Short answer
A new website only makes sense if the positioning is fundamentally different. If not - one strong, well-structured site almost always performs better than several smaller ones.
Here's how we usually look at it.
When a separate site does make sense
Creating a new site is justified when at least one of these is true:
- Different positioning or audience Different ICP, different buying triggers, different language, different value proposition.
- Different business model For example: marketplace vs agency, product vs consulting, B2B vs B2C.
- Different intent universe The site answers fundamentally different questions for users (not just variations of the same one).
- Brand separation is intentional You want Google and users to treat it as a separate entity with its own authority.
In these cases, a new site is not "splitting SEO" - it's creating a new semantic and commercial space.
When a separate site usually hurts more than it helps
Multiple sites are usually a bad idea when:
- The services are closely related.
- The audience overlaps.
- The keywords and search intent overlap.
- The content would be 70-80% similar.
- One site would inevitably be weaker.
In this scenario, you end up:
- splitting authority,
- duplicating effort,
- competing with yourself in search,
- slowing down trust building.
Google strongly prefers one authoritative source per topic, not several half-strong ones.
The "one strong site" model (most common for B2B)
For most B2B platforms and services, the winning approach is:
One domain -> clear structure -> strong internal linking
Instead of multiple sites, you separate by structure, not by domain.
Common structural models (and when to use them)
1. Hub-and-spoke model (recommended in most cases)
- Hub pages Broad, authoritative pages (e.g. "Coworking Platform", "Office Marketplace", "Flexible Workspace Software").
- Spokes Supporting pages:
- locations,
- use cases,
- industries,
- features,
- guides,
- case studies.
This tells Google:
"This is the main topic - everything else supports it."
This model scales extremely well and avoids cannibalization.
2. Service-led structure
Used when people search primarily by solution type.
- Core services as main pages.
- Supporting content below:
- who it's for,
- how it works,
- examples,
- locations.
Good for agencies, platforms, and consultative products.
3. Content / authority-led structure
Used when organic growth is driven by education and research.
- Strong knowledge base / library.
- Guides, comparisons, explainers.
- Commercial pages supported by authority content.
This works well when the buying cycle is long and research-heavy.
4. Geo-led structure (careful)
Works when location truly changes the offering (pricing, availability, regulation).
Must be:
- deeply localized,
- clearly differentiated,
- internally linked to a global hub.
Otherwise it quickly turns into thin or cannibalizing pages.
Content strategies are more important than site count
What actually matters for growth is how content is organized, not how many sites exist.
Typical strategies we see working:
- Commercial hubs + educational support Services + guides + proof.
- Programmatic pages with intent control Locations / industries that clearly answer different questions.
- Case-driven authority Real examples reinforcing commercial pages.
- Knowledge base explaining "how we think" Builds trust and long-tail visibility.
All of this works better on one strong domain than across many weak ones.
Our general recommendation
If the question is:
"Should we create another site or another section?"
Our default answer is:
Start with one strong site. Split only when positioning truly diverges.
It's much easier to:
- expand structure,
- add hubs,
- segment content,
than to merge authority later after splitting it.
Bottom line
- Multiple sites != better SEO.
- Authority compounds best on one domain.
- Structure and intent separation matter more than domains.
- New sites are for new positioning - not for organizational convenience.
If you want, next step could be:
- mapping your current topics into clear hubs,
- or stress-testing whether any part of the business really deserves a separate site.
That's usually where the right answer becomes obvious.