Marshmallow bouquets: the bot vs Instagram and Facebook
05.10.2026
Comparing two sales channels for marshmallow bouquets: a Telegram bot with a mini-app versus a storefront on Instagram and Facebook. Each point has a table, then an honest look at where social media are stronger, and finally a brief conclusion.
Why this comparison makes sense at all
A marshmallow bouquet is a date-driven product: people buy it for March 8, birthdays, anniversaries. So three things decide the outcome: how fast an order can be placed, whether it gets lost at peak time and whether the reminder reaches the customer on time.
Social media are strong at showing, the bot is strong at the deal. These are two different halves of one funnel, and confusing them is the main mistake when choosing a channel.
Point-by-point comparison
Bot — Social media — Draw
| # | Criterion | Via the bot | Instagram and Facebook | Edge |
|---|---|---|---|---|
| 1 | 1. Placing an order | Via the botThe customer assembles the bouquet in the app: catalog, cart, ingredients, total — everything on screen | Instagram and FacebookOrders described in words in DMs: "can I have roses and no poppies?" — ingredients and price are confirmed in a back-and-forth | Edge Bot |
| 2 | 2. Up-to-date price and availability | Via the botOne place to edit — the admin panel; the buyer sees exactly what the operator sees | Instagram and FacebookThe price lives in the post caption and pinned stories and quietly goes stale | Edge Bot |
| 3 | 3. Losing an order | Via the botThe order lands in the database: the operator gets a list, the client gets a pickup code | Instagram and FacebookThe message drowns in DMs; it gets lost in the March 8 rush | Edge Bot |
| 4 | 4. Speed of ordering | Via the botA few taps — the only thing to type is the phone number | Instagram and FacebookMessaging: the client waits for a reply, the operator answers everyone by hand | Edge Bot |
| 5 | 5. Repeat orders | Via the botOrder history in the app — "same as last time" in two taps | Instagram and FacebookYou have to describe the order all over again | Edge Bot |
| 6 | 6. Notifying the client | Via the botThe bot writes on its own: accepted, ready, ready for pickup — with a pickup code | Instagram and FacebookThe operator writes to everyone by hand | Edge Bot |
| 7 | 7. Reminders before the date | Via the botTechnically possible to send to those who bought before last March 8 | Instagram and FacebookBroadcasts reach only followers, and even that reach is cut | Edge Bot no broadcast yet |
| 8 | 8. Your own client base | Via the botPhone, history, language — in your own database on your own server | Instagram and FacebookThe audience belongs to the platform; exports are limited | Edge Bot |
| 9 | 9. Client data | Via the botThe phone number comes from Telegram with consent and is stored by you | Instagram and FacebookThe phone number stays in the chat, i.e. with the platform | Edge Bot |
| 10 | 10. Fee per order | Via the botNone | Instagram and FacebookNone, but promotion is paid | Edge Draw |
| 11 | 11. Cost of the channel | Via the botOne server ≈ $3.34 per month (per this project's estimate) | Instagram and FacebookThe account is free, ads are optional | Edge Social media |
| 12 | 12. Risk of being blocked | Via the botThe bot isn't blocked for content, but the channel lives inside Telegram | Instagram and FacebookA ban or shadowban zeroes out the channel in a day | Edge Bot |
| 13 | 13. Showcasing work | Via the botOnly a catalog: a photo in the product card | Instagram and FacebookFeed, reels, stories: process, emotion, "how it's made" | Edge Social media |
| 14 | 14. New clients | Via the botNo organic discovery: people arrive via a direct link | Instagram and FacebookGeotags, hashtags, recommendations, reposts | Edge Social media |
| 15 | 15. Peak time | Via the botThe rush is handled as a list; nothing gets lost | Instagram and FacebookThe more orders, the higher the chance of drowning some in DMs | Edge Bot |
| 16 | 16. Payment | Via the botOn pickup only: this app has no online payment | Instagram and FacebookA card transfer, but also manual and without marking it paid | Edge Draw |
| 17 | 17. Entry barrier for the client | Via the botTelegram is usually already installed; ordering takes one tap from the bot | Instagram and FacebookYou need an account and a follow; part of the audience doesn't use Instagram | Edge Bot |
| 18 | 18. Analytics | Via the botYour own numbers: orders, totals, repeat purchases | Instagram and FacebookBuilt-in reach statistics, but not sales | Edge Social media |
| 19 | 19. Speed of launch | Via the botYou need a server, a domain, a bot — i.e. development | Instagram and FacebookAn account and posts — you can start today | Edge Social media |
| 20 | 20. Multiple locations | Via the botThe catalog and pickup codes are shared; a location is added as one line in the order | Instagram and FacebookEach location needs a separate account and a separate feed | Edge Bot |
| 21 | 21. Seasonal load | Via the botOne order list for everyone, statuses visible | Instagram and FacebookChat volume grows faster than the ability to reply | Edge Bot |
| 22 | 22. Trust in the order | Via the botYou see the ingredients, price, pickup code and address — before paying | Instagram and FacebookThe client confirms the price and ingredients in chat | Edge Bot |
Where social media are objectively stronger
This isn't a verdict on the bot but the limit of where it applies — without it the comparison would just be an ad.
| What | Why the bot can't replace it |
|---|---|
| Showcasing work | Why the bot can't replace itA marshmallow bouquet is bought with the eyes. The feed and reels show process and emotion; a catalog shows only the result |
| Inflow of new clients | Why the bot can't replace itThe bot has no organic reach: it can't show itself to someone who doesn't know it. Someone has to bring the link to the bot |
| Cost | Why the bot can't replace itThe account is free, the server costs money — not much, but ongoing |
| Quick start | Why the bot can't replace itA feed can be started today; a bot needs development and a launch |
| Reach statistics | Why the bot can't replace itPlatforms have ready audience analytics; the bot has it only for its own orders |
A brief conclusion
Social media are the storefront, the bot is the checkout. For marshmallow bouquets the bottleneck isn't showing the product but ordering, peak time and date reminders — exactly what the bot handles and DMs don't.
So the sensible setup is not "a bot instead of social media", but the bot as the main sales channel with social media as the storefront: the feed and reels show the bouquets and bring people in, the link leads to the bot, where the order is placed in seconds, never gets lost and builds history for repeat sales.
Starting with "a bot instead of social media" isn't worth it: there's nowhere for the bot to get clients. Fully swapping social media for the bot isn't either: you lose the showcase that sells marshmallow in the first place.
| Where the bot wins | What the bot can't do yet |
|---|---|
| The order isn't lost, the price doesn't go stale, the status updates itself | What the bot can't do yetNo online payment — pickup only |
| Your own client base and order history | What the bot can't do yetNo date-based broadcast — nothing to remind with yet |
| Zero fees, the channel won't be blocked for content | What the bot can't do yetNo organic reach — social media brings the traffic |
Those last three points aren't a verdict but a list of what's worth improving if the bot becomes the main channel.