add the Cognito Forms app in ChatGPT:

1. Click your account banner and open Settings.
2. Click Apps and search for Cognito Forms.
3. Select Cognito Forms and click Connect.
4. Sign in to your Cognito Forms account if prompted.
5. Select your organization and authorize the connection.

Learn more about ChatGPT apps.

Follow these steps to add the Cognito Forms connector in Claude:

1. Click the Customize icon (looks like a briefcase) and select Connectors.
2. Click + Add Connector and search for Cognito Forms.
3. Select Cognito Forms and click Connect.
4. Choose your organization and click Authorize.

Learn more about Claude connectors.

Follow these steps to set up a custom MCP connection:

1. In your AI app, add a custom connector or MCP connection.
2. Enter the Cognito Forms server URL: https://mcp.cognitoforms.com/mcp
3. Set the OAuth Client ID: 366c9904-def2-451a-af2f-8e08d751088c
4. Log in to your Cognito Forms account when prompted.
5. Select your organization and click Authorize Organization.

If you have feedback forms or surveys with several open-ended text fields, these can take hours to read manually. Try asking the AI to analyze all responses across several submissions, so you can anchor your team review or planning sessions in customer feedback.

Try this prompt, customizing it with context on your business and your specific form’s name:

I manage customer service for a small business and want to understand what's frustrating our customers most. Review my Customer Feedback form submissions from the last 30 days. Group responses into core themes and rank them from most frequently mentioned to least frequently mentioned. To support you findings, give each theme a concise title, a 1 to 2 sentence description or summary of the issue, and include three relevant quotes taken verbatim from the submissions.

Start your day or week with an automated triage to surface what needs your attention or requires a human response before anything else. Adjust the prompt to match what matters most to you process (i.e. urgency language, missing fields, high-value indicators, or specific service types).

Try this prompt, changing the name of the form with your own:

Review the last two weeks of submissions on my Client Intake form. Flag any entries where the customer mentioned a tight deadline, expressed urgency, or left the budget field blank. List them by name, include their entry numbers, order them by urgency, and summarize what each one needs from us.

Before assigning work, sending proposals, or acting on a batch of new submissions, prompt the AI to catch gaps that can slow you projects down before they become a problem. You could use this prompt daily, weekly, or monthly depending on your submission volume and what cadence makes the most sense for you.

Try this prompt, customizing it with your desired time frame and the name of your form:

Review new submissions from this week on my Project Request form. Flag any entries that are missing a budget, have a deadline that's already passed, or describe a scope that seems inconsistent with the service type selected. List what's wrong with each one and what information I'd need to follow up. Include the entry number for each one.

Ask the AI to help you understand demand patterns, plan capacity, or identify process gaps. The more specific you are about what you’re looking for (i.e. time patterns, request types, geographic trends), the more actionable the output will be.

Try this prompt, customizing it with your role, what you’re analyzing, the form name, and time frame you want it to focus on:

I'm a service operations manager reviewing our intake trends. Look at the last 60 days of submissions on my Service Request form. Are there patterns in when requests arrive, like certain days of the week, times of month, or request types that spike? Summarize what you find and provide data to back up your findings.

Get a quick volume picture without exporting or counting entries yourself. Specify the output format upfront (such as a table, bullet list, bar chart, pie chart, etc.), so you get something ready to use rather than raw data you still have to reformat yourself.

Try this prompt, customizing it with the name of your forms and the output format you want:

Pull the last 30 days of submissions across my Sales Inquiry and Request for Quote forms. Show me how many came in each week, broken down by form. Format it as a bar chart showing the number of submissions for each form.

Use AI to help you better understand the distribution of your workload, like which service types are most common, which categories are growing, or where your team is spending most of their time. You could run this prompt quarterly for planning or monthly to catch shifts before they become resource problems.

Try this prompt, changing the name of the form, the time frame to surface, and your grouping criteria:

Show me a breakdown of all submissions on my Work Order form during Q1, grouped by service type. List each category, the total number of submissions, and what percentage of the total it represents. Format it as both a pie chart and a table.

Compare submission counts for different time periods to better understand the growth, seasonality, or impact of a process change of your online forms.

Try this prompt, customizing the form name, time frame, and output you want:

Compare submission volume on my Submit a Support Request form between April and May. Show me total submissions for each month and the week-over-week trend within each period. Format as a line chart and write a brief summary to go with it so I can share this in a team meeting.

Ask AI to combine a written summary with a few charts to help you present updates to your leadership or a client.

Try this prompt, customizing it with your form name, purpose, time frame, and criteria:

I'm preparing a brief update for my director on our company's reported incidents this quarter. Pull data from my Workplace Incident form for Q2. Summarize total submissions received, how many we received in each severity level, how many are resolved vs. still open, and any patterns worth flagging. Write it as a short one page document and include a bar chart for the quantitative values you found.

Instead of manually scanning multiple forms’ Entry Views to figure out what’s new or what needs your attention, ask AI to put this together for you at the start of your work day. Adjust the aging threshold and date window to match your team’s normal response time. For example, three days might be too long for urgent workflows and too short for complex projects.

Try this prompt, customizing:

Give me a quick pipeline check on my Service Request form. List everything that's currently open, grouped by status. Include the entry number for each entry mentioned. Flag anything that's been in the same status for more than three days, and anything where the requested completion date is within the next five business days.

Catch items that have quietly stopped moving, like approvals that were missed, reviews that have stalled, or requests that got buried. Run this prompt once a week to prevent things from sitting without anyone noticing until a client follows up.

Try this prompt, changing the form name, status name, and other details:

Review my Project Request form and show me any entries that have been in Pending Approval status for more than one week. List the entry number, client name, request type, and how long each one has been waiting. Sort by oldest first.

If you need a brief workload snapshot, but don’t need a full-blown project management tool, AI can help you with capacity planning or checking current assignments before assigning new work. If your form uses a Person field or a dropdown for assignment, mention that field name in the prompt for a more accurate result.

Try this prompt, customizing the form name, field name, and any other criteria you have:

Show me all open entries on my Project Request form, grouped by the team member assigned in the 'Assigned Member.' For each person, list how many open items they have and flag anyone carrying more than five active requests.

Set up a scheduled task in Claude or ChatGPT to run automatically every morning. The MCP Connector will pull your live Cognito Forms data at the time the task runs, so the briefing it gives you will always reflect current entries.

Try this prompt, customizing:

Every weekday at 8:00 AM, pull all new submissions from the last 24 hours across my Custom Order Request form and my Wholesale Orders form. Group them by form type, flag anything where the customer mentioned urgency or a deadline, include the entry number, and format it as a short daily brief.

Let AI help you close out your day with a clear picture of where things stand. This is especially useful if you manage high-volume submissions where things move throughout the day frequently and you want a quick recap or wrap-up without manual tracking.

Try this prompt, customizing:

Every weekday at 5:30 PM, review my Service Request form and give me a short end-of-day summary: how many new submissions came in today, how many entries changed status, how many are still open with no activity, and anything I should prioritize first thing tomorrow with the entry number included.

If you find yourself pulling numbers manually across forms or for a high number of submissions, replace that habit with a weekly meeting preparation prompt. Schedule it for whatever cadence and whichever day of the week you need, so it’s ready for you to quickly review before your team meeting.

Try this prompt, customizing:

Every Friday at noon, pull this week's activity across my Volunteer Sign-Up form and Sponsorship Application form. Show me: total new submissions, total resolved, anything still open from last week, and any submissions that came in but haven't been touched yet. Format it as a brief I can share in our Monday team meeting.

Make sure no new submission falls through the cracks with a weekly check on who hasn’t been contacted. This is especially useful if follow-up depends on one or two people staying on top of it.

Try this prompt, customizing with your form name, status name, and any other details you want to include:

Every Monday morning, review my Demo Request form and show me any submissions from the past seven days where the status is still Submitted and no note or update has been added. List the submitter's name, the entry number, what they requested, and when they submitted. I want to catch anyone who hasn't heard from us yet.

Let AI handle the manual task of transferring client information from a form submission into your CRM contacts. This prompt requires that you’re connected to your CRM’s MCP server. Be sure to mention specific field names on your form for the highest accuracy.

Try this prompt, customizing with your field names and specific details:

Pull the most recent submission from my New Client Intake form. Using the submitter's name, email, company, and service interest from the entry, create a new contact in [YOUR CRM TOOL]. Set the lead source to 'Intake Form' and assign it to the account manager listed in the entry.

Bridge your intake process with your team’s actual work with AI’s help. This prompt is especially useful if you receive a high volume of requests. Be sure to specify which ones to include by status, date, or service type so you’re not creating tasks for items already in progress.

Try this prompt, customizing with your unique criteria and details:

Pull all new submissions from this week on my Service Request form with a status of Submitted. For each one, create a task in [YOUR PROJECT OR TASK TOOL] with the client name as the task title, the request description in the task notes, and the requested completion date as the due date.

AI can help you catch leads who expressed interest but never made it into your follow-up process. Run this prompt weekly or after a marketing campaign. Both your Cognito Forms and CRM connections need to be active in your AI app at the same time for this to work.

Try this prompt, customizing with your CRM tool’s name and details to match your needs:

Pull all submissions from my Lead Inquiry form from the last 30 days. For each submitter, check [YOUR CRM TOOL] for a matching contact record using their email address. List anyone who filled out the form but doesn't have a contact record or open deal in the CRM yet.

Ask AI to pull information from both Cognito Forms and your project management or task tracking tool, so you can make decisions with data from both systems in one view.

Try this prompt, customizing the details to fit your tools, process, and needs:

I'm a service manager planning capacity for next month. Pull all open submissions on my Service Request form in Approved or Pending Assignment status. Then pull the current open task list from [YOUR PROJECT OR TASK TOOL] for my team. Based on what's already assigned, does my team have capacity to take on all pending service requests? Flag anyone who looks overloaded.

Let AI generate a template with field references and structure you can refine in Word. Once you’re satisfied, upload it to Cognito Forms and connect it to your Workflow so proposals generate automatically when a new intake is submitted.

Try this prompt, customizing it with the name of your form and what you want to be included in your document:

Using publicly available Cognito Forms documentation on Custom Document Generation, help me draft a Word document template for client proposals. The template pulls from my Client Intake form and should auto-populate the client's name, company, project description, requested timeline, and budget range. Include an introduction section, scope of work, timeline, and a signature block. Keep the tone professional.

AI can help you come up with a template structure for service agreements or contracts that pulls from your form data. Once finalized, you can have it sent to your team to review and make edits in Word or connect it to your Workflow so clients receive a pre-filled agreement automatically after submitting their intake form.

Try this prompt, customizing:

Using publicly available Cognito Forms documentation on Custom Document Generation, draft a service agreement template for my Client Onboarding form. Auto-populate the client's name, company, service type, start date, and monthly rate. Include a scope of work section with a placeholder for custom terms, a payment terms section based on the billing frequency selected, and a signature block for both parties.

If you want your form to generate invoices automatically, you can use AI to help you come up with the template that you’ll use with Cognito Forms’ Document Generation feature. Be specific about which fields map to which parts of the document — the more detail you give, the less cleanup you’ll need to do in Word before uploading it to Cognito Forms.

Try this prompt, customizing:

"Using publicly available Cognito Forms documentation on Custom Document Generation, draft an invoice template for my Order Request form. Include an itemized table that pulls product name, quantity, and unit price from a repeating section. Add a subtotal, a tax line at 8%, and a total. Set the due date to 30 days from submission. Include a header with a company name and logo placeholder. Be sure the colors and format look professional and match my brand.

Once the email and Cognito Forms MCP connections are both active, AI will check for matching emails (on whatever scheduled cadence you set up) and create the inquiry entries automatically. New leads hit your form pipeline without you manually transferring over each one.

Try this prompt, customizing it with:

I want to set up an ongoing task: whenever I receive an email in [your email tool] that looks like a new client inquiry (someone asking about services, pricing, or availability), pull the sender's name, email, and what they're asking about, and create a new entry in my Client Inquiry form. Flag the entry as 'Email Source' in the notes field.

Schedule an automated task to kick off your onboarding process for sales that closed within the last 24 hours, without anyone having to remember to create the entry. Pair it with Cognito Forms’ workflow automation to trigger the next steps (task assignments, client emails, document generation, etc.) the moment the entry appears.

Try this prompt, customizing:

Every morning, identify deals marked 'Closed Won' within the last 24 hours. Pull the client's name, company, email, service type, and contract start date from the deal record and create a new entry in my Client Onboarding form in Cognito Forms. Set the entry's status to 'Onboarding Initiated' and add a note with the deal ID for reference.

Run this at the end of your weekly report (Method 2) or as part of a scheduled task (Method 4). Instead of manually opening each overdue entry and adding a note, AI can handle the annotation across all of them at once, and give you a list so nothing falls through.

Try this prompt, customizing:

Pull all entries on my Project Request form that have been in 'Pending Approval' status for more than seven days. For each one, add an internal note: 'Follow-up required: flagged by weekly overdue report on [today's date]. Assigned reviewer has not responded.' Then give me a list of the entries you updated with entry numbers so I can follow up directly.

The MCP Connector is available on Pro, Team, and Enterprise plans. Actions performed through the connector use API requests, which count toward your organization’s usage. See Integration billing for details.

You also need Admin access to the Cognito Forms organization you want to connect.

Your AI assistant only has access to data you specifically request during a conversation. Your existing Cognito Forms user permissions apply and the connector respects the same access controls you’d have when logged in manually. Results from analysis prompts are also scoped to the default Entry View (the first View listed for the form), unless a different Entry View is specified in your prompt.

Any change made through the connector (such as updating an entry or changing form availability) is recorded in the Audit Log. Entry-level changes appear in that entry’s Audit Log and include the user who triggered the action and the service used. Read-only queries, like viewing or analyzing entries, are not logged. You can access the Audit Log from your entry or from Organization Settings.