9
7 Comments

How to automatically score partnership opportunities and draft outreach with AI

A good partnership can bring revenue, users, or distribution. The challenge is deciding which opportunities are worth your time — whether you find them yourself or they come to you.

Here’s how to build an AI workflow that scores each opportunity and drafts the next email when one is needed.

The tools

  • Jotform: to record each partnership opportunity in one standard form.
  • Google Sheets: to store opportunities, scores, decisions, and results.
  • Zapier Professional or higher: to connect the tools and send opportunities down different paths. Zapier’s Paths and AI by Zapier are not available on its Free plan.
  • AI by Zapier: to score the information you provide about audience fit and write an outreach email or response.
  • Gmail: to create a draft for you to check before sending.
  • Calendly: to give qualified partners one booking link.

Step 0 — Decide what makes a good partnership

Before AI reviews partnership opportunities, decide how you'll score them.

Here's an example scoring system you could use:

  • Audience fit (0–5)
  • Potential value (0–3)
  • Clarity (0–2)

Add the scores together.

  • 8–10: Strong fit
  • 5–7: Needs more information
  • 0–4: Pass

You can also set minimum requirements for different types of partnerships. For example, a sponsorship and a software integration don't need the same budget. If you use different requirements, add them explicitly to the AI prompt.

Step 1 — Create the opportunity form

First, create a short internal form that you can complete whenever you identify a potential partnership.

Open Jotform, create a new Classic Form, and add these fields.

Required
These fields give AI enough information to score most opportunities.

  • Opportunity source
  • Name
  • Work email
  • Company or brand
  • Website
  • Audience size
  • Describe their audience
  • Partnership type
  • Why their audience may overlap with yours
  • Proposed collaboration
  • What can you offer them?
  • What do you hope to receive?

Recommended

  • Budget or estimated commercial value
  • Target customer
  • Timeline
  • What result do you want?

Optional
Only add these if you think they'll be useful later.

  • Supporting notes
  • Supporting link

For Opportunity source, use these options: Discovered by us, Warm introduction, Inbound request.

For Audience size, use these options (or something similar): Under 1,000, 1,000–10,000, 10,001–50,000, More than 50,000, Unknown

For Partnership type, use (or something similar): Sponsorship, Affiliate deal, Content swap, Webinar, Product bundle, Integration, Newsletter promotion, Other

For Budget, do not use a blank text box. Use (or something similar): No cash budget, Under $500, $500–$1,499, $1,500–$4,999, $5,000+, Unknown

Structured answers are easier for AI to understand than free-text responses.

Make sure you complete the important fields before submitting the form. Open each field's settings:

  • Go to General
  • Enable Required

Make Work email required if you want Zapier to create a Gmail draft automatically. Leave Supporting link optional.

Because this is an internal form, you don't need to send a confirmation email after each submission. When someone contacts you about a partnership, simply enter the details into the form yourself. If important information is missing, the workflow can draft an email asking for it.

Now, every partnership opportunity is stored in the same format, making it much easier for AI to review.

Step 2 — Create your tracking sheet

Open Google Sheets and create a spreadsheet called Partnership Pipeline.

Rename the worksheet Requests.

Add these headers to the first row:

Date Added, Opportunity Source, Name, Email, Company, Website, Partnership Type, Audience, Audience Size, Audience Overlap, Target Customer, Proposed Collaboration, What We Can Offer, What We Hope to Receive, Budget or Estimated Value, Timeline, Expected Result, Supporting Notes, Supporting Link, Audience Score, Value Score, Clarity Score, Total Score, Classification, Reason, Draft Subject, Draft Email, Final Decision, Outreach Sent, Reply Received, Outcome.

Leave the headers in the first row. Zapier uses them to save each value in the correct column.

Update the Final Decision and Outcome columns after you've reviewed an opportunity. As you send emails and receive replies, update Outreach Sent and Reply Received.

You now have one place to see what was pursued, passed on, and successful.

Step 3 — Connect Jotform to Zapier

Before you connect Jotform to Zapier, submit one test opportunity through your form.

Complete the form using a realistic example. It could be a founder you want to approach, a warm introduction, or an inbound proposal. Make sure every field is filled in. This gives Zapier some sample data to use during setup.

Then create a new Zap and use these settings:

  • Trigger app: Jotform
  • Trigger event: New Submission

Then connect your Jotform account, select your form, and click Test trigger.

If the setup is correct, you'll see the data from your test submission.

Step 4 — Score and classify the opportunity

Now, let AI apply your scoring rules and classify the opportunity.

  • Click + under the trigger.
  • Search for AI by Zapier and select it.
  • Choose the Analyze and Return Data action.

Paste this prompt — or something similar:

You evaluate partnership opportunities for [BUSINESS NAME].

  • Our product is: [DESCRIBE YOUR PRODUCT]
  • Our audience is: [DESCRIBE YOUR IDEAL AUDIENCE]
  • Our accepted partnership types are: [LIST YOUR ACCEPTED TYPES]

Score the opportunity:

Audience fit: 0-55 = direct match3 = some meaningful overlap0 = no clear overlap

Potential value: 0-33 = clear value for both sides1 = value is unclear or one-sided0 = no clear value

Clarity: 0-22 -- The collaboration, expected result, and next step are clear1 -- One of those elements is missing or unclear0 -- Two or more are missing, or the idea is too vague to assess

Total score = audience fit + potential value + clarity.

Classify it:

  • STRONG FIT: 8-10
  • NEEDS INFO: 5-7 or key details are missing
  • PASS: 0-4, the type is not accepted, or audience fit is 0-2

Return:

  • Each score
  • Total score
  • Classification
  • Short reason

Write the appropriate next email based on the opportunity source.

For STRONG FIT:

  • If Opportunity Source is DISCOVERED BY US, draft a concise initial outreach email. Mention the relevant audience overlap, propose the collaboration, explain what we can offer, and suggest one simple next step.
  • If Opportunity Source is WARM INTRODUCTION, draft a concise follow-up that acknowledges the introduction, explains the possible collaboration, and suggests one simple next step.
  • If Opportunity Source is INBOUND REQUEST, thank them, respond to the proposal, and include this booking link when a call is appropriate: \[CALENDLY LINK\]

For NEEDS INFO:

  • If the opportunity is inbound, ask only for the specific information that is missing.
  • If we discovered the opportunity or received an introduction, draft a brief email that starts the conversation and asks only for the information needed to evaluate the idea. Do not make unsupported claims about their audience or results.

For PASS:

  • If the opportunity is inbound, draft a polite response explaining that it is not a fit right now.
  • If we discovered the opportunity or received an introduction, do not write an outreach email. Briefly explain why we should not pursue it.

Write a short subject line when an email should be drafted.

Use only the information provided. Do not invent facts or claims.

OPPORTUNITY

  • Source: [MAP FIELD]
  • Company: [MAP FIELD]
  • Contact: [MAP FIELD]
  • Audience: [MAP FIELD]
  • Partnership type: [MAP FIELD]
  • Audience overlap: [MAP FIELD]
  • Proposed collaboration: [MAP FIELD]
  • What we can offer: [MAP FIELD]
  • What we hope to receive: [MAP FIELD]
  • Budget or estimated value: [MAP FIELD]
  • Timeline: [MAP FIELD]
  • Expected result: [MAP FIELD]
  • Supporting notes: [MAP FIELD]

Replace each bracketed opportunity field by clicking inside the prompt, typing /, and selecting the matching Jotform field.

Under Settings → Output Fields, add:

  • audience_score — Number
  • value_score — Number
  • clarity_score — Number
  • total_score — Number
  • classification — Text
  • reason — Text
  • draft_subject — Text
  • draft_email — Text

Require the first six fields. Leave draft\_subject and draft\_email optional because some Pass results should not produce an email.

Click Preview.

Preview and test Strong Fit, Needs Info, and Pass examples before clicking Finish.

Step 5 — Save every opportunity

Next, save the opportunity and its scores in your tracking sheet.

In Zapier:

  • Click + and choose Google Sheets
  • Select Create Spreadsheet Row
  • Choose your Partnership Pipeline spreadsheet and the Requests worksheet

Map:

  • Opportunity Source into Opportunity Source
  • Jotform answers into the intake columns
  • audience_score into Audience Score
  • value_score into Value Score
  • clarity_score into Clarity Score
  • total_score into Total Score
  • classification into Classification
  • reason into Reason
  • draft_subject into Draft Subject
  • draft_email into Draft Email

Leave Final Decision, Outreach Sent, Reply Received, and Outcome empty.

Click: Test step.

If everything is working, a new row should be added to your spreadsheet.

Step 6 — Route each opportunity

Next, send each opportunity down a different path based on the AI's classification.

Add Paths to your Zap by clicking + → Paths.

Create three paths:

  • Strong Fit
  • Needs Info
  • Pass

Set each path to run when the Classification field exactly matches:

  • STRONG FIT → Strong Fit path
  • NEEDS INFO → Needs Info path
  • PASS → Pass path

In the Strong Fit and Needs Info paths, add Gmail → Create Draft.

Map:

  • To: Contact email
  • Subject: draft_subject
  • Body: draft_email

For the Pass path, add a filter first. Continue only if Opportunity Source is Inbound request. Then add Gmail → Create Draft and map the same fields.

This way, inbound requests receive a polite reply, while opportunities you discovered yourself don't create unnecessary drafts.

For now, create drafts instead of sending emails automatically. Review each one before you send it.

Step 7 — Create a review workflow

After you've collected enough partnership opportunities, create a second Zap to review the results. You could run it every month, every quarter, or after a major campaign.

  • Choose Schedule by Zapier → Every Month.
  • Select day 1 and your preferred time.
  • Add Google Sheets → Get Many Spreadsheet Rows (Advanced).
  • Select your Partnership Pipeline spreadsheet and Requests worksheet.
  • Include Date Added, Opportunity Source, Classification, Final Decision, Outreach Sent, Reply Received, Outcome, scores, and Reason.
  • Return all rows as one JSON value.

If you're reviewing only one month or one campaign, filter the sheet first. Otherwise, the report will include all rows. This action can return up to 1,500 rows.

Next, add AI by Zapier.

Ask it to review your partnership records and summarize what happened during the review period.The report should include:

  • How many opportunities were classified as Strong Fit, Needs Info, and Pass.
  • Which opportunity sources produced the best results.
  • Where your final decisions differed from the AI's recommendations.
  • Which outreach emails received replies.
  • Which partnerships led to successful outcomes.
  • Common traits of your best partnerships.
  • Any scoring rules that may need adjusting.
  • A few ideas to test before the next review.

Tell the AI to base its report only on the data provided and not to infer missing info.

Finish with Gmail → Create Draft.Send the report to yourself. Review the report before sending it.

Before you go live

A few things to keep in mind:

  • Don’t let AI make the final decision. In this workflow, it cannot confirm that a person’s audience size, budget, or claims are true.
  • Don’t ask AI to judge vague phrases such as “good opportunity.” Give it clear and numeric rules.
  • Don’t change your form fields after publishing the Zap without testing it again. Changed fields may need to be remapped.
  • Review the early results. For example, when promising opportunities keep landing in Needs Info, improve the form or adjust the threshold. When weak opportunities reach Strong Fit, tighten your audience definition.

The better your rules, the better your results.

on July 22, 2026
  1. 1

    One field I’d add before the AI scores an opportunity: an evidence level for each input. Audience size, traffic and partnership value are often self-reported, so the workflow should distinguish verified, externally observed and applicant-provided data rather than treating every number equally.

    I’d also split the outcome into two checkpoints: first qualified action (reply, intro, trial or booked call) and eventual result. That makes the monthly review useful even when a partnership has a long sales cycle. Otherwise the scorer can look accurate while the pipeline quietly optimizes for impressive-looking forms instead of outcomes.

  2. 1

    Solid pipeline, and the rigour drops off exactly where it matters most: those last three columns. Reply Received and Outcome are free text, yet that's the part that decides whether a partnership was worth doing. Most partnerships don't die at outreach. They die at the first "so how's this going" call two months in, when neither side can point to a number and both quietly stop investing.

    You've instrumented who to approach in fine detail and left "did it work" as a note to self. The highest-leverage field in that sheet isn't the AI score, it's a partner-specific tracking link or code you decide at deal time, before the first customer moves. Retrofit attribution after a partner has sent traffic for a month and you're guessing, and "I sent you loads of people" against an empty dashboard is how a good partnership turns into an argument.

    Score fit up front by all means. Just instrument the outcome with the same discipline, or you're optimising a funnel you can't see the bottom of.

  3. 1

    The scoring plus draft pipeline is solid in concept — the failure mode to watch for is optimizing the wrong signal.

    Partnership fit scores based on reach or audience overlap surface partners who look good on paper but have often already been approached by every competitor in your space. The ones who actually convert tend to be partners actively writing about a problem your product solves — they are already primed for the conversation before you reach out.

    A sharper pre-filter: before running the AI scorer, check whether the candidate mentioned a relevant pain point in their own recent content. That intent signal narrows a list of a thousand candidates to maybe forty, and the outreach draft can open with a specific reference to their own words. Reply rates improve meaningfully when the first line proves you read something they wrote rather than just scored them algorithmically.

    What data source is feeding your candidate list? That usually determines whether the intent signal is even there to work with at all.

  4. 1

    This is a great example of AI expanding the leverage of a solo founder.

    Partnerships are usually one of those areas where founders know the value, but don't have enough time to research opportunities and personalize outreach.

    Curious — after building this, do you think the biggest bottleneck for AI-powered BD is finding the right opportunities, or creating enough trust for people to respond?

  5. 1

    self-reported fields are where this gets risky imo. audience size and traffic are the exact numbers people inflate, so if the AI scores off the form answers you're mostly automating trust in whatever they typed. i'd feel better about it pulling one signal you can actually check (real traffic, engagement) and using the form just to route. are you scoring on anything verifiable or is it all applicant-provided?

  6. 1

    automating vetting partners is HUGE. thanks for this

Trending on Indie Hackers
How to rank #1 on ChatGPT? User Avatar 112 comments I built a startup-idea scanner. It just told me none of my 3,400 ideas are easy wins. User Avatar 66 comments I Tested Agenmatic for Finding Customers in Communities — Here’s What I Learned User Avatar 63 comments Building a Shopify bundles app for stores with real fulfillment: here's the wedge User Avatar 42 comments “I’ll just post on Upwork” is not a client strategy. Here’s what I built instead. User Avatar 40 comments I recorded myself using 200+ indie SaaS products cold. Here are the 7 conversion killers that keep showing up. User Avatar 32 comments