Exporting Form Submissions to CSV and Excel (And What to Do With Them)

Short answer: exporting submissions to CSV or Excel is standard, included on every BotForms plan from $29/month as of September 2026. The part that determines whether the export is worth anything happens earlier, when you decide which fields are constrained and which are free text.

The export button is not the interesting part of this. Almost every form tool has one, they all produce broadly the same file, and the file is almost always disappointing the first time somebody opens it with an actual question in mind.

That disappointment is a form design problem wearing a reporting costume.

The reason exports are unusable

Open a typical export from a form that has been running for a year and you will find one or two columns doing real work and eight columns of prose.

The prose columns are not the problem. The problem is that the columns you want to count are also prose. "How did you hear about us?" as a free-text field produces Google, google, Google search, googled you, GOOGLE, a friend, referral, Referral - Dave, and forty other values, all of which mean six things. No pivot table recovers that. You end up doing it by eye, once, and never again.

The rule is simple and it has to be applied when the form is built:

Fields you intend to count must be constrained. Dropdowns, radio buttons, checkboxes. A closed list you chose.

Fields you intend to read stay free text. Describe the problem. Anything else we should know. These are for humans, and any attempt to analyse them at scale is a project, not a formula.

Most forms get this exactly backwards — they constrain the fields nobody analyses and leave the analytical ones open because a dropdown felt restrictive. The cost lands a year later.

If you are building a form now, this is also the single most useful correction to make when a tool drafts your questions for you: generated field sets default to free text far too often, which is one of the specific pitfalls covered in what AI gets wrong when it writes form questions.

What a usable export contains

Beyond your own fields, four columns do most of the analytical work. Check your export has them before you build anything on top:

Column	Why it matters	Common failure
Submission timestamp	Volume over time, day-of-week and hour patterns	Stored in a locale-dependent format that reorders on import
Source or campaign	Which channel actually produces work, not just traffic	Never captured, so every lead looks equally good
Form name or ID	Comparing performance across several forms	One giant export with no way to split it
Status or outcome	Conversion rate rather than raw volume	Lives only in a CRM, so the export cannot answer it

The third and fourth columns are the ones businesses most often lack. Without an outcome field somewhere, an export tells you how many enquiries arrived and nothing about how many were worth having.

Four questions worth running every month

Once the file is clean, these four cover most of the value and take about twenty minutes:

Volume over time. Submissions per week, plotted. You are looking for a trend and for the weeks that break it, not for a number.

Submissions by source. Group by the source field. The interesting output is usually not the biggest source — it is the source with high volume and low conversion, which is where money is being spent on the wrong traffic.

Completeness rate. What share of submissions have enough information to act on without a follow-up call? A low rate points at a missing required field or an ambiguous label, and it is worth more than any conversion tweak because it compounds across every lead.

Time to first response. Timestamp of submission against timestamp of your reply, if you can get both. This is nearly always the largest single lever on conversion and nearly always unmeasured. The reason it matters is covered in speed to lead: SMS versus email notifications; the reason it goes unmeasured is that the reply timestamp lives in a different system to the submission.

For the metrics worth watching continuously rather than monthly, form analytics: the metrics that actually matter covers what belongs on a dashboard instead of in a spreadsheet.

Export versus connect

These get conflated, and they are different tools for different jobs.

Export is for analysis. Point-in-time, batch, ad hoc. You pull a file because you have a question. Its strength is that a spreadsheet can do anything; its weakness is that it is stale the moment it lands and nobody remembers to run it.

Connect is for operations. A submission arrives and something has to happen — a CRM record, a task, a notification, a row appended to a shared sheet. This has to be live, and doing it by export means doing it by hand. The routes into a CRM or an automation platform are covered in connecting form submissions to a CRM with Zapier, and the developer-side equivalents in form webhooks and API access.

A useful test: if a person is copying rows out of a spreadsheet into another system, you needed a connection and built an export.

The mechanics that trip people up

Three small things cause most of the wasted time.

Import the CSV, do not open it. Double-clicking a CSV hands every column to the spreadsheet's guesser. Phone numbers lose leading zeros, long IDs become scientific notation, and dates rearrange according to the machine's locale. Use the import flow and set the phone, ID and postcode columns to text explicitly.

Filter before exporting. Most submission views let you filter by date range, form and status. A file containing only the rows the question needs is faster to work with and less of a liability sitting in a downloads folder. On BotForms this is what the filtering and bulk actions in the submissions view are for — see features.

Name the file with the range. submissions.csv, submissions(1).csv and submissions_final.csv are the natural end state otherwise, and nobody can tell you what period any of them cover.

Exports are the loosest copy of your customer data

Worth saying plainly: an export is usually the least protected copy of everything a customer told you. It leaves a system with access controls and lands in a downloads folder that has none. It gets emailed to a colleague, forwarded to a contractor, and sits on a laptop for two years.

Keep them for the period the question needs, share them through something with access control rather than as mail attachments, and clear old ones on the same schedule you would clear anything else containing customer details. If a formal data protection regime applies to you, the retention and handling specifics are a question for someone qualified in your jurisdiction. General information here, not legal advice.

If the answer is a list of people, not a number

One case where the export is the wrong instinct: when what you actually want is to contact everyone who submitted. Downloading a CSV and pasting addresses into a mail tool is the manual version of a thing most platforms do directly — submissions can flow into an audience and be mailed from there, which keeps consent and unsubscribes attached to the record rather than to a spreadsheet. That path is covered in turning form submissions into an email audience. On BotForms, audiences run from 500 contacts on Starter to 25,000 on Business, with campaign sending on Pro and above.

Getting started

Pick your highest-volume form and export the last ninety days. Before analysing anything, count how many of its columns are free text that you wish were dropdowns. That number is your work list, and changing three fields will do more for next quarter's reporting than any amount of spreadsheet effort on this quarter's file.

CSV and Excel export is included on every BotForms plan, starting at $29/month with a 30-day trial. Plan limits and the current add-on list are on pricing; if you are weighing the entry tier against alternatives, the cheapest AI form builder options for a small business compares them. Otherwise, start a trial.