Gartenglück24 GmbH, 25 employees (online retailer for home & garden)
Sells about 6,000 home and garden items via its own Shopware store, Amazon, eBay and Kaufland; the biggest time sinks are customer mail, returns and product data.
How it works
- Check the numbers. Each step shows the hours your company spends on it per week. Click a number to change it, or set it to 0 if the step does not exist for you.
- Insert building blocks. Click a coloured block in a step. Below it you read in plain words what the AI would do there and what stays with your people. The shadow under the step shrinks.
- Read the result. The dark bar at the bottom shows the key figures; the full calculation with cost, return and a curve is below the departments.
Block colours show how you get it:
A dashed box means no building block exists yet.
Where AI helps most here
Start with what saves the most. Many blocks run on the same platform (e.g. one AI workplace for the office): you buy it once, and every further step only needs configuring.
Biggest single levers
Packages: many steps in one go
Link copied. It holds all your settings, nothing is stored on a server.
The company from above
Each building is a department. Its height shows how much work happens there, each window is a recurring task, and the shadow on the ground is the manual work. Click a building to open it. Windows light up where AI helps, and the shadow gets shorter.
Customer Service & Returns
13 steps, 219 h per week
Customer enquiries
- Answer shop mailboxService agent h/week
- Where is my parcelService agent h/week
- Answer marketplace messagesService agent h/week
- Handle phone callsService agent h/week
- Pre-sales product questionsService agent h/week
Returns handling
- Register and inspect returnsReturns clerk h/week
- Decide refund or replacementService agent h/week
- Regrade and restock returnsReturns clerk h/week
- Warranty and damage claimsService agent h/week
- Analyse return reasonsService lead h/week
Order checks
- Check suspicious orders (invoice purchase, prepayment)Service agent h/week
Reviews & reputation
- Answer shop and marketplace reviewsService agent h/week
- Check seller ratings and claimsService lead h/week
Product Data & Marketing
12 steps, 184 h per week
Product data
- Import supplier dataProduct data clerk h/week
- Write product textsContent editor h/week
- Edit product photosPhoto editor h/week
- Map marketplace categories and attributesProduct data clerk h/week
- Maintain prices and promotionsCategory manager h/week
Online marketing
- Plan and send newslettersMarketing manager h/week
- Manage Google and marketplace adsMarketing manager h/week
- Write SEO category textsContent editor h/week
- Social media postsMarketing manager h/week
- Weekly sales reportingMarketing manager h/week
Season & range planning
- Observe competitor pricesCategory manager h/week
- Plan seasonal rangeCategory manager h/week
Purchasing & Warehouse
11 steps, 281 h per week
Purchasing
- Plan replenishment needsBuyer h/week
- Create purchase ordersBuyer h/week
- Chase suppliers and confirmationsBuyer h/week
- Negotiate terms and compare quotesBuyer h/week
Inbound goods
- Match delivery notes to ordersWarehouse clerk h/week
- Unload and put away goodsWarehouse worker h/weekHands-on work, AI can hardly help
- Cycle counts and inventoryWarehouse clerk h/weekNo building block yet: Counting is done by hand or with handheld scanners in JTL-WMS; camera or drone counting does not pay off in a small warehouse. AI could at most flag stock discrepancies from the booking data.
Pick, pack, ship
- Pick ordersWarehouse worker h/week
- Pack parcelsWarehouse worker h/weekHands-on work, AI can hardly help
- Print labels, book freight forwardersShipping clerk h/week
- Handle shipping exceptionsShipping clerk h/week
Office & Administration
11 steps, 105 h per week
Accounting & payments
- Process incoming invoicesAccountant h/week
- Reconcile PayPal, Amazon, Klarna payoutsAccountant h/week
- Dunning and open itemsAccountant h/week
- Month-end and VAT/OSSAccountant h/week
Compliance & legal
- Keep legal texts and prices compliantManaging director h/week
- Report packaging to LUCIDOffice clerk h/week
- Answer privacy requestsOffice clerk h/week
People
- Plan shifts and seasonal helpWarehouse lead h/week
- Screen applicationsManaging director h/week
IT & shop operations
- Fix interface errors Shop-Wawi-marketplaceIT / shop admin h/week
- Employee IT supportIT / shop admin h/week
Your calculation
Example calculation for orientation. Prices, hours and savings are estimates for a fictitious company, not an offer, a promise or advice. Your real figures depend on your processes, your software and the providers you choose.
Insert building blocks or a package to see the calculation.
What you put in
Click a row to see the individual building blocks.
What comes back
- Time saved
- Value of that time
The value is working time that becomes free for other work; it is not automatically lower wages. The curve counts the freed time from the first month on. All figures are rough estimates.
Parts list for your selection
| Block | Used in | Saves | Value per year | One-off | Monthly | Data | Risk of failure |
|---|
Every row shows the amount that goes into your calculation; the rows add up to the totals. Blocks on a platform (AI workplace, automation, transcription, accounting AI, ...) only cost their configuration; the platform itself appears once with its rollout and its seats or base fee (prices researched 2026). Stand-alone solutions use the middle of the catalog price band, scaled to the size of the company.
What is still missing
Steps where AI could help but the catalog has no building block yet.
- Cycle counts and inventory Purchasing & Warehouse
Counting is done by hand or with handheld scanners in JTL-WMS; camera or drone counting does not pay off in a small warehouse. AI could at most flag stock discrepancies from the booking data.
Assumptions and sources
Fictitious but typical firm: about 9 in warehouse and shipping, 4 in customer service, 4 in product data and marketing, 2 in purchasing, 3 in accounting and admin, 1 in IT/shop, 1 managing director and part-timers/seasonal help. Roughly 400-500 orders per day in peak season, 250 on average, return rate about 12-15 percent in garden furniture and 20 percent in textiles-like decor. Hours are team-wide weekly estimates (average week, spring peak not included) and deliberately stay below full capacity because of meetings, holidays, breaks and unmodelled small tasks. Sources were used only as orientation for orders, return rates and cost levels and were not freshly re-verified; all hours are careful estimates. Legal frame: GoBD, DSGVO, Widerrufsrecht (14 days), VerpackG/LUCID, Preisangabenverordnung, OSS for VAT.