Airtable Prototype

What is an Airtable Prototype?

An Airtable prototype uses Airtable, a spreadsheet-database hybrid with forms and automations, as the back office of a marketplace or matching service. Two sides sign up through forms (say, home bakers and cafés, or tutors and parents), their answers land in tables, and you make the matches by hand inside the base. Customers see forms and emails. You see a tidy control room.

The point is to prove the hardest thing about any marketplace: that you can attract BOTH sides and that a match leads to a real deal. A custom platform with profiles, search and payments takes months. An Airtable base takes a weekend, and the questions it answers (who signs up, how fast you can match, whether matches turn into repeat business) are the same.

When should you use it?

Use it when:

• Your idea connects two groups: buyers and sellers, clients and freelancers, hosts and guests, mentors and mentees.
• Matching needs a few facts per person (location, budget, availability, skills) that fit into columns.
• You expect dozens to a few hundred participants in the first months, which fits the free plan.

Skip it when one side is already served well by a big platform, since your test then measures how well you compete with that platform’s reach. Leave it aside too when matching happens in real time (ride-hailing, instant delivery), because manual matching cannot keep up. And if your product is a list people browse without being matched, a Notion database prototype is simpler.

How do you do it, step by step?

1. Write down what a good match is. Three to five rules in plain words: “Same city. Café orders at least 10 cakes a week. Baker can deliver before 8:00 and holds a food safety registration.” These rules decide your fields, so do this before you open Airtable.

2. Create a base with three tables. Supply (one row per seller or provider), Demand (one row per request or buyer) and Matches. In Matches, add two “Link to another record” fields, one pointing to Supply and one to Demand, plus a single-select Status: Proposed, Introduced, First deal, Repeat, Failed. Linked records let you open a match and see both sides at once.

3. Build one form per side. In the Supply and Demand tables, add a Form view and drag in only the fields your match rules need, plus email and phone. Keep each form under 10 questions. Share the form links on your page or in messages. Airtable forms on the free plan carry Airtable branding, which is fine for a test.

4. Make working views. In Demand, a grid view filtered to “Status is empty”, called Needs match. Then a Kanban view in Matches, grouped by Status, so you see the whole pipeline at a glance. Sort by “Created time” so the oldest request is always on top.

5. Match by hand, fast. Every morning, open Needs match, filter Supply by your rules and create a Match record. Send the introduction from Gmail with a template: “Hi Sara, meet Café Noord. They want 12 carrot cakes a week, delivered before 8:00, and you’re 2 km away. Sara, Café Noord, over to you.” Our rule of thumb: match within 48 hours, or the asking side loses interest.

6. Follow up after 7 days. One message to both sides: “Did you two agree on anything?” Update the Status. This is the number that matters, and it only exists if you chase it.

7. Automate only the boring step. Once you have done 20 matches by hand, automate one thing, say a confirmation email when a form comes in. The free plan includes 100 automation runs per workspace per month and its built-in email action is limited, so a free Zapier connection to Gmail is a handy alternative.

Know the free plan limits before you start. Airtable’s plans overview lists 1,000 records per base and up to 5 editors on the free plan. For a test with a few hundred participants, that is plenty.

What does a worked example look like?

Say Mei wants to build a marketplace that connects home bakers with independent cafés in Utrecht that want local cakes. She builds the base on a Saturday: Bakers, Cafés and Matches, with forms for both sides.

She recruits bakers through two local baking groups and cafés by walking into 30 of them with a printed QR code to the form. In 4 weeks: 14 bakers and 9 cafés sign up. She proposes 11 matches, 7 lead to a tasting, 4 to a first order and 2 cafés reorder every week.

The base shows her something no interview could: cafés fill in the form in 3 minutes, but bakers take 2 days to reply to an introduction, because many bake in the evening and check messages late. The slow side is supply. Her next test focuses on fast-replying bakers and a “reply within 24 hours” promise on the café side.

Which free tools help?

What mistakes should you avoid?

• Building profiles nobody reads. If a field does not affect a match, delete it. Every extra question costs you sign-ups.
• Recruiting only one side. A base with 80 bakers and 2 cafés tests nothing. Recruit the scarce side first and pace the other.
• Automating before matching by hand. You do not know the match rules yet. Twenty manual matches teach you what to automate.
• Counting sign-ups as success. The question is whether matches turn into deals and repeat deals.
• Forgetting to follow up. Without the 7-day check, a match that went nowhere looks the same as a match that became a weekly order.

How do you know it worked?

Our rule of thumb for a go: at least 30% of introductions lead to a first deal, and at least a third of those deals repeat within a month. Also watch your time to match. If you need a week per match, the side you are short of is your real business problem.

No-go: you get sign-ups on both sides, yet introductions rarely lead to a deal. Interview the people who said no before you change the matching rules.

What should you do next?

Use your Status pipeline as a conversion funnel to find the step where matches die. If one group of buyers converts far better than the rest, narrow down with beachhead segment selection. When matching works, try charging one side with a form-based service fee. The startup game for women by Fe/male Switch walks you through testing a two-sided idea with an AI co-founder.

FAQ

Is Airtable free for a startup prototype?

Yes. The free plan covers a test like this: 1,000 records per base, up to 5 editors and 100 automation runs per month. You upgrade when you outgrow those limits, which is a good problem to have.

Can I build a marketplace with Airtable?

You can run the back office of one: sign-ups, matching, status tracking and simple automations. For customer logins, search and payments inside the product, you will need a front end tool or custom code later.

Airtable or Google Sheets for a matching prototype?

Google Sheets works for a single list. Airtable is better for matching because linked records let one Match row point to a person on each side, and forms and Kanban views come built in.

Updated 26 October 2026 · Written by the Fe/male Switch team

Mark this method as done