Neweb

Neweb is a bet that local businesses do not want a website builder. They want a website, and the builder is the obstacle. Everything commercially interesting about this product follows from that: if the outcome arrives in ninety seconds, the onboarding funnel stops leaking, and if the same data keeps Google and the site in agreement afterwards, the subscription has a reason to renew. Those are two different businesses stacked on one intake form, and they fail for different reasons.

CMS share held by WordPress

60.7%

43.3% of all surveyed sites. Neweb drives WordPress as an engine rather than selling it, which is why the deployed site is portable.

Source · W3Techs data, October 2025, reported by Search Engine Journal

The LCP threshold that counts as good

2.5s

Google's own bar for Largest Contentful Paint. The deployed sites are built to sit well inside it.

Source · Largest Contentful Paint (LCP), web.dev

Questions before anything is built

4

Business, location, industry, name. No design decisions are asked of the customer at all.

Model · Counted from the intake flow documented in this case study.

Surfaces kept in sync from one record

3

The site, Google Business Profile and Bing Places. One wrong phone number cannot survive in two places.

Model · Counted from the delivered scope in this case study.

The problem is the blank canvas, not the tooling

Site builders are good products that most of their signups never finish, because they hand a local business owner an empty page and a design decision. Neweb's intake asks four questions and makes none of the decisions available to the customer. That is a conversion mechanism disguised as a limitation, and it is the whole reason the ninety-second claim matters commercially: for a self-serve product with no sales team, onboarding completion is the funnel.

1. 01Four questions

   Business, location, industry, name. The shortest form that still contains enough to build from.

   ▾

2. 02Domain claimed and pointed

   Provisioned inside the flow rather than sent as homework. A registrar redirect at this step is where most self-serve site products lose the customer.

   • ↳ Preferred domain taken → alternatives offered inline
   ▾

3. 03Site built from an industry layout

   A block layout chosen for the trade, filled with copy generated about that specific business rather than lorem text.

   ▾

4. 04Sync the business profiles?decision

   Google Business Profile and Bing Places populated from the same record as the site.

   • ↳ Profile already claimed → ownership verification queued
   • ↳ Verification pending → site ships, sync retried
   ▾

5. 05Autopilot on

   Weekly SEO passes and performance monitoring start immediately, which is the part that converts a one-off build into a subscription.

Commission an agency

~4 weeks

Brief, design rounds, content requests. The highest quality, and the reason most local businesses never start.

Use a site builder

~2 weeks, if finished

Elapsed rather than active time. The work is spread across evenings, and a large share of drafts are abandoned.

Four questions

~90 seconds

One session, no design decisions, and the outcome exists before the owner's attention does anything else.

Model · Our estimates of elapsed calendar time for each route, from decision to a public URL. Illustrative comparison — no completion-rate data was shared with us.

Why the recurring half is the real product

A website is a one-off purchase that people resent paying for twice. Presence is not. The commercial distinction between Neweb and a site builder is that the thing being sold monthly is agreement between the site, Google and Bing — the details that go stale on their own and cost a business calls when they do.

Profile and data sync40%

The only line the customer can verify themselves by searching for their own business.

Ongoing SEO passes30%

Highest value, lowest visibility. The reason churn analysis for this product is hard.

Hosting and performance20%

Table stakes, and priced as such by every competitor.

The original build10%

Enormous value in month zero, close to none by month three.

Model · Our weighting of the four delivered capabilities by their contribution to renewal, based on which of them keep producing value after launch.

Performance is a local-search argument

Speed is in the delivered scope because of who is on the other end: someone searching for a trade on a phone, on mobile data, with a competitor one result away. Google's own threshold for a good Largest Contentful Paint is 2.5 seconds 1, and the deployed sites are built with LiteSpeed and edge caching to sit well under it. On a fleet of automatically generated sites this has to be structural rather than per-site tuning — nobody is going to hand-optimise the four hundredth restaurant.

1. L1

   Intake and dashboard

   Four questions, then a status surface. Deliberately thin, because the customer should not be making decisions here.

   • Next.js 14
   • TypeScript
   • Tailwind CSS
   • Framer Motion
2. L2

   Orchestration

   Sequences domain, install, theme, copy and plugins, and owns the failure handling for each one.

   • Node.js
   • Express
3. L3

   Site engine

   WordPress driven programmatically. The customer never sees an admin panel, but the output is a standard site they could take elsewhere.

   • WordPress
   • Generated block layouts
   • LLM-written copy
4. L4

   Presence integrations

   Where the recurring value lives: the same business record pushed to the places people actually search.

   • Google Business API
   • Bing Places
   • Schema markup
   • Sitemaps
5. L5

   Delivery

   Holds page performance on a fleet of generated sites without per-site work.

   • LiteSpeed
   • Edge caching
   • Cloudflare

What we would watch

References

1. 1.Largest Contentful Paint (LCP) · web.dev, Google
2. 2.Top 10 Content Management Systems (W3Techs data, October 2025) · Search Engine Journal