A florist order chatbot that knows your cut-off, your delivery zones and when to call you

It's 1.40 on a Thursday. You've got a funeral tribute half-wired on the bench, a bride's mother due in at two, and the shop phone ringing with someone asking whether they can still get flowers to an address in the next village today. Your same-day cut-off is two o'clock. If you answer, the tribute waits. If you don't, the order goes to the first florist on Google who does pick up. A florist order chatbot is meant for exactly that moment: someone else answering, applying your rules, and writing the order down properly while your hands are full of oasis and wire.

This page is about Klugon's AI assistant for flower shops. It answers your messages on Instagram, WhatsApp, Facebook Messenger and your website, and can answer the shop phone too. It knows your cut-off times, the postcodes you deliver to and what each costs, your price bands, and what goes on a card. It takes the order down, checks it against those rules, and hands it to you to confirm and arrange payment. It doesn't design bouquets, and it doesn't take money in the chat.

Florist order chatbot basics: what customers actually ask

Flower shop enquiries are more varied than they look from outside. A florist hears all of these in a week:

• Same-day gifts. "Can you deliver something nice to my mum in Knaresborough today? About £40." The two things that matter are the time and the postcode.
• Next-day and dated orders. Birthdays, anniversaries, get well soon, new baby, "sorry I forgot". Customers often want to choose a date a week ahead.
• Sympathy and funeral flowers. Tributes, sprays, wreaths, letters, coffin sprays, posies for children to carry, and flowers sent to the family home rather than the service.
• Wedding enquiries. "Are you free on 14 June next year?" followed by a long conversation about colour, budget and venue.
• Event and corporate work. Weekly reception arrangements for offices, hotels and restaurants, and one-off events like a product launch or a christening.
• Stock and price questions. "Do you have peonies?" "How much is a dozen red roses?" "Do you sell plants?"
• Order chasing. "Has my order been delivered?" "Can I change the message on the card?"
• Workshops. Wreath-making in December, seasonal classes, hen-party workshops.

Most of these arrive as messages rather than calls now, and a lot of them arrive on Instagram, because that's where people found your work in the first place.

When the questions come

The daily pattern in a flower shop is front-loaded. The same-day rush runs from opening until your cut-off, with a peak late in the morning when people at work remember a birthday. After the cut-off, questions turn to "first thing tomorrow" and to planned orders. Evenings and weekends bring wedding enquiries and funeral orders from families sitting down together to plan.

The florist's calendar

You know this better than anyone, but it's worth setting it against message volume:

• Valentine's Day. The busiest few days of the year for most florists, with orders arriving weeks early and a flood on the day itself. Pre-orders, delivery capacity and red rose prices are the questions.
• Mothering Sunday in the UK falls three weeks before Easter, so the date moves each year and customers ask about it constantly. It's often the second-busiest day.
• Wedding season, roughly May to September, with enquiries for next year's weddings arriving all autumn and winter.
• Christmas: door wreaths, table centrepieces, grave pots and wreaths for cemeteries, corporate arrangements and wreath workshops.
• Graduation season, Easter, and Remembrance, depending on your area.
• Funerals don't follow a calendar. They arrive any week and can't wait.

On a Valentine's morning, any florist who's had to choose between answering the phone and finishing the order in their hands knows which one loses.

What the AI handles

Klugon's assistant works from three things you give it: your instructions, a knowledge base of documents and text, and a product list. For a florist, here's how those map onto the work.

Same-day cut-offs, enforced every time

Cut-offs are where flower shops lose orders and where they make promises they can't keep. You give the AI the rules once, in plain words, and it applies them to every message:

• "Same-day delivery must be ordered by 2 pm, Monday to Friday, and by 12 noon on Saturday. No delivery on Sunday."
• "On the day before Valentine's Day and on Mothering Sunday weekend, same-day orders close at 11 am."
• "If someone asks for same-day after the cut-off, offer first thing tomorrow or collection from the shop before 5 pm."

One honest limit: don't rely on the AI to read the clock. Write the rule so it works without knowing the exact minute: the AI states the cut-off ("same-day orders need to be in by 2 pm"), takes the order down, and the florist confirms whether it made it. For orders close to the line, that confirmation comes from you, flagged straight to your Slack. What the AI does reliably is tell every customer the same rule, every time, instead of a "we'll try" that turns into an awkward phone call at four.

Keep the wording simple, and write the exceptions out in full. If your Valentine's cut-off depends on driver numbers that you only decide in early February, update the instructions when you know.

Delivery areas and charges

Delivery questions are mostly postcode questions. Give the AI a short table, and it can answer instantly:

(That table is an example for a made-up shop. Use your own.)

Also tell the AI the things that trip up deliveries: hospitals that don't accept flowers on wards, care homes with a front desk, workplaces that need a name and floor, and whether you'll leave flowers with a neighbour or in a safe place.

Price bands, not bouquet design

Most florists don't sell fixed products for everyday orders. They sell a price point and a style: "a hand-tied in pinks and whites, about £45", "florist's choice, seasonal, £35". The AI is good at taking that kind of order down if you tell it your bands:

• Hand-tied bouquets from £35, £45, £60 and £80.
• Arrangements in a box or vase from £40.
• Florist's choice: the customer gives a price and colour preference, and you design.
• Add-ons: vase, chocolates, balloon, card upgrade.

You can also add specific products, such as your signature dozen roses or your Christmas wreath, with prices and photos. The AI can search those and suggest related ones, such as a vase with a hand-tied bouquet.

What it shouldn't do is promise particular stems. Seasonal flowers depend on what came in at the market this week. Tell the AI to say "peonies are usually available from May to early July; the florist will confirm on the day" rather than "yes, peonies".

Card messages, written down exactly

The card message is the part the customer cares about most and the part most often garbled on a phone call. In a chat, the customer types it, and the AI repeats it back word for word to check. Tell the AI to:

• ask for the message exactly as it should appear, including the sign-off;
• ask whether the card should be anonymous;
• read it back and ask for confirmation;
• never correct spelling or grammar unless the customer asks.

Funeral orders: the details that matter

Funeral work is careful work, and the order needs more information than a birthday bouquet. Give the AI a checklist so it collects everything before handing over:

• The name of the person who has died, as it should appear on any card.
• The date and time of the funeral, and the name of the funeral director.
• Where to deliver: usually the funeral director's premises, by a stated time before the service, not the church or crematorium. Some families want flowers at the house instead.
• The type of tribute: spray, wreath, heart, letters ("MUM", "GRANDAD"), cushion, posy, coffin spray.
• Colours, and any flowers the person loved.
• Whether it's from the family (often the coffin spray) or from a friend.
• The card message.
• A contact name and phone number for the person ordering.

Tell the AI to write gently and keep the questions short. People ordering funeral flowers often message late at night, and a calm, clear reply matters more than speed. Then flag the order for you straight away. Funeral tributes with letters and large sprays need you to confirm them personally, and the time pressure means you'll want to know about them early.

Wedding enquiries: book the consultation

Wedding work rarely gets decided in a chat. What the chat can do is qualify the enquiry and book the consultation. Klugon has a built-in booking system that uses your opening hours, service lengths and blocked dates. Set up a "Wedding flowers consultation, 45 minutes" service and the AI can offer real free slots and book one.

Before booking, the AI can collect:

• the wedding date and venue (and whether you've worked there before);
• the ceremony and reception venues, if different;
• rough guest numbers and the budget for flowers;
• what's needed: bridal bouquet, bridesmaids, buttonholes, corsages, ceremony arrangements, table centres, an arch;
• colour scheme and style;
• whether the date is already in your diary.

On that last point, tell the AI to check your capacity rule ("we take no more than two weddings on a Saturday") from your instructions, and to treat the date as provisional until you confirm. Klugon's booking system manages consultation slots, not your wedding capacity, and it doesn't read a separate wedding diary unless you connect a calendar.

Questions with fixed answers

A big share of florist enquiries have simple answers:

• Opening hours, and holiday closures.
• Whether you deliver to a given postcode, and the price.
• Whether you do plants, dried flowers, sympathy hampers or gifts.
• How to look after the flowers once delivered.
• Whether you sell gift vouchers.
• When your next workshop is, and how to book.
• Parking and where the shop is, with a Google Maps link shared from your links library.

What the florist order chatbot should hand to you

Some things need your eye. Write these into the instructions as "flag the conversation for a person".

Anything that changes the design or the budget

"I want something really special, about £200, she loves blue." That's a conversation you'll enjoy and the AI should set up, not finish. Tell it to take the details down and flag it.

Same-day orders close to the cut-off

At 1.50 with a 2 pm cut-off, the AI can take the order but you might not be able to make the drive. Tell it to say the order is received and the shop will confirm within a set time, and flag it at once. Klugon notifies you through Slack or Zapier; it can't ring you or send you a text, so pick the alert you'll actually see while working.

Changes and cancellations

Changing a card message or delivery address after the order's gone to the bench needs a person. So do cancellations, especially for flowers already made up. Florists often point out in their terms that flowers made to order can't be returned. The Consumer Contracts Regulations 2013 set out exceptions to the usual 14-day cancellation right for distance sales, including goods made to the customer's specifications and goods liable to deteriorate rapidly. How those apply to your orders is something to check with your own terms and an adviser; the AI should quote your terms, not interpret the law.

Complaints

"The flowers arrived wilting." "They were left on the doorstep in the rain." Apologise, collect the order details and a photo if the customer has one, and flag it. Don't let the AI offer refunds or replacements unless you've told it exactly what you offer.

Relay orders and out-of-area deliveries

If you're part of a relay network and use it for out-of-area orders, the AI shouldn't promise one. Relay terms, prices and timings are yours to decide. Tell it to take the details and flag the conversation.

Sensitive situations

Occasionally a funeral order comes with a story: a sudden death, a child, a family argument about who orders the coffin spray. The AI should take down the order details plainly and let you handle anything beyond them.

Sample conversations

The shops and customers below are made up. These are sample conversations written for this page, not recordings or real transcripts.

A sample Instagram DM: same-day, just before the cut-off

The shop is Wren & Thistle Flowers, an invented florist in Harrogate. Same-day cut-off is 2 pm on weekdays.

Behind the scenes, the AI saved Jo as a lead with the phone number, flagged the conversation and posted to the shop's Slack. The owner replied from the Klugon inbox with a payment link from the shop's own card provider. Klugon didn't take the payment.

A sample WhatsApp: funeral tribute, late evening

The shop is Old Mill Flowers, an invented florist in a market town.

The owner had told the AI not to quote prices for coffin sprays and letters, because they vary so much with size and flowers. The AI collected everything needed and flagged it. The florist rang Claire at 9 am.

A sample website chat: wedding enquiry to booked consultation

The consultation landed in Klugon's built-in booking list. The AI's reply in the chat is the only confirmation; Klugon doesn't send booking confirmations or reminders by email or text. If you want a reminder to go out, you'll be sending it yourself.

Valentine's Day and Mothering Sunday with an AI on the messages

Peak days deserve their own plan, because the AI's job changes. On a normal Tuesday it's answering the odd enquiry while you work. On 13 February it's the front of a queue.

Three weeks before

• Decide your peak rules and write them in. Pre-order deadline, same-day cut-off on the day (most florists bring it forward), delivery windows ("deliveries between 9 am and 6 pm, we can't promise a time"), and whether you'll take timed deliveries at all.
• Simplify the menu. Peak days work best with a short list: a dozen red roses, a mixed hand-tied at three prices, a florist's choice. Add those as products with photos and prices so the AI can quote them exactly.
• Set the price change. If your rose prices rise for Valentine's week, say so in the instructions with dates, so the AI doesn't quote January prices on 12 February.
• Write the "sold out" message. If you cap deliveries, tell the AI what to say once you've reached it. You'll update one line in the instructions on the day.

On the day

• Watch one place. Every order from Instagram, WhatsApp, Messenger, the website and the phone lands in the Klugon inbox, flagged. Keep the inbox open on a tablet by the bench rather than checking four apps.
• Confirm in batches. The AI has written each order up with the recipient, address, card message and price band. Confirm and send payment links in batches between bouquets rather than one by one as they arrive.
• Remember the 24-hour window. WhatsApp and Instagram replies need to go out within 24 hours of the customer's last message. For pre-orders taken a week earlier, that window will have closed, so confirm pre-orders when they come in, not on the day.

Mothering Sunday is different

Mothering Sunday orders tend to come from adult children who live elsewhere, ordering for a mum in your town. That means more out-of-area postcodes, more "I don't know her postcode, she lives near the church" and more questions about whether you deliver on Sundays. If you don't, the AI needs to say clearly that Saturday delivery is the option, and from when Saturday's slots close.

Corporate and weekly accounts

Outside the peaks, regular corporate work (hotel reception arrangements, restaurant table flowers, a weekly office vase) is good steady income that often starts as a message. The AI can take the first enquiry down: the business, the address, how often, the budget per week, and who to invoice. Tell it to flag these for you rather than quote, because weekly pricing depends on the space and the flowers you'll use.

Setting up a florist order chatbot

What to gather first

• Your cut-off rules, including Saturday and holiday exceptions, written as plain sentences.
• Your delivery table: areas, postcodes, charges, which are same-day.
• Your price bands for bouquets, arrangements, funeral tributes and plants.
• Your terms: cancellations, substitutions ("we may substitute flowers of equal value and similar colour if something is unavailable"), and what happens if nobody is home.
• Your funeral and wedding checklists of the details you need.
• Care instructions and your FAQ.
• Photos of signature products if you want them in the product list.

The checklist

1. Start the 7-day trial of the product that matches where your enquiries come from. For Instagram and WhatsApp, that's the social agent. For your website, the chat agent. For the phone, the voice agent. A card is taken at sign-up and charged when the trial ends unless you cancel.
2. Create an assistant and write its instructions: your shop, your tone, your rules, and the list of things to flag.
3. Add your cut-offs and delivery table to the instructions, and your terms, FAQ and care guide to the knowledge base.
4. Add products for anything with a fixed price and design.
5. Set up services for bookable things: wedding consultations, workshop places. Set your opening hours, slot length and closed dates.
6. Connect the channels. Instagram DMs need a professional account linked to a Facebook Page, and Meta's app review for the messaging permission before the general public can message the assistant; Meta's Instagram messaging documentation explains the requirements. WhatsApp uses Meta's WhatsApp Cloud API with your own WhatsApp Business number. The website widget is one script tag pasted into your site, and works whether your site is on Wix, Squarespace, Shopify or WordPress. It's a paste, not a plugin.
7. Connect Slack or Zapier for alerts. These need the Business tier.
8. Test every awkward order: a same-day request at 1.59, an out-of-area postcode, a funeral with letters, a card message with an emoji, a Valentine's order in July.
9. Before Valentine's and Mothering Sunday, update the cut-offs and add the pre-order details.

If your enquiries come mostly from your website, the AI chatbot for your website page explains how the widget works, how to set its colours and where it appears.

On Instagram specifically

Florists live on Instagram, and DMs are where a lot of orders start. Be aware that connecting Instagram DMs needs a Meta app with an approved messaging permission. Until Meta approves it, only people with a role on your Meta app can message the assistant. Budget time for that review, and keep answering DMs yourself until it's through.

Costs, worked through

Klugon's prices are in US dollars and are taken from our pricing as of October 2026.

Choosing the plans

• Social agent (Instagram, WhatsApp, Messenger): Starter $79 a month for 300 conversations, Business $249 for 1,500. WhatsApp adds $19 a month for the number.
• Chat agent (website widget): Starter $49 a month for 100 conversations, Business $149 for 400.
• Voice agent (phone): Starter $99 a month with 1,000 minutes, plus $9 a month for a phone number, which you can add once the trial ends.

Yearly billing is 20% less. Slack and Zapier alerts need Business on the voice or social plan, so a florist who wants an alert for every same-day order should budget for social Business.

A worked example

Assumptions, which you should replace with your own numbers:

• A one-shop florist with the owner and one part-time assistant.
• 120 order enquiries a month across Instagram and WhatsApp in a normal month, and 400 in February.
• 1 in 5 messages arrives while you're both on the bench or out delivering, and today half of those get answered too late to win the same-day order.
• Your average same-day order is £55 including delivery.

In a normal month that's 24 enquiries arriving at bad moments and 12 that go elsewhere: about £660 of orders. In February, using the same proportions, it's 40 lost orders, or about £2,200. Klugon's social Business plan plus WhatsApp is $268 a month.

You won't win every one of those back. Some customers wanted a different style, some were shopping around on price. But the cut-off is a hard deadline, and an enquiry answered at 1.15 has a chance that one answered at 2.30 doesn't.

Compared with the alternatives

• A Saturday or seasonal helper on the phone costs wages for the hours they're in, and doesn't cover 10 pm funeral messages.
• An answering service can take a message but usually doesn't know your cut-offs, postcodes and price bands, so you still call every customer back.
• A website order form works for fixed products and takes payment, but most enquiries that start with "can you do something for…" end up as a message anyway.

Data, privacy and the card in the bouquet

Florists handle more personal information than they think:

• the sender's details and the recipient's name and address;
• card messages that reveal relationships, apologies and illnesses;
• funeral details, which can reveal a family's religion through the service type.

Religious belief is special category data under UK GDPR, according to the ICO's guidance on special category data. You don't need to ask about it to make a tribute, so tell the AI not to. If a customer volunteers it, it's part of the order conversation; keep access to your inbox limited to staff who handle orders.

Klugon is run by a German company, Softaar Technologies UG, under the GDPR, with sub-processors and transfer safeguards listed in its privacy policy. You can export conversations as CSV and delete them in the dashboard. Full erasure is by email request.

How it compares with florist order systems

Florists usually run some mix of these. Klugon slots into the gaps rather than replacing them.

If you've used a general chat widget before and found it couldn't do much beyond collect an email address, our Tidio alternative comparison covers how an AI assistant that answers from your own knowledge differs.

Florists overlap with several other kinds of shop in Klugon's retail AI assistant pages. Gift shops share the seasonal peaks, clothing boutiques share the Instagram-first customers, and grocery stores share the delivery-area questions.

Knowing whether it's working

After the first month, and especially after your first big holiday with it, check:

• The busiest hours chart. If the peak sits just before your cut-off, the AI is earning its keep.
• Leads by channel. See how many orders started on Instagram against WhatsApp and the website.
• Bookings. Count wedding consultations booked by the AI and how many became weddings.
• Questions without a good answer. Usually price questions you haven't covered, or new venues. Add them.
• Handover share. Every order should be flagged for you to confirm. Questions answered without you are the time you got back.

Frequently asked questions

Can the florist order chatbot take payment?

No, and we don't recommend it. The AI takes the order down and hands it to you. You confirm it and take payment the way you already do, by card over the phone, a payment link from your own provider, or in the shop.

Does it know my same-day cut-off?

Yes, if you write it into the instructions, including different times for Saturdays and the peak days you list. It tells every customer the rule and takes the order down. Whether a borderline order made the cut-off is confirmed by you, so don't rely on the AI to judge the exact minute.

Can it check whether I have a particular flower in stock?

No. It doesn't see your cooler. Tell it which flowers are seasonal and when, and to say the florist will confirm on the day.

How does it handle funeral orders?

It collects the details you list: funeral date and time, funeral director, tribute type, colours, card message and a contact number. It writes gently and flags the order for you to confirm, rather than quoting a price for a complex tribute.

Can it book wedding consultations?

Yes. Set up a consultation service with a duration and your opening hours, and the AI offers free slots and books them in Klugon's booking list. It can't decide whether you'll take the wedding date; that stays with you.

Will customers get a confirmation or reminder by text or email?

No. The AI confirms in the conversation itself. Klugon doesn't send texts, emails or reminders to customers. Your own confirmation, with the price, comes from you.

Can it answer the shop phone too?

Yes, with the voice agent. It answers in the same way and every call shows up in your inbox. It can't put a caller through to you; it flags the conversation so you can call back.

Does it work with Instagram DMs?

Yes, after Meta setup. You'll need a professional Instagram account linked to a Facebook Page, a Meta app, and Meta's approval for the messaging permission before the public can use it.

What about Valentine's Day?

Update your cut-offs, price bands and pre-order dates in the instructions in January. The AI takes every enquiry as it comes in, and you'll see them all flagged in one inbox rather than across three apps.

Can I make it sound like my shop?

Yes. Write the tone into the instructions and give it your own greeting. Most florists want it warm and short, and gentle for anything to do with funerals.

Klugon's AI receptionist for small business is easiest to judge with your own enquiries, so try it during the trial with last February's messages. Plans are on the pricing page.