D&D LANDSCAPING- From word of mouth to a digital home.
FIRST AI-ENABLED WEB BUILD
AUCKLAND, NZ
2026
Project Summary
The trust already existed. The website didn’t.
D&D had the things a local service business cannot fake: practical experience, honest communication and a reputation built through real work. What they lacked was one owned place online that carried the same confidence. I shaped a responsive, search-ready website that clarified services, showed how the team works and made the first enquiry feel less like a form and more like the start of a conversation.
AI angle: This was my first AI-enabled website build. Figma Make, MCP and ChatGPT accelerated structure, copy and interface exploration, while I stayed responsible for the product decisions, tone and final experience.
Problem
Instagram could show the work. It couldn’t explain the business.
Discovery relied on referrals, Instagram and a basic Microsoft 365 form. New customers could see fragments of the work, but had to infer the service range, process, team and what would happen after they made contact. For landscaping, that uncertainty is a trust problem.
So the real question became: how might we make a small team’s trust visible before the first phone call?
My Role
I owned the story as much as the screens.
I worked end to end across the proposition, customer journey and delivery. The job was not simply to “make a website”; it was to turn D&D’s offline strengths into an experience people could understand before meeting the team.
• Experience positioning, information architecture and page hierarchy
• Website copy, service framing and local SEO structure
• Figma Make-led build workflow, MCP prompting and rapid iteration
• Responsive review, enquiry flow and handoff-ready page structure
Outcomes
One website. Three practical gains.
TRUST HAS A HOMETeam, values, process and service coverage are visible before a customer commits to a conversation. | ENQUIRIES HAVE CONTEXTThe form captures suburb, project needs, preferences and optional photos, giving D&D a better starting point for follow-up. | SEARCH HAS SOMEWHERE TO LANDService-specific and Auckland-specific content creates an owned foundation for organic discovery and future optimisation. |
40
Description
354
description
54
Intent
Credible, not corporate.
D&D’s advantage was never scale. It was direct communication, fair advice and two people who care about the work. The design needed to amplify that, not sand it into generic “premium landscaping” language.
“You don’t need a landscaping brief. Start with the problem.”
Challenges
Small team. Big trust gap. New AI workflow.
1 TRUST IS BEHAVIOURALHonesty and follow-through are hard to prove in a hero banner. The experience had to show them through language, expectations and process. | 2 CUSTOMERS SPEAK IN PROBLEMSHomeowners often know what feels wrong, not whether the answer is paving, planting, fencing or maintenance. | 3 AI SPEED CAN FLATTEN PERSONALITYFaster generation was useful only if the output still sounded like D&D. Generic copy and generic patterns had to be edited out. |
Approach
Prompt fast. Edit harder.
LISTEN + FRAMEPull out the real signals: why customers call, what D&D believes, and where uncertainty appears before a quote. | STRUCTURE + MAKEOrganise the journey around natural questions, then use Figma Make, MCP and ChatGPT to explore pages and interactions quickly. |
HUMANISECut generic agency language. Use actual service situations, local Auckland context and plain explanations. | CONNECTLink homepage, services, process, about and contact into one story: discover → understand → trust → enquire. |
08 KEY WORK / SOLUTIONS
Four decisions did most of the heavy lifting.
01 Reassurance before capabilityThe homepage leads with a people-first promise, then quickly explains process, services and next steps. It answers “can you help me?” and “what will dealing with you feel like?” before asking for a lead. |
02 Services framed around situationsEach service explains what it solves and who it suits. Customers can recognise their own problem without first learning landscaping terminology. |
03 The invisible process made visibleThe “How We Work” experience turns an opaque contractor journey into clear expectations, from enquiry and site visit through quoting, delivery and follow-up. Uncertainty is explained rather than hidden. |
04 Lower-friction enquiryThe contact experience says customers do not need a finished brief, then captures only what helps the next conversation: suburb, need, timeframe, preferences and optional photos. |
09 TRANSFORMATION
Not more pages. A more coherent path to trust.
DIMENSION | BEFORE | NOW |
Discovery | Instagram + referrals | Owned website with a clear customer journey |
Services | Mostly inferred | Scenario-led service structure |
Process | Largely invisible | Expectations explained before contact |
Enquiry | Basic form | Guided context + optional project details |
Trust | Built mostly live | People, values and behaviours visible up front |
10 LEARNINGS
First AI build. Three things I’ll carry forward.
AI COMPRESSES PRODUCTION, NOT JUDGEMENTFigma Make made options cheap. The designer’s job shifted toward framing, editing, taste and knowing what not to ship. | CONTEXT IS THE REAL PROMPTThe strongest outputs came from specific business context, customer language and local situations, not clever prompt syntax. | SMALL-BUSINESS UX IS TRUST DESIGNPlain language, visible next steps and honest expectations did more work here than novelty ever could. |
A small local business, now with a digital presence that feels as human as the people behind it.
