← All work
HospitalityMar – Dec 2026Live · London

Demi’s Kitchen

Running a restaurant
on numbers, not instinct.

2.7×Weekly revenue, Mar – Dec 2026
Demi’s admin · door summary
Patient dashboard
Role
Product Lead
I owned
Strategy, requirements, build + launch
What we built
ClockIn for staff, plus an admin for bookings, doors + buffets
Timeline
Mar 2026 – present
Status
Live in daily use
01
Chapter 01

The problem

Demi’s is an owner-run Nigerian restaurant in London. When I came in as Product Lead in March 2026, the food was great but the numbers weren’t. Weekly revenue sat under £3,000, too much food was going in the bin, and nobody could really see how many hours staff were working, or what that was costing.

This one is a little different from my other case studies, because I didn’t spend most of it in Figma. I set the strategy, wrote the requirements and led the build of two internal tools, then paired them with new offers and marketing to bring people through the door. Design stayed lean so the tools could ship fast.

Growth wouldn’t stick while the kitchen was leaking money.

02
Chapter 02

The strategy

01Fix margin before chasing volume
02Tools staff can use mid-shift
03New reasons to visit, not discounts
04Every channel gets a number
05The owner sees it all at a glance

2 internal tools · 2 new offers · social, ads + influencers · Mar – Dec 2026

03
Chapter 03

Who it had to work for

Any change had to work for three groups of people, and they wanted very different things.

Runs the business

The owner

Needs to see the business without standing in the kitchen all day.

Needs
  • Revenue and costs at a glance
  • Labour costs she can control
  • Less waste, same quality
Risks I planned for
  • A busy weekend wiping out stock
  • Hours creeping up unseen
  • Decisions made on gut feel
On shift · shared devices

Kitchen + floor staff

Clock in, check stock, get on with service.

Needs
  • Clock in and out in seconds
  • Know what’s in stock
  • Nothing that slows service
Risks I planned for
  • Shift swaps
  • Forgetting to clock out
  • Busy service, no time for admin
New + returning

Customers

Need a reason to come in, and a reason to come back.

Needs
  • An offer worth the trip
  • Hearing about it where they scroll
  • Something new on a weekend
Risks I planned for
  • Not knowing what’s on
  • A one-off visit, no return
  • Ad spend with nothing to show
And behind it all

Numbers the owner could act on.

Every offer, ad and influencer campaign was tracked against visits and weekly revenue. If it worked, we kept it. If it didn’t, it went.

04
Chapter 04

How I implemented it

I led requirements and delivery here, not pixel-pushing. Four moves did most of the heavy lifting.

Screenshot · Stock tracking before new menus

Stock tracking before new menus

Before chasing more customers, I wanted to stop losing money on the ones we already had. I scoped and led the build of inventory tools so ordering followed what actually sold, and the door summary now shows exactly what sold each day. Food waste dropped 68.7%, which protected our margin as volume grew.

Screenshot · A clock-in platform for labour

A clock-in platform for labour

Hours were the biggest cost nobody could see. Staff clock in and out on ClockIn, the rota sits next to a daily labour budget, and timesheets work out pay for the owner. No more guessing what a shift cost.

Screenshot · Buffet and brunch, not discounts

Buffet and brunch, not discounts

Discounts would have trained people to wait for a deal. Instead we launched a buffet and a brunch: new reasons for first-timers to come in, and for regulars to come back. A door kiosk checks buffet guests in and prices them by arrival time.

Screenshot · Every channel gets a number

Every channel gets a number

Each launch got social content, paid ads and influencer campaigns, and every one of them was tracked. Booking forms even ask how people heard about us. If a channel didn’t fill tables, the budget moved.

05
Chapter 05

What went live

Staff · ClockIn
ClockIn · staff
Owner · Demi’s admin
Demi’s admin

Two sides of the same restaurant. Staff clock in, check the rota and see who’s on shift. The owner sees bookings, covers, takings and what sold at each door, across Cricklewood and Streatham Hill. And it’s still running every day: in the last 30 days alone, 79 bookings and 289 covers came through it.

Live board
Live board
Rota + labour budget
Rota + labour budget
Staff roles
Staff roles
Owner dashboard
Owner dashboard
Booking calendar
Booking calendar
Bookings
Bookings
Set menu groups
Set menu groups
Buffet bookings
Buffet bookings
Buffet door kiosk
Buffet door kiosk
Sunday Feast
Sunday Feast

Scroll sideways for 10 key screens →

06
Chapter 06

The results

£8K+Weekly revenue, from under £3k
68.7%Less food waste
2.7×Revenue growth since March 2026

From the restaurant’s own records, March to December 2026.

Why it worked

Cost control first, then growth.

Cutting waste first meant the extra covers from the buffet and brunch turned into profit, not just busier shifts. Tracking every campaign meant the marketing money went where it worked. And the tools are still in daily use.

Photo · rota vs daily budget
rota vs daily budget
Photo · buffet door kiosk
buffet door kiosk
07
Chapter 07

What I’d do differently

01

Track repeat visits from day one.

We knew which offers filled tables, but not which ones built loyalty. Measuring repeat visits earlier would have told us.

02

Design the hand-off.

Documenting the tools and training someone else on them sooner would have made them less dependent on me.