Designing and building an in-home tasting service
- Company
- Decanto
- Role
- Founder, designer and developer
- Team
- Just me
- Timeline
- May to October 2026
- Outcome
- Live in Stockholm, first tastings delivered
TL;DR
Decanto brings a sommelier to your home in Stockholm for a wine tasting. I did the end-to-end product design and built it myself: brand identity, site, booking and back office.
Manual before marketplace
The obvious move was a marketplace, where sommeliers log in and guests book themselves. I didn't build that. I started with service design, mapping how guests, sommeliers and I talk to each other, and matching by hand won: a faster MVP, and nicer for guests. The emails run themselves, and I do the matching.
Fewer, clearer tastings
Too many themes make people hesitate, and each one needs a sommelier who can deliver it. So the content strategy is eight tastings, with a standard one nobody has to choose. Every page comes from one template, so a new tasting is a few lines of text.
One page to book
Booking is a one-page checkout flow with a fixed price: 599 kr per person, paid when you book. A request form would have let me price each group by hand, but a fixed price lets people book without asking first, one step less before conversion. Below the button sit only the payment methods we accept.
A separate path for companies
Companies don't buy like friends planning a birthday. Group size and theme vary, so they get their own B2B page built for lead generation: the whole quote request is the first thing they see. No checkout. I started on a business login too, but it was overkill before we'd proven the value, so I cut it.
A shopping list instead of delivery
In Sweden, guests buy their own wine at Systembolaget, so I leaned into it. After every booking, the sommelier fills in a shopping list on one page, with the guest count, budget and wishes right above it. The list then goes to the guest by email, so the end-to-end experience carries on after payment.
The back office
I run everything from one back office, designed as internal tools: bookings arrive paid, I assign a sommelier, and payouts and gift cards run through Stripe. It's a static site with a Next.js back office on Supabase, Stripe and Resend.
In short
- Key decision
- Matching by hand instead of a marketplace. It was faster to launch and nicer for guests.
- Changed course
- I started on a business login, then cut it. It was overkill before we'd proven the value.
- Outcome
- Live in Stockholm, with the first tastings delivered. Designed and built solo. Partnership talks are in progress.
- Takeaway
- Launch the manual version first. Automate the parts that repeat, like the emails.


