Migrating Your Forms to a New Builder Without Losing Submissions

Short answer: rebuild the form, migrate the data separately, and run both platforms in parallel for a week. Export your historical submissions before you cancel anything — that is the only step that cannot be undone later. Rebuilding is the fast part; on BotForms a form can be generated from a description or extracted from a PDF of the old one, on every plan from $29/month.

People put off switching form platforms because they imagine a migration project. In practice there is no such thing as importing forms between builders — the formats are proprietary and nobody supports it — so the "migration" is three separate, small jobs that get conflated into one intimidating one.

Rebuild. Move the data. Swap the embeds. Only the second is irreversible, and it is the one most often left until last.

Do the export first

Before anything else, while the old account is fully active and paid:

Export every submission to CSV. All forms, all history, no date filter. This is the file you cannot recreate.

Download the attachments. Uploaded photos and documents usually are not in the CSV — they are links pointing at the old platform's storage, and those links die with the account. If your forms accepted uploads, this is a separate and larger download.

Screenshot the configuration. Notification recipients, conditional logic, redirect-after-submit URLs, integrations. Nobody remembers the confirmation email wording six weeks later, and it is easier to photograph than to reconstruct.

Do this before you build anything. Migrations stall, priorities change, and the account you meant to keep for another month gets cancelled by whoever is trimming subscriptions. Then the history is gone.

Once exported, the file is your least protected copy of customer data, so treat it accordingly — the handling points in exporting form submissions to CSV and Excel apply directly.

Rebuild rather than replicate

The building step is where AI generation genuinely earns its keep, because you are not designing from nothing — you know exactly what the form contains.

Three routes, roughly in order of speed:

Describe it. Write what the form is for, who fills it in, and list the fields. A generator produces a structured draft with sensible field types and validation attached. Minutes, not an hour.

Upload a PDF of the old form. If the old platform can export one, or the form started life on paper, extraction pulls the fields out directly — this is the route covered in converting a PDF to an online form, and it is the strongest option for long forms with many fields.

Build by hand. Reasonable for a five-field contact form, slow for anything larger.

Whichever route, budget a handful of AI runs per form for the iteration; how that allowance works and what a realistic month looks like is in AI runs explained.

Then — and this is the part people skip — do not reproduce the old form faithfully. Look at your export and count how often each optional field was actually filled. Fields with near-zero completion are fields to delete. Free-text fields you wish you could group are fields to convert to dropdowns, for the reasons in validation rules that stop bad data. A migration is the only moment where changing the form costs nothing extra, and rebuilding a mediocre form exactly wastes it.

The cutover, in order

Step	Action	What it prevents
1	Export submissions and attachments from the old platform	Permanent data loss
2	Build the new forms, editing rather than copying	Carrying over a bad form
3	Configure notifications, integrations and confirmations	Silent leads arriving nowhere
4	Submit five test entries yourself, on a phone	Discovering a broken field with a customer
5	Swap the embed on one low-traffic page	A site-wide outage from one mistake
6	Watch for a few days, both platforms live	Confirms real submissions arrive, not just tests
7	Swap remaining pages	—
8	Leave the old account live and empty for a fortnight	Catches the embed you forgot
9	Cancel the old subscription	—

Step 4 deserves emphasis. Test on a phone, not only a desktop, and check that the notification actually reaches a human — not that it was sent. The classic failure is a notification going to an address nobody monitors, which produces a form that appears to work perfectly and silently swallows every enquiry.

Step 8 is where the forgotten embeds surface: the landing page from last year's campaign, the link in an email signature, the QR code on a printed flyer. Watching the old platform for a fortnight tells you exactly which ones exist.

Embeds and URLs

Two different situations, often confused.

Embedded on your own site. The page URL does not change; only the embed code does. There is no SEO consideration and no redirect to set. The only risk is missing a page, which is what the parallel fortnight catches. Platform-specific embed details for the common site builders are in embedding a form on WordPress, Webflow and Squarespace.

Hosted form pages. If the old platform gave each form its own public URL, and those URLs were shared, printed or indexed, they need redirects to the new equivalents. Deleting them produces broken links in emails, on business cards and in search results — for as long as those artefacts exist, which is longer than you think.

The integrations are the sleeper task

Rebuilding a form takes twenty minutes. Rebuilding the things attached to it is the part that overruns.

Work through the list explicitly: CRM connections, automation platform connectors, webhook receivers, spreadsheet syncs, email tool handoffs. Each needs recreating and each needs testing with a real submission rather than a test event, because the field names have changed and a mapping built against the old names will fail quietly.

If you had a webhook receiver, the payload shape is different and your parser needs updating — the design points in form webhooks, API access and Zapier cover what to check. If your setup was purely a no-code connector, this is usually a rebuild of the same Zap against a new trigger, which is fast but not free.

Where the historical data should live

Once exported, decide deliberately where the history goes rather than leaving it in a downloads folder.

Most platforms do not import third-party submission history into their submissions view, and it is worth being clear-eyed about that rather than hunting for a feature that generally does not exist. The realistic options are keeping the CSV as an archive, loading it into your CRM or a spreadsheet where it can actually be queried, or accepting that history older than a certain point is reference material rather than working data.

If the reason you wanted the history was to contact those people, that is a different job with its own consent considerations — the mechanics of turning submissions into a mailable audience are in turning form submissions into an email audience.

Choosing what you are migrating to

Worth doing before any of the above rather than during it. The specifics of the common switches — what carries over, what does not, and what the pricing actually compares to — are covered per platform in the best Jotform alternative, Formspree alternatives for static sites and alternatives to conversational form tools.

The one thing to check that most comparisons skip: the export path out of the new platform. You are migrating because leaving somewhere was harder than it should have been. Confirm the next exit is easy before you commit, and if the vocabulary in the category is confusing the comparison, what a bot form actually is sorts out which products are which.

Getting started

Pick your simplest live form, run the full nine steps on it this week, and time it. That number tells you what the rest of the migration costs, and it is almost always smaller than the estimate that has been deferring the decision.

BotForms includes AI form generation and PDF-to-form conversion on every plan from $29/month, with a 30-day trial to run in parallel while you cut over. See features, pricing, or start a trial.