Case study

How SFD turns 26%
of chats into cart-adds

SFD is one of Poland's best-known sports-nutrition retailers — protein, creatine, vitamins and pre-workouts under its own Allnutrition, SFD Nutrition and Alldeynn brands, sold online alongside a chain of stores and gyms.

4 min read

SFD Assistant
Type your question
Powered by spot

Catalog

2,216 supplements in the bio range

Helpdesk

Freshdesk

Results

Key results

Measured in AskSpot across 1,709 chats over 30 days — what the agent does for the basket, and how much of the queue it closes on its own.

26%

of chats fill a basket

3.1

items in an average chat basket

56%

product recommendation click-through

73%

of chats resolved with no human

“AskSpot is a genuine asset to our customer support team. It efficiently takes over routine inquiries via chat and email, freeing up our agents to concentrate on more challenging matters. As a result, customers enjoy prompt, clear and consistent responses, which has noticeably enhanced both their experience and our overall service quality.”

Krzysztof Cop

★★★★★

Krzysztof Cop

Customer Service Manager, SFD

Before AskSpot

A wall of supplements, and a shopper who needs a person to choose

SFD sells thousands of supplements that all promise something — a shelf of creatines, proteins split by goal and by diet, pre-workouts that read like a chemistry set. A shopper who already knows they want to cut fat or add muscle still has to work out which product, in which form, at which dose actually fits — and that is exactly the question a category page answers badly.

The other half is logistics and rules: where an order is, whether it makes the InPost cut-off, how a discount code stacks with loyalty points. Demand runs into the evening and the weekend, around training — precisely when the phone line and the inbox are shut. And a slice of it is health-adjacent, where the responsible answer is sometimes not to answer at all.

Cover outside opening hours

none

Questions that repeat most

product choice & delivery

Peak enquiry time

evenings & weekends

Queues to staff

chat & Freshdesk email

Rollout

How they rolled it out

Storefront advice first, because that is where the sale is won; then live order data, because delivery and status are a third of everything asked; then the email queue, so the same answers serve both channels.

AI Sales Agent

Step 01

Catalogue & Stock Sync

Pre-sale advice on the storefront

The full bio range went in — Allnutrition, SFD Nutrition and Alldeynn — so the agent can weigh a monohydrate against a malate, read a label for what a pre-workout actually contains, and filter by diet such as gluten-free. Where a question turns medical, it says so and points to a specialist rather than guessing.

What shoppers ask

  • I want creatine but I don’t know which one
  • Does Redox Hardcore 2.0 contain omega-3?
  • Gluten-free products safe for celiac?

AI Sales Agent

Step 02

Order API Integration

Delivery dates and order status

Delivery terms and order status are more than a third of everything asked, and the InPost cut-off decides whether a parcel ships today — so both come from live order data, not a policy page.

What shoppers ask

  • Order before 6 pm — will InPost arrive tomorrow?
  • When will my order arrive?
  • Can I still change the delivery address?

AI Inbox Agent

Step 03

Freshdesk Integration

The post-sale email queue

2,532 tickets a month joined the same catalogue and order data, so a returns rule or a delivery cut-off only has to be right once to be right on chat and in the inbox — answered in the customer’s language, SFD’s foreign markets included.

What shoppers ask

  • When will my order from 14 July be dispatched?
  • I need a VAT invoice for my order
  • Sent my return two weeks ago, still no refund

Measured in AskSpot

What the agent does in a month

4,200

conversations handled a month

58%

of chats come in after hours

10

messages per conversation

6 s

average first response

From real conversations

What SFD shoppers actually ask

Two things shape a supplement sale: which product fits the goal, and when it turns up. Choosing between a monohydrate and a malate, checking what's in a pre-workout, or asking whether a protein suits a teenager is the single biggest slice — and delivery terms and order status make up most of the rest.

Product advice & specs29.8%
Delivery & shipping27.0%
Discounts, codes & loyalty15.9%
Order status & tracking13.8%
Changes & cancellations7.1%
Returns & complaints6.1%

Based on 1,706 tagged conversations over 30 days · about 10 messages each · 58% arriving outside business hours.

Post-sale inbox

The email queue runs on the same product knowledge

SFD's Freshdesk queue — order status, delivery chases, returns, complaints and invoices — runs through the AI Inbox Agent alongside the on-site chat. Routine tickets are answered end to end, 24/7 and in the customer's language; anything that needs a judgement call is triaged to the team.

23%

593 of 2,532 support emails resolved end to end by AI over 30 days — the rest are triaged and routed to the right person.

Resolved by AI

  • PLWhen will my order from 14 July be dispatched?Order status
  • PLFrom what amount is delivery free?Delivery
  • PLI need a VAT invoice for my orderInvoice
  • DEPlease cancel the order I placed todayCancellation
  • PLWrong delivery address — please correct itAccount

Sent to your team

  • PLPackage arrived damaged — photos attachedComplaint
  • PLRedeemed points but was still charged full priceBilling
  • ENSent my return two weeks ago, still no refundRefund

More case studies

See how other stores use AskSpot

More stories from teams that automate support and grow sales with AskSpot.

All case studies

4FIZJO

How 4FIZJO automated 50% of customer inquiries with AI

4FIZJO cut website calls and messages by 40% and automates a third of its helpdesk tickets. AskSpot's AI Chat resolves 74% of its 21,360 monthly chats with no human, across 6 markets, and its Inbox Agent clears email and Allegro tickets in Thulium.

Read the story →

KODANO

How KODANO resolves 81% of eyewear chats with no human

KODANO's AskSpot agent resolves 81% of chats with no human across 10 European storefronts — diopters, PD, lens behaviour and where the order is. 13% of chats add to the cart, and the inbox agent auto-resolves 43% of post-sale email.

Read the story →

badaj.to

How badaj.to turns health questions into booked lab tests

badaj.to's AskSpot agent finds the right blood test from a name or symptom, strictly non-diagnostic — 8.4% book, €29,900/mo, 81% resolved with no human.

Read the story →

Morele.net

How Morele.net cut customer service costs by 90%

Morele.net cut customer service costs by 90% and went from 50 support agents to 5. AskSpot resolves 71% of its 37,600 monthly chats with no human, and answers Allegro marketplace questions from the same catalog.

Read the story →

Militaria.pl

How Militaria resolves 83% of chats in 9 languages

Militaria.pl runs AskSpot AI chat across three domains in nine shopper languages: 48% product recommendation click-through, 27% add-to-cart after a chat and 83% of chats resolved with no human — 16,600 conversations a month, over half after hours.

Read the story →

Sport-Shop

How Sport-Shop turns 17% of chats into cart-adds

Sport-Shop.pl's AskSpot chat agent puts an item in the basket in 17% of conversations and resolves 88% with no human — delivery timing, sizing and fit for camping, beach and garden gear. 15,800 conversations a month across chat and email.

Read the story →

25case studies

See every story

Every merchant we have measured — what shoppers asked, what the agent did, and the numbers it moved.

Browse them all →

Test before you trust

Worried AI will cost you customers?

You never flip a switch and hope. Every answer is checked by our AI, our specialists and your own team before launch — and if you want extra reassurance, you can go live after hours first.

Layer 01

Self-testing AI

Every reply is checked automatically against your catalog and your rules, so off-topic or off-brand answers are caught before anyone sees them.

Layer 02

Vetted by experts

Our specialists challenge it with all kinds of questions, drawing on experience from dozens of e-commerce projects we've run, then sign off on the quality.

Layer 03

Tested on staging

Your own managers get a private staging environment to try the chat end to end — ask anything, push the hardest cases — and approve it before go-live.

Layer 04

After-hours launch

Want extra reassurance? Optionally go live only outside working hours at first, review every answer and the full statistics, then expand when you're ready.

You decide when — and if — AskSpot goes live.
No leap of faith, no lock-in.

Ready to make every conversation sell?

Book a demo and see AskSpot on your own catalog. Onboarding takes a few days.