DEV Community

Wbcom Designs 💡
Wbcom Designs 💡

Posted on AI-assisted

Test the way out before you move in: an exit drill for any site builder, course platform or hosted tool

Most of us choose a tool by its front door. The demo is smooth, the first page goes live in an afternoon, and the pricing fits. Almost nobody walks round the back to check where the exit is.

Then one day the exit matters. Across the communities we follow, the stories arrive in different shapes but land in the same place:

  • A small business owner built three sites of well over a hundred pages each on an AI site builder, then found the builder would not hand the pages over in any form another system could use. Moving meant rebuilding, and the money already spent was simply gone.
  • A course creator in a billing dispute with their platform realised that their contacts, offers, videos and subscriptions all lived inside the account under dispute, and started asking what to export first, before anything else happened.
  • A teacher with dozens of courses and over a hundred students had to move off the LMS plugin everything was built on, and the question was not "what next?" but "what moves with me, and what do I retype?"
  • A developer community watched a much-loved hosted coding tool announce it was winding down. The reactions split between sympathy for the team, relief that the code would be opened later, and a sharper point: the tool sat on a runtime whose own future had just changed, so the risk was two layers deep.
  • Company owners who moved a site off a page builder, or off WordPress to plain static files, often in a day with an AI assistant, after agencies quoted far more. Some were thrilled. Others found the forms, search and redirects they had forgotten about a week later.

None of these people chose badly. They chose without testing the exit. So here is a drill we now suggest to clients and run ourselves before putting real work into any hosted tool. It takes an afternoon, and you do it in week one, not in year three.

Step 1: Write down what you would need to take with you

Not "the data". Be specific. For a website that usually means pages, posts, images, the URL of every page, forms and where their submissions go, redirects and email settings. For a course business it is students, progress and completions, certificates, purchases and active subscriptions, plus the lessons themselves. For a community it is members, their posts, who moderates, and any paid access.

If the list is longer than you expected, good. That list is the real size of the switch.

Step 2: Do a real export, not a theoretical one

Find the export button and press it now, with sample content. Then open what came out.

The questions that matter:

  • Is it a format something else can read? A spreadsheet of members and an XML or Markdown dump of content are useful. A PDF of your pages is a souvenir.
  • Does it include relationships? A list of students without which course they bought, or posts without who wrote them, is half an export.
  • Does it include the money side? Subscriptions often live with the payment processor tied to the platform's account. Ask whether recurring payments can move to a new system, or whether every subscriber has to sign up again.

A short exercise here saves a long argument later. If the export is poor, you have learned it while switching is still cheap.

Step 3: Check who holds the keys

Write down who owns the domain registration, the DNS, the email records, the payment account and the admin logins. The pattern we see again and again: the platform or an agency registered something "to make setup easier", and now the business has to ask permission to leave.

The rule is boring and it works: the organisation owns the accounts, and people or vendors get access. When the domain, the email records and the payment account are in your name, leaving a builder is a project. When they are not, it is a negotiation.

Step 4: Ask what happens to the URLs

Every page that earns search traffic has an address. If you move and those addresses change without redirects, you start again in search. So: can you export a list of every URL? Can the new home redirect old addresses to new ones? If the current tool uses addresses you cannot reproduce, that is a cost to note now.

Step 5: Look one layer down

The hosted coding tool story is the reminder here. Ask what your tool depends on: a runtime, a parent company, a plugin ecosystem, a single payment provider. You cannot control those layers, but you can know about them. A tool built on something that is itself changing is not automatically a bad choice. It is just a choice with two exits to plan for instead of one.

Step 6: Price the exit, then decide

Now put a rough number on leaving: hours to move content, redirects, re-inviting members, re-subscribing payers, retraining whoever edits. Compare that with what the tool saves you every month.

Sometimes the honest answer is "this tool is worth being locked in for". Lots of teams happily rent convenience. The difference is that they decided it with the exit price in front of them, instead of finding out during a dispute or a shutdown notice.

What goes wrong even with a drill

  • Exports drift. A tool that exported well when you joined may not after a redesign. Rerun the drill once a year, or when the vendor changes ownership or pricing.
  • The forgotten jobs. People who leave a CMS for static files often find it was quietly handling forms, search, spam filtering and scheduled posts. Add those to the step 1 list.
  • Subscribers do not move themselves. Plan the email that asks people to re-subscribe, and expect some of them not to.
  • Speed hides gaps. Moving a site in a day with an AI assistant is real. Checking that every form, redirect and email record still works takes longer, and that part is the one that matters.

The short version

Test the exit while it is cheap. Own the domain, the email records, the payment account and the member list. Keep a URL list. Know what your tool stands on. Then rent whatever convenience you like, with your eyes open.

Your turn

  1. Have you ever tried to leave a tool and found the export was not what you expected? What was missing?
  2. Which one thing do you now check before committing to any hosted builder, course platform or community tool?
  3. When subscribers or members had to move with you, how did you handle re-subscribing, and how many actually came along?
  4. Is there a tool you knowingly stay locked into because it is worth it? What makes that trade fair for you?

Top comments (0)