Website Migration Service: How to Choose the Right One
A migration rarely fails because a team forgot to move files. It fails because someone treated a business transition like a deployment task, then discovered too late that redirects, analytics, consent, indexing, data flows, or rollback ownership were never locked down.
For agencies carrying client risk, and for enterprise teams inheriting a portfolio under deadline, a website migration service has to do more than make a new site live. It has to preserve revenue paths, search visibility, and operational control while the old environment is still serving users and bots. That is the standard WebinOne takes, and it is the right standard to demand from any provider.
Table of Contents
- Why a Website Migration Service Is a Continuity Decision
- Lift-and-Shift or Rebuild, How to Decide
- Scoping, Inventory and a Migration Plan You Can Audit
- Redirects, SEO and the Expected Recovery Timeline
- Rollback, Uptime and What the SLA Must Actually Promise
- What a Migration Service Should Cost and What Changes the Number
- Questions to Ask a Migration Provider and What Comes Next
Why a Website Migration Service Is a Continuity Decision
A website migration service is a continuity decision because it moves more than code. It shifts hosting, infrastructure, URLs, content, redirects, analytics, and search-engine indexing, so a launch can look successful while revenue, attribution, or search visibility starts to slip.
Google's own guidance says even a medium-sized site can take “a few weeks or more” for new URLs to replace old ones in search results, and larger sites can take longer because recrawling and reindexing happens progressively Google site move guidance. That alone should kill the idea that migration is a same-day file transfer. Cutover is the visible moment, not the finish line.
What the move really includes
A serious migration service has to handle the mechanics behind the page. When domain or subdomain changes are involved, Google guidance calls for permanent server-side 301 redirects, including redirects for canonical pages, and its Change of Address tool can transfer signals from the old site to the new one.
That matters because the obvious win, a new site online, is not the business outcome. The business outcome is continuity. If redirects, crawl access, canonical tags, internal links, or sitemaps are incomplete, a deployment can still damage indexed URLs, organic visibility, or traffic.
The same standard shows up in switching platforms with ECORN, where platform change is treated as a move that affects content, structure, and search visibility. That is the right lens for any stack.
Practical rule: if a provider talks only about launch day and not about the weeks after cutover, that provider is selling a deployment, not a continuity plan.
A working definition of success should include preserved rankings, intact content, valid redirects, stable analytics, and no unexplained break in customer journeys. If a migration service cannot describe those outcomes clearly, it is not ready for enterprise or portfolio work.
Lift-and-Shift or Rebuild, How to Decide
WebinOne treats this choice as a business decision, not a design preference. The right model depends on whether the existing structure is worth preserving or whether the structure itself is the problem.
The two models are not the same
A lift-and-shift keeps the legacy shape as intact as possible. It preserves structure, templates, and redirects close to one-to-one, which usually means less risk, less mapping complexity, and a faster path to cutover.
A rebuild uses the migration to clean templates, consolidate modules, and modernize the data model. That gives more room to improve the estate, but it also multiplies the mapping, rebuild, and testing burden.
| Axis | Lift-and-shift | Rebuild |
|---|---|---|
| URL preservation | Highest, because the legacy path is mirrored closely | Lower unless the new structure is carefully mapped |
| Data integrity | Easier to protect for ecommerce, members, and CRM data | Strong, but only if the rebuild is disciplined |
| Timeline to cutover | Shorter and easier to control | Longer, with more moving parts |
| Cost shape | More predictable up front | Higher discovery and rebuild effort |
| Acceptable rework | Minimal, focused on continuity | Greater, because the structure is being changed |
A rigidly good answer does not exist. The right answer is the one that fits the deadline, the risk profile, and the amount of structural change the client can absorb.
Where the line should be drawn
If the move is tied to an end-of-life deadline, or the estate is large and messy, lift-and-shift is usually the disciplined choice. It protects continuity and reduces the number of unknowns before cutover.
Rebuild belongs where the current structure is the actual obstacle and the client is willing to pay for the extra time, QA, and stakeholder review. That is a valid move when the business wants better governance, cleaner templates, or a more sustainable operating model.
For a deeper look at phased planning and content handoff, Streamkap's migration playbook is useful context even outside web work, because the same discipline applies, inventory first, movement second, cleanup last.
If the client cannot explain why the rebuild risk is worth it, the answer is probably lift-and-shift.
That position is blunt for a reason. A migration service should defend the model that protects the business, not the model that creates the most billable rework.
Scoping, Inventory and a Migration Plan You Can Audit
WebinOne and TeamOne approach scoping the same way every time, because repeatable discovery is what makes portfolio work manageable. The first deliverable is not a launch date, it is a written map of what exists, what moves, and what gets tested.
Start with inventories, not assumptions
The content inventory should cover pages, templates, media, structured data, and anything else the site publishes publicly. The data inventory has to include orders, members, CRM records, consent state, and other records that affect business operations after cutover.
The integration inventory matters just as much. APIs, webhooks, payment gateways, identity systems, and form flows need to be listed before anyone starts rebuilding. If those dependencies are hidden until the end, the launch becomes a repair job.
The URL map is the spine of the whole effort. Every legacy URL needs a target, and every redirect decision should be documented so the launch can be audited after the fact.
Audit standard: if the provider cannot show a traceable relationship between old URLs, target URLs, and content owners, the migration is under-scoped.
Historical platform shutdowns prove why this structure matters. When Adobe Business Catalyst was announced for end of life and later shut down on a fixed deadline, portfolios had to move in waves, not one by one, and repeatable discovery became the only sane way to handle the load Business Catalyst end-of-life timeline.
What a credible first pass should produce
A provider worth trusting should produce a small set of artifacts early:
- A content map that shows what is in scope and what is excluded.
- A redirect map that ties old URLs to new targets.
- A dependency list covering integrations and data flows.
- A test plan that defines staging checks and launch gates.
- A rollback plan that says who can stop the cutover.
Staging has to be more than a visual review. It needs full environment parity, a redirect test suite, a content-delta review between staging approval and launch, accessibility checks against WCAG 2.2, and a defined rollback point WCAG 2.2 guidelines.
If those artefacts do not exist in the first two weeks, the service is improvising.
Redirects, SEO and the Expected Recovery Timeline
WebinOne treats SEO preservation as a business KPI, not a checkbox. Redirects matter, but they are only one part of the move, and they do not protect a site by themselves.
Why redirects alone are not enough
301 redirects cover the path, not the full experience. Canonical tags, internal links, sitemaps, structured data, conversion paths, attribution, and consent behavior also move with the site, and all of them can break in ways that are invisible on launch day.
A practitioner analysis of 892 domain migrations reported an average 523-day period for organic traffic to return to its prior level, and 17% had not recovered after 1,000 days migration benchmark analysis. Those are observational benchmarks, not universal causal rates, but they are useful risk indicators. They tell teams to stop assuming recovery is quick or automatic.
If the site includes Shopify paths, learn how to manage Shopify URL redirects before launch so old product and collection links do not fall through the cracks.
For teams building a checklist, WebinOne's site migration SEO checklist is the reference to keep open during planning and QA.
Where effort actually pays off
The first control is an automated URL test suite. Every high-value legacy URL should be checked for exactly one hop, a relevant destination, a permanent redirect status, and no loops.
Then crawl the new site and look for orphan pages, broken internal links, duplicate canonicals, missing titles, malformed structured data, blocked assets, and sitemap inconsistencies. Segment the results by URL type and template so one healthy section does not hide failures in product, article, regional, or account pages.
Before launch, take a baseline of traffic, rankings, conversions, revenue, crawl coverage, page speed, and accessibility. After launch, keep a post-launch control period with daily anomaly detection and rollback criteria. That is how a migration service earns the word “service.”
The weekly energy should go where the risk lives. In the week before cutover, finish the URL map, lock the redirect rules, and freeze last-minute changes. In the four weeks after, watch the deltas, not the vanity impressions.
Rollback, Uptime and What the SLA Must Actually Promise
WebinOne's SLA language is the kind of thing buyers should demand from any migration provider. A promise of uptime only matters if the architecture can survive failure, and if the rollback path is real.
Zero downtime means tested redundancy
AWS defines availability as the percentage of time a workload is available for use, and it describes a single point of failure as a component whose failure disrupts the application because it lacks redundancy AWS resilience guidance. That's the heart of the matter. A migration plan that copies files and changes DNS without tested redundancy is not a low-downtime plan, it is a hope.
If a workload depends on a database and that database fails, requests fail unless a standby and failover path exist. The same logic applies to application, storage, and routing layers. The rollback decision also needs a named owner, not a group chat consensus.
Read the SLA like a buyer, not a salesperson
A strong SLA should separate the availability commitment from maintenance windows. It should also explain how credits are calculated, what evidence is required to claim them, and which exclusions remove coverage.
WebinOne publishes a 99.95% availability commitment and reports 99.99% uptime over the last 12 months, hosted on AWS across six regions. That is the sort of written commitment a migration buyer should expect a provider to stand behind.
The operational details matter just as much as the number:
- Pre-launch freeze so the target does not drift during QA.
- Change-window approval so cutover has explicit authorization.
- Incident escalation so the right people act immediately.
- Rollback ownership so the cutover can be stopped without debate.
- Breach-response ownership so data incidents are handled within the reporting clock.
The UK GDPR guidance from the Information Commissioner's Office says a notifiable personal-data breach must be reported without undue delay and no later than 72 hours after the organization becomes aware of it, where feasible ICO breach guidance. That turns migration credentials, exports, and imports into operational risk, not admin work.
For a broader view of support expectations and escalation paths, managed services support should be part of the buyer's reading list before any contract is signed.
If the SLA does not show how the provider behaves when something breaks, the SLA is marketing.
What a Migration Service Should Cost and What Changes the Number
A migration quote should read like a continuity budget, not a single line item. WebinOne is transparent about starting price because the cost is the mix of project work, risk exposure, and ongoing run cost.
The quote has to be broken apart
A credible provider separates discovery and inventory from mapping and rebuild work. It also isolates staging and QA, cutover and rollback support, post-launch monitoring, and the platform plan that sits under the migrated site.
That split matters because the one-time project is only part of the bill. Patch work, plugin upkeep, scaling, and handoffs can cost more over time if the new stack still behaves like a maintenance burden.
What transparency looks like
WebinOne starts at $10 per site per month and charges zero transaction fees on ecommerce. That does not make every project cheap, because custom work is still scoped separately, but it does make the platform cost legible managed service pricing.
The right way to read any migration proposal is to ask three questions:
- What is being rebuilt now?
- What will still need maintenance later?
- What is the risk of getting the cutover wrong?
A managed platform with TeamOne delivery changes the cost shape because it removes the long tail of upkeep that patched-together estates tend to accumulate. Buyers often miss that when they compare only the headline project fee.
For a clearer view of how ongoing costs are framed, review the managed service pricing guide alongside the proposal itself.
If a quote looks simple, it is probably hiding the expensive part.
Questions to Ask a Migration Provider and What Comes Next
WebinOne uses the same scorecard for every serious conversation, because the questions reveal whether the provider understands continuity, not just launch mechanics.
Ask questions that expose the real process
Start with discovery. How is the inventory built, and who signs off on the URL map?
Then ask about SEO preservation. How are redirects tested, how are recovery signals measured, and what gets monitored after launch?
Move to rollback. Who owns the call, and what specific condition stops the cutover?
Read the SLA next. What is published, what is excluded, and what does a credit claim require?
Finally, ask about platform and scale. Where does it run, who supports it, what replaces fragile add-ons, and how does the process repeat across many sites under a deadline?
A good provider answers these questions without improvising.
For agencies, the next step is the WebinOne reseller program. For systems integrators, AWS partners, and enterprise teams, the right next move is a conversation with the WebinOne team about a partnership or migration through WebinOne enterprise.
If the provider can't answer these questions cleanly, keep looking. If it can, move the conversation from “can this launch?” to “can this business stay stable after launch?”