Case study · Agriculture · FarmInvest

A second brand on the same engine.

Agricultural land investment, sold to an audience that reads before it enquires. The brand needed its own lead system, its own content, and its own sending identity — without a second team, a second stack or a second monthly bill for tooling.

464

leads under management

2

brands, one system

100%

sends brand-routed in code

Weekly

reporting, unattended

Before

  • A second brand with no system of its own, borrowing time from the first
  • Founder research sitting in notes because formatting it was a whole afternoon
  • Manual choice of sending identity on every campaign — one slip and the wrong brand sends
  • No reliable count of who was on the agricultural list

After

  • 464 leads under management, tagged, scored and queryable
  • Raw research becomes a formatted draft the founder only has to approve
  • Brand identity derived from the project, enforced in code
  • Weekly reporting arrives without anyone assembling it

What we built

Lead system on shared rails

FarmInvest runs on the same lead engine as the property side: canonical project and source tags, scoring, recency windows, and audience queries in plain English. A second brand cost a fraction of the first because the machine already existed.

Backend CMS the founder can use

Content, campaigns and email sequences are edited from an admin portal rather than a developer's inbox. Drafts are reviewable, versioned and previewed exactly as they will land.

AI content pipeline for agricultural research

Founder-written notes and agricultural research go in raw. The pipeline structures them into properly formatted, publishable articles in the brand's own voice — headings, summaries, and the framing an investor audience expects — with the founder approving before anything publishes.

Deterministic brand routing

One system, two identities. Anything agricultural sends as FarmInvest; everything else sends as the property brand. The project decides the brand, not a human dropdown, and a wrong sending address fails loudly instead of quietly.

Approval-gated bulk sending

Audience counts first, branded preview second, explicit approval third. Do-not-contact is filtered, bulk sends above a threshold need confirmation, daily volume is capped, and every send is logged.

Reporting on a schedule

Weekly performance drafts covering sends, engagement and appointments, produced automatically for review rather than assembled by hand.

Running more than one brand?

The second one should cost a fraction of the first. That only happens if the first was built as a system.

See the automations service →