Tally Digital venture · Multi-site estate
A network of service-brand sites built to answer the questions people actually type, rank for them, and route the resulting enquiry to the firm that can do the work. Built to prove a repeatable pattern: how fast can one person put a credible, fast, genuinely useful site into search — and does it produce real enquiries?
Search for a specialist service in a rural county and the results are bleak. Directory aggregators that sell your details to six firms at once, national brokers with no regional knowledge, and a long tail of genuinely good operators whose websites are one page from 2011 with a fax number on it.
The customer wants a straight answer — what does this cost, is it urgent, who covers my postcode. The local firm wants the phone to ring. Neither is well served.
Seven sites, each covering one trade across Cornwall:
Each is a hub page, a set of service pages, thirteen town pages and a blog running to roughly 50 URLs. All of it is generated from a single Python template — content lives in plain data structures, the generator renders static HTML, and there is no CMS, no database and no WordPress to patch.
The first site took around ten hours including a full fact-check. The second, using the clone-and-rebrand pattern, took about six. That ratio is the whole point of the exercise.
The uncomfortable part. A fact-audit of the first site found eight factual errors in copy that read perfectly plausibly — including the VAT treatment of a service and the exact Environment Agency permit fee. Every one would have been believed by a reader.
That audit became a standing step. Regulatory detail, trade-body names and statutory fees are checked against primary sources before a site goes live, because a local-services site that gets the permit fee wrong is worse than no site at all.
Each site also carries an explicit footer disclosure that it is a lead generation service connecting customers with vetted local professionals — never implying we hold the tankers or turn the spanners.
A quote form posts to an n8n webhook, which validates the submission, filters bots behind Cloudflare Turnstile, emails the enquiry out through Brevo and writes it into a tracking dashboard alongside search and analytics data for that site.
flowchart LR
classDef extern fill:#fdfcfb,stroke:#c4873b,stroke-width:1.5px,color:#0a0f0f
classDef hub fill:#0a0f0f,stroke:#c4873b,stroke-width:1.5px,color:#fdfcfb
classDef out fill:#f3f2f1,stroke:#9a9a96,color:#0a0f0f
G["Python generator
+ content data"]:::hub
S["7 static sites
~50 URLs each"]:::extern
F["Quote form
+ Turnstile"]:::extern
N["n8n
validate & route"]:::hub
E["Email to the
local firm"]:::out
D["Lead dashboard
+ search data"]:::out
G --> S
S --> F
F --> N
N --> E
N --> D
One generator, seven sites, one enquiry pipeline behind all of them.
Since the first enquiry arrived in late May 2026, the network has captured 25 real enquiries — eleven for septic tanks, five removals, three pest control, three EV chargers, two holiday-let management and one locksmith. Real names, real postcodes, real urgency levels, spread across west and east Cornwall.
The sites are fast because there is nothing to them: a 16KB gzipped hub page and a 116ms time to first byte, from static files on a single small server.
And the honest bit. Turning captured enquiries into a paying local partner has been harder than building the sites. Approaches so far have run into a sector at full capacity in peak season — when a firm cannot take the work it already has, more enquiries read as a burden rather than an opportunity. That is a distribution problem, not a build problem, but it is the real state of play and worth saying plainly.
Want a site that answers the question people actually typed? Fast, static, fact-checked, with the enquiry landing in your inbox. See how we build websites →