Import and Export: Easy to Arrive and Easy to Leave

When people choose software, they usually focus on what it does while they use it. Far fewer ask the question that matters most in the long run: what happens when I want to stop? How easily can I get my data out, and how easily can I bring it in from somewhere else?

For anyone building an alternative to an established product, import and export are not side features. They are the way you earn trust. This article explains why, and how to do them well.

Why they matter so much

Switching is hard

Wikipedia describes switching costs as the economic and psychological costs of moving from one product to another. They include the uncertainty about whether the new product will perform worse, the effort of setting up something new and the loss of what has been invested in the old one. For software, the largest part of that investment is usually data: years of documents, contacts, records and history.

If that data cannot move, switching becomes close to impossible, and the incumbent keeps its customers regardless of how good the alternatives are. If it can move, the market becomes fairer, and good products can win on merit.

Customers notice

People who evaluate alternatives look for signs of how a vendor treats them. A clear, honest export option says: we are not holding your data hostage. A clear, honest import option says: we want to make this easy for you. Both are signals of confidence and respect.

It is not only about business

For some data, there are legal considerations. The Information Commissioner's Office explains that the right to data portability allows individuals to obtain and reuse their personal data across different services, in structured, commonly used and machine-readable formats, and to ask for it to be transmitted directly to another provider where technically feasible, applying to data they provided where processing is based on consent or contract and carried out by automated means. If you handle personal data, check what applies to you and take advice. Even where there is no legal duty, the principle is a good guide to design.

Designing a good export

Make it complete

An export that leaves out half the data is worse than none, because it creates a false sense of safety. Decide what counts as the customer's data, and include all of it: content, structure, settings that matter, history, attachments and metadata such as dates and authorship. Document anything that cannot be exported, and why.

Use open and widely supported formats

Formats that many tools can read give customers real freedom. Plain text, CSV, JSON, standard document formats and open standards for the type of data in question are better than proprietary files that only your product opens. Wikipedia notes that open-source software is software whose source code is publicly available for others to use, study, modify and distribute, and that independence is one of its benefits. A similar spirit applies to data formats: the more open, the more independent your customers are.

Offer more than one format where helpful. A human-readable version and a machine-readable one serve different needs.

Preserve structure and relationships

Flat lists lose meaning. If your data has folders, tags, links between items, comments or versions, represent them in the export in a way that can be understood and, ideally, re-imported. Document the structure.

Make it self-service

Do not require customers to write to support or wait weeks. A button in the account settings, with a clear explanation, is best. For large exports, send a link when it is ready. Make sure it works for accounts that are about to close.

Keep it available

Do not hide export behind a paid plan or remove it when an account lapses. Allow a reasonable grace period after cancellation for customers to collect their data.

Secure it

Exports may contain sensitive information. Authenticate the person requesting it, expire download links and tell customers to store the file safely.

Test it

Run your own export and then try to use it. Open it in other tools. Re-import it into your own product. Check counts, dates, special characters and attachments. Many exports look fine and fail on real data.

Designing a good import

Support the formats people actually have

Start with the products your customers are leaving. Study their export formats, and build or document a path from each. Where there is no direct path, support common intermediate formats such as CSV.

Show a preview and a mapping

Before importing, show what will happen: how fields will map, what will be created, what will be skipped. Allow customers to adjust the mapping. This catches errors before they happen.

Validate and report

Check the data as it comes in. Report problems clearly: which rows failed and why. Do not silently drop data. At the end, give a summary: how many items were imported, how many were skipped and a way to download the list of exceptions.

Make it safe to try

Offer a test import into a sandbox or a way to undo an import. Customers will be more willing to try if a mistake is reversible.

Be honest about limits

Say what will not carry over. If comments, attachments or history will be lost, say so before the customer starts. This honesty prevents anger later and builds trust.

Handle scale

Customers with a lot of data need imports that do not time out. Process in the background, show progress and let people leave and return.

Document it

Documentation is part of the feature. Write plain guides that cover:

• What can be exported or imported, with formats.
• Step-by-step instructions with screenshots.
• What carries over and what does not, as a table.
• Common problems and fixes.
• Limits, such as file sizes and record counts.
• How to get help.

Date the documentation, and update it when formats or products change. A separate article describes how to write a switching guide.

Share the policy publicly

Put a short statement on your site: "You own your data. You can export it at any time in these formats. After you close your account we keep it for this long so you can collect it." This is simple, honest and persuasive. Link to it from your listing and your pricing page.

On this site, a maker can say in the listing's description how import and export work. The categories page groups products by purpose, the search page helps people find those that meet particular needs and the submit page starts a listing that makers then complete from their account.

Avoid dark patterns

Some products make leaving deliberately hard: hidden export buttons, exports that are incomplete, long delays, repeated retention pop-ups, threats of data loss. These tactics may reduce churn in the short run, but they damage trust, invite complaints and in some contexts may breach rules. If customers want to leave, help them. Many will return.

A worked example

A small company makes a contact manager as an alternative to a popular but expensive product. From the start, they treat import and export as core features.

Export: a button in settings produces a ZIP file with contacts in a standard format, notes as plain text, a list of tags and a readable summary page, plus a README explaining the structure. They test by importing the export into two other tools and fixing the issues they find. Customers who cancel have thirty days to download their data.

Import: they study the leader's export file and build a guided import that shows a mapping screen, flags duplicates and reports any rows that could not be read. A sandbox mode lets users try with a hundred contacts first. Their documentation includes a table of what carries over (names, emails, phone numbers, tags) and what does not (activity history, email threads), with dates.

Customers frequently mention that the honest table and the easy exit persuaded them to try. Several say that knowing they could leave is the reason they stayed.

A checklist

• Complete export, open formats
• Preserved structure
• Self-service and available after cancellation
• Secured and tested
• Import for the formats customers have
• Preview, mapping and validation
• Clear report of exceptions
• Safe to try, reversible
• Honest about limits
• Documented and dated
• Public data policy

Measuring whether your doors really are open

It is easy to believe your import and export are good because the buttons exist. Measure them. Track how many export requests succeed, how long they take and how often customers contact support afterwards with a problem. Ask customers who cancel whether they were able to take their data and whether it was usable. Run a yearly exercise in which someone outside the team tries to move a realistic account out and into another tool, then writes up what went wrong. The results are often humbling, and they give you a concrete list of improvements that customers will notice and talk about.

Frequently asked questions

Does making export easy lose me customers? Possibly a few, but it wins many more, and the ones who stay do so by choice.

Should I charge for export? Charging for access to a customer's own data tends to damage trust. Consider the effect carefully.

How often should I test import and export? With every release that touches the data, and at least quarterly.