First count decision paths, not pages
I do not pick a CMS from a pretty template. Problems show up later: you need a second language, change a canonical, set a 301, connect CRM, and figure out why the form did not submit.
The mistake I would remove first: Choosing structure from a template instead of from the number of intents.
What to prepare and what result to expect
- Outcome: Structure does not block growth and does not force a visitor to hunt one service inside a universal canvas.
- Count services/products, separate audiences, regions, filters, content, and the number of future pages.
- Keep source data and access rights separate from the result so you can verify what site structure choice did.
- Draw the path for three typical clients and check whether each can get a separate URL and CTA.
Match three structures to how the business really works
- Choose a landing page
A landing page fits one product, one audience, a short decision path, and few standalone queries. Build sections for problems, proof, terms, and CTA.
Проверьте: The whole scenario can be completed without a second page.
Если не сработало: If different services and questions appear, move them to separate URLs.
- Choose a multi-page structure
For several services create /services/ and a separate URL for each service. Add contacts, case studies, FAQ, and internal links between related pages.
Проверьте: Each page has its own intent and an owning CTA.
Если не сработало: Do not split one service by city without local content.
- Choose a store
A store is needed when you have a catalog, inventory, prices, variants, a cart, and repeat purchases. Describe filters, categories, and indexing rules up front.
Проверьте: A visitor can find a product and understand availability/delivery.
Если не сработало: For 3–5 products use a product page or a small catalog, not a heavy platform.
- Calculate total cost of ownership
Add support, content, integrations, SEO, forms, payments, and data migration. A cheap start with expensive manual upkeep is not savings.
Проверьте: There is an owner of changes after launch.
Если не сработало: Remove a feature nobody will be able to maintain.
A test page before you buy
On each platform create `/test-service/` and check: title, description, slug, canonical, noindex, JSON-LD, sitemap, 301 redirect, form, and GA4 without double counting. Verify the published URL, not the preview.
In WordPress open `Settings → Reading` and `Settings → Permalinks`. In Webflow — `Pages → Page settings → SEO settings`. In Wix — `SEO & GEO → Tools and settings → SEO Settings`. In Tilda — page SEO settings and the published `robots.txt` and `sitemap.xml`.
curl -I https://example.com/test-service/
curl -s https://example.com/test-service/ | grep -Ei 'title|description|canonical|robots'I check the cost of exit
Before launch request content, image, and URL export, a backup, a redirect table, and domain transfer. If every change needs a contractor, that is not automatically a bad platform—just add internal time and support cost into TCO.
- The domain and recovery accounts must belong to the company.
- A new site must not change every URL without a 301 map.
- Analytics and Search Console access are verified before launch.
I check the emergency path, not only the editor
Ask to create a backup, delete a test page, set a 301, restore the page from backup, and export content. If support cannot say where the sitemap lives or how to transfer the domain, record that as an ownership risk.
Score the platform on a real service page and form, not a demo template. One hour of that check is cheaper than a migration after traffic drops.
When you need a landing page, a site, or a store
| Criterion | Question | Good sign |
|---|---|---|
| Input | What exactly enters site structure choice? | Count services/products, separate audiences, regions, filters, content, and the number of future pages. |
| Action | What is the system allowed to do on its own? | Only prelisted actions, without access to the entire account |
| Verification | How do you know the result is acceptable? | Draw the path for three typical clients and check whether each can get a separate URL and CTA. |
| Failure | Where do unclear cases go? | Start with a smaller structure and leave clean URLs for future services instead of creating sections “just in case.” |
What should change after setup
Structure does not block growth and does not force a visitor to hunt one service inside a universal canvas.
How structure starts blocking search and sales
Choosing structure from a template instead of from the number of intents.
Trying to fit five services into one landing page.
Buying a store with no owner for catalog and inventory.
Not planning for content and URL portability.
When site choice cannot be made on budget alone
Bring in a specialist if the choice affects a catalog, multiple markets, URL migration, or payment and CRM integrations.
What to ask yourself before buying a template
Who is this site structure choice approach for?
A landing page, a multi-page site, and a store differ not by screen count, but by number of intents, products, rules, and purchase scenarios. Choose structure for how the business works.
Where do you start if everything is still manual?
Count services/products, separate audiences, regions, filters, content, and the number of future pages.
How do you check that the setup will not hurt you?
Draw the path for three typical clients and check whether each can get a separate URL and CTA.
What if the result is unclear?
Start with a smaller structure and leave clean URLs for future services instead of creating sections “just in case.”
