The pain and the solution

The pain

  • Orders came through a flower marketplace. It kept about 30% of every bouquet, and the buyer remained its customer.
  • To post a bouquet, it was photographed, moved to a computer and typed into someone else's dashboard. A bouquet lives three days; the listing took half an hour.
  • Cost price was guessed, and trimming and wilting were never counted.
  • Florist software such as Posiflora charges a monthly fee per user and keeps the data in its own cloud.

How we solved it

  • Its own website and bot. An order through them pays nothing to the marketplace, and those 30% stay with the studio.
  • A bouquet is posted from the phone in a minute; florists post their own and never see purchase prices.
  • The cost of every bouquet comes from its contents and purchase prices, and the dashboard suggests a selling price.
  • Catalogue, customers, cost price, money and staff run on the studio's own server, with no subscription fee.

Recognise your task? We will solve it for you too.

Discuss a similar project

What we did

Storefront and cart

A catalogue filtered by occasion, flower and recipient. A bouquet removed from the cart can be put back right there; Moscow delivery is free from 10,000 roubles.

A bouquet from the phone

Photo from the camera, name, contents and price - about a minute later the bouquet is on the site and in Telegram. Florists post their own bouquets as drafts.

Bouquet cost price

A purchase list with a percentage for trimming and wilting. When roses go up, the cost of every bouquet with roses is recalculated by itself.

Customers and money

A customer base with mailing consent, expenses by category, wages, marketplace commissions and a check of their payouts.

Bot and workshops

The bot picks a bouquet from a phrase like "peonies under 20 thousand", books people into a workshop and opens the storefront inside Telegram.

The project in detail

Doch Sadovnika (“The Gardener’s Daughter”) is an author’s flower studio on Kutuzovsky Prospekt in Moscow. Bouquets are made to order, and workshops take place upstairs.

We built the studio its own software: a storefront with a cart, an owner’s dashboard with cost prices and money, a Telegram bot and a workshop page. Feature for feature it covers what cloud florist software such as Posiflora does, but it belongs to the studio: it runs on the studio’s server, the data goes nowhere, and there is no monthly fee per user.

What a marketplace costs

Before us, sales went through a flower marketplace. It takes about 30% of each order and keeps the buyer: no name, no phone number, no way to reach the person again.

For example, if 400,000 roubles a month go through the marketplace, about 120,000 roubles go to commission. Every order that comes to the studio’s own site keeps that money. That is why together with the site we offer search and content work: articles, pages for local search queries, Telegram posts. We know how to do it, and every customer who arrives this way saves the studio at least 30% of the order. The marketplace stays as one sales channel among several.

The owner’s day

In the morning a delivery arrives and roses have gone up from 190 to 210 roubles a stem. The owner changes one number in the purchase list, and the cost of every bouquet with roses is recalculated.

The first bouquet is ready. “Post a bouquet”, the phone camera, the name “Peonies with eucalyptus”, contents, price 8,900, the “one of a kind” tick, “Publish”. About a minute later it is on the site and in Telegram.

A florist posts their own bouquet. They are not allowed to set prices, so it goes to drafts marked with who made it. The owner types the price right in the row, and the bouquet goes live.

The bouquet sells: “Sold out” in the row, and it leaves the storefront but stays in the dashboard. Made another one like it: “Back on display”. Nothing is deleted for good.

In the evening the summary shows revenue and orders, what went on flowers and labour, what the marketplaces took and what is left. A separate line shows what marketplaces still owe.

Cost price and selling price

The purchase list holds more than flowers: greenery, wrapping, decor, the florist’s time. Items are counted in pieces, stems, metres or hours. Each item has a percentage for trimming and wilting, usually about 8%. Without it the cost price is understated by roughly a tenth.

The dashboard shows a bouquet’s cost, its price and the margin in roubles and percent, and suggests a typical price for an author’s bouquet at that cost. If the contents are not set, the cost stays empty rather than zero, because zero would show a 100% profit that does not exist. The cost is frozen at the moment of sale, so past profit does not change when purchase prices rise.

The owner’s dashboard

Eight sections: summary, bouquets, flowers and contents, customers, requests, money, access and a how-to. Summary, money and access are visible only to the owner; a florist does not even have them in the menu.

“Money” holds expenses in nine categories, wages (monthly, per shift, a share of each bouquet or hourly), marketplace payouts and their commissions. If a marketplace pays a rouble less than it owes, the row is flagged as a mismatch. Every action has an undo next to it.

Website, bot and workshops

The storefront has three independent filters: recipient, occasion and flower. The number on each filter matches the results, sold bouquets are not counted, and a bouquet without a price never reaches the storefront. The header plays a video from the studio, with sound only when the visitor turns it on.

The bot understands ordinary phrases like “for mum’s birthday” or “peonies under 20 thousand” and offers real bouquets with catalogue prices. If nothing fits, it says so and asks for a budget. The bot also books people into the “Living Painting” workshop: five short questions, and the request reaches the owner in Telegram at once. Staff get into the dashboard through a one-time link from the bot, with no password.

Who else this fits

Flower shops and studios, confectioners, bakeries, jewellery makers: anywhere the goods are one-off, the range changes daily, the owner stands behind the counter, and sales go through a marketplace with a commission or through subscription software.

Discuss a similar project

Tell us about your task, and we will suggest how to build something like this for you.

  1. 01
    Analysis

    We study processes, data and the points where value is lost, and fix the result we are working towards.

  2. 02
    Architecture and prototype

    We design the system and build the first version on your real data. Stages and cost are agreed before work begins.

  3. 03
    Pilot

    We launch in one area. Every stage is checked in a test version before it goes into operation.

  4. 04
    Scale and support

    We connect the remaining areas, take responsibility for how the system runs and keep developing it.

How should we contact you

Could not send the request. Please try again a little later.