Switching From Tidio: A Practical Migration Guide
Most articles about leaving Tidio are lists of other tools. This one is the boring part nobody writes down: what to export, what quietly breaks, what to test, and how to run both widgets at once so you never drop a message mid-move.
The short version
- Half a day of work, spread over ten days. The hands-on part is two to three hours. The rest is watching.
- Export first, cancel last. Access to your history usually ends when the subscription does.
- Do not copy your bot flows. Rebuild the five that matter and delete the rest.
- Run both for a week. One extra month of the old bill buys away almost all the risk.
First, be clear about why you are moving
Almost everyone who searches for a way off Tidio is looking at an invoice, not a feature list. It is the shape of the bill rather than the size of it: three separate meters running at once.
The first is the plan itself, which is the predictable one. The second is Lyro, Tidio's AI, which is billed apart from the plan and starts at around $32.50 a month for 50 AI conversations. The third is the branding: removing the Tidio watermark from your own widget is a further add-on of roughly $16.67 a month on the lower tiers.
Two of those three move with how well things go. A busy week means more AI conversations and a bigger invoice. The thing you want, AI handling more of your inbox, pushes the cost up. That is what makes it hard to budget: you cannot forecast next quarter, because the number depends on traffic you have not had yet. On the free plan the 50 Lyro conversations are a one-off allowance rather than a monthly reset, so free AI stops for good once they are spent. We work through what that does to a real monthly bill in Tidio pricing explained.
On EzyConn, every plan has a monthly message allowance: 500 on the free plan, 5,000 on Starter and 25,000 on Professional. The difference is that the allowance is a number you can read on the pricing page before you pay, not a charge that appears after the month is over. Paid plans are Starter at $25 and Professional at $95 a month, AI runs on GPT-4o and Claude on every plan including the free one, prices are per workspace rather than per seat, and annual billing takes 20% off.
Step 1: write down what you actually use (30 minutes)
Do this before you look at any other tool. Answer six questions on one page:
- Which channels are actually live? Website chat, Messenger, email, WhatsApp?
- How many people reply to chats in a normal week, including weekend cover?
- How many bot flows fired at all in the last 30 days? Most teams find three or four out of twenty.
- Where is the widget code installed? Theme file, plugin, or tag manager?
- Who gets notified when a chat comes in, and where do they read it?
- What integrations are connected, and does anything downstream depend on them?
That page is your migration scope. Painful switches usually start with someone skipping it and rebuilding the wrong things.
Step 2: export while the account is still paid
This is the one irreversible step, so do it first and properly. What you can export depends on your plan. Get these five out:
- Conversation transcripts. Usually a CSV. You want them for continuity, and for spotting the questions your AI should be trained on.
- Contacts and email addresses. The real asset. Everything else can be rebuilt; a lost email list cannot.
- Canned and saved replies. The first thing your team will miss on day one.
- The help content your bot answered from. FAQ articles, policy text, shipping and returns wording.
- Screenshots of every bot flow. Flow logic does not transfer between vendors in any format, so a screenshot is the export.
Do not downgrade or cancel first to save a month. That is usually the point at which the history stops being reachable. The full checklist, including what happens to an unused Lyro allowance, is in how to cancel Tidio safely.
Step 3: rebuild, do not copy
A flow is a decision tree: if the visitor clicks this, show that. Teams end up with dozens of branches because a rule-based bot cannot read your website, so every answer gets spelled out by hand.
A modern AI chatbot is trained on your own pages instead, so most of those branches stop being necessary. Point it at your site, upload your help content, and ask it your twenty most common questions. Then rebuild only the flows that genuinely need to be flows: booking, order lookup, or routing to a specific team. Expect about a fifth of what you had.
Step 4: run both tools side by side for a week
This is the step most guides leave out, and the one that makes the difference. Two chat bubbles on one page look ridiculous, so split by section instead.
Put the new widget on a slice of the site where the stakes are lower but the traffic is real: your blog, your help pages, one product category, or one regional site. Leave the old widget serving everything else. A tag manager page-path rule does this in minutes and lets you change your mind just as fast.
Watch three numbers on both sides: conversations started, the share resolved without a human, and time to first reply. A week covers a weekday pattern and one weekend. The cost is one extra month of the old subscription, which is the cheapest insurance in this whole process.
The five tests worth running
| Test | Why it matters | A pass looks like |
|---|---|---|
| Widget loads fast on a phone | Most chat starts on mobile | Appears in under two seconds and never covers your checkout button |
| AI answers your five hardest questions | This is the entire reason you are paying | Correct answer, or an honest handover instead of a guess |
| Handover to a human | This is where switches quietly fail | The message reaches a real person within seconds, on the channel they watch |
| Out-of-hours message capture | After hours is where enquiries leak away | Email captured, confirmation sent, conversation waiting in the morning |
| A returning customer asks about an old order | Continuity of history | Your agent can find the imported transcript without leaving the inbox |
The handover test matters most, because it is where teams silently lose enquiries after a switch. Ask where your people genuinely spend the day. If the answer is Microsoft Teams or Slack, the chat should arrive there rather than in another dashboard someone has to remember to open. EzyConn replies work natively inside Microsoft Teams, Slack and Zoom, so nobody has to change where they work.
What actually breaks
- Links into old transcripts. Any URL you pasted into a CRM note or an internal doc stops resolving.
- Tag manager triggers and analytics goals. Anything listening for the old widget's events goes quiet without erroring, which is the worst kind of failure. Check your conversion goals on day one.
- Inline store data. If your old widget showed cart contents or order status in the chat, confirm the replacement does the same before you switch.
- Team habits. People keep the old tab pinned for weeks. Remove the app from phones and browsers on switch day.
- Reporting history. Your averages restart at zero. Screenshot your last 90 days of response times first, or you have nothing to compare against.
A realistic timeline
| When | What you do | Time |
|---|---|---|
| Day 0 | Audit and export | 1 hour |
| Day 0 | Install the new widget on a small section of the site | 30 minutes |
| Day 1 | Rebuild only the flows you genuinely use; run the test list | 1 to 2 hours |
| Days 2 to 8 | Both tools live. Watch the numbers, fix answers as you go | 10 minutes a day |
| Day 9 | Move the new widget site-wide, remove the old script | 20 minutes |
| Day 30 | Cancel the old subscription once you are sure | 5 minutes |
When to move from Tidio
If you are mid-way through an annual term, finish it: audit and export now, and switch at renewal when it is free to do so.
A cheaper invoice is not by itself a reason to move. Move when the shape of the bill stops matching the shape of your business, which usually means the month your AI usage became unpredictable. If you are weighing specific options, our Tidio alternative page sets out the comparison line by line, and the Shopify guide covers what an ecommerce store needs from a chatbot before it switches anything.
Questions people ask before switching
How do I migrate from Tidio to another live chat tool?
Work in this order: write down what you actually use, export your data while the Tidio account is still paid and active, install the new widget on a small part of your site, run both tools side by side for about a week, then remove the old script and cancel. The hands-on work is two to three hours. The week of overlap is just watching.
What should I export from Tidio before I cancel?
Five things: conversation transcripts, your contact list with email addresses, canned or saved replies, whatever FAQ and help content your bot was answering from, and screenshots of every bot flow. Export options differ by plan, so check what yours allows before you downgrade. Cancelling first is the classic mistake, because access to history usually ends with the subscription.
Can I run Tidio and a new chat tool at the same time?
Yes, and you should. Two chat widgets on the same page look messy, so put the new one on a smaller section of the site instead: your blog, your help pages, or one product line. Leave Tidio serving the rest. You pay one extra month of the old bill and remove almost all the risk of a silent gap in coverage.
How long does switching from Tidio take?
Plan for half a day of real work spread across roughly ten days. Installing a new widget and pointing the AI at your website takes under 30 minutes. The audit and export take about an hour. Rebuilding the handful of flows you actually use takes an hour or two. The rest of the calendar time is the side-by-side week, where you are watching rather than working.
Will I lose my Lyro chatbot flows if I switch?
Bot flows do not transfer between vendors, and you should not try to recreate them one for one anyway. Most flow branches exist to work around a bot that cannot read your website. A modern AI chatbot trained on your own pages answers those questions without a branch, so rebuild only the flows that do something a flow has to do, such as booking, order lookup, or routing to a specific team.
Run the side-by-side week for $0
The EzyConn free plan gives you 2 seats, 1 chatbot and 500 messages a month, free forever. Enough to test a real slice of your traffic before you cancel anything.
Start freeCompetitor pricing here is limited to figures published on the vendor's own site, last checked July 2026. Where a number is not published we describe the pricing model rather than guess. Check current pricing before deciding either way.