Context
Tanzania Wildmakers is a small, deliberately small, safari operator running tailored expeditions through the Serengeti, Ngorongoro Crater, and the Kilimanjaro forest belt. Their guests do not arrive through performance marketing. They arrive through referrals, through magazine features, through a long conversation with a concierge in Geneva or Singapore who has been on one of their trips and now sends two clients a year.
When we were brought in, the brand had been quietly winning for a decade but had no website to match. The existing site was the kind of thing that happens when a wonderful operator hires the cheapest available developer in a hurry — busy, generic, with stock photography of lions photographed in South Africa. None of it matched the product. None of it earned a referral from a guest who had just paid twenty-six thousand dollars to spend nine days in the bush.
The brief was unusual in one specific way. The owner did not want more bookings. He wanted better bookings. Fewer enquiries, qualified more sharply, from guests willing to spend on the upper end of the catalogue. The website was being asked to do filtration, not amplification.
We do not need a louder website. We need a website that quietly says no to ninety percent of the people who land on it.
Decisions
01
Decision
Refuse the safari-website template.
Every safari website in East Africa looks the same. A hero with a roaring lion. A grid of itineraries. A booking widget. The genre had calcified. We wrote a one-page brand argument explaining why the existing form was the wrong form for the price point, and got buy-in to start from a magazine layout instead. The site reads top-to-bottom like an essay about Tanzania, not a catalogue of trips you can buy.
02
Decision
Spend the budget on photography, not animation.
Cinematic websites in this category lean on hover effects and parallax to perform luxury. We did the opposite: we put almost the entire production budget into commissioning new photography in the Serengeti — a four-day shoot with a stills photographer who normally shoots editorial fashion. The site has almost no animation. The photography does the work.
03
Decision
Treat the itinerary page as a long-form story, not a product page.
Each of the four signature itineraries is published as a 1,500-word piece written in the operator's own voice, describing each day with the kind of detail that only someone who has actually led the trip can write. Booking is a discreet link at the bottom — not a sticky CTA, not a price calculator. Guests at this price point want to be persuaded, then trusted to find the door.
04
Decision
Strip the navigation to four words.
Trips · Camps · Journal · Enquire. That is the entire navigation. We deleted the About page, the Sustainability page, the Team page, the FAQ. Each of those became a paragraph somewhere on the homepage, written with care. The site has six pages instead of twenty-three.
Outcome
The redesigned site went live in September 2025. Three months in, the operator's own metrics tell the story better than ours do.
2.3×
Average enquiry budget
−61%
Total enquiry volume
38%
Enquiry-to-booking rate
The total number of enquiries went down, which was the point. The enquiries that came in arrived from guests who had already been pre-qualified by the writing — they knew what they were enquiring for, and they had budgeted for it. The conversion rate to booking nearly tripled. The team's time stopped being absorbed by tyre-kicker enquiries and went back to the work they were actually good at: planning trips.
- Next.js 14
- Tailwind CSS
- Sanity (headless CMS)
- Cloudflare R2 (photography CDN)
- Vercel (Fluid Compute)
What I’d do differently
-
I would have pushed harder on the photography brief. We commissioned beautiful images, but I let the operator narrow the shot list during the shoot. A handful of images came back too literal — a giraffe is a giraffe — when what the site needed was atmosphere. Next time the shot list gets locked before the photographer flies in.
-
The Journal section is underpopulated. We launched with three articles. The plan was four a year. At time of writing there are still only four. A long-form magazine site needs a content cadence the operator can actually sustain — that's something I should have negotiated as a deliverable, not a hope.
-
The enquiry form is too short. We deliberately shortened it to feel quiet and confident, but the operator now does an extra round of email back-and-forth to surface budget and dates. A four-field form would have been the right answer. Two was too few.