armincx works on a flat-rate model. You pay no per-agent seat costs and no per-ticket costs. This sets armincx apart from Zendesk (seat costs) and Gorgias (ticket costs). For Shopware shops with growing volume this means: your costs stay predictable even when ticket volume triples in Q4.
Meta costs for WhatsApp messages are 11 cents per marketing message in Germany and around 5 cents in Austria and Switzerland (as of 2026). Service messages within the 24-hour window are free. WhatsApp generates 2 to 3x higher revenue per recipient than email. After deducting the Meta fee, you retain 139% higher net margin than via email.
armincx software pricing is not publicly communicated. In a demo we calculate your individual business case together based on ticket volume, number of channels and desired automation level.
Many Shopware retailers use Zendesk or Freshdesk for customer service. Both are solid ticketing systems. What they lack is the depth of shop integration.
Zendesk displays your Shopware data in a side panel. But every action (cancellation, address change, refund) requires manual clicking in Shopware. The Zendesk AI generates text suggestions but does not execute backend actions. Large Zendesk installations cost an additional EUR 20,000-40,000 in setup through implementation partners like Leafworks.
Freshdesk has the same problem: the Freshdesk AI (Freddy) is a separate, paid add-on and hard to train. Shopware data is displayed, but actions require manual intervention.
armincx was built from the ground up as an AI-first platform. The AI executes actions directly in Shopware. No copying between tabs, no manual clicking. And the onboarding is handled personally by the Chatarmin team, in German, with no external agency.
The AI automatically recognizes when it cannot confidently answer an inquiry. In these cases it escalates to a human agent. The AI delivers a prepared draft and all collected customer data. The agent only needs to review and send.
For incomplete inquiries (e.g. missing order number, no photos for a complaint), the AI first asks for the missing information before escalating. This saves your team the back-and-forth and ensures a complete case arrives at the agent.
You configure the escalation rules yourself. Example: refunds under EUR 50 are handled by the AI independently. Above EUR 50 it creates a draft for approval. Complaints on high-value products go directly to a senior agent. Legal threats are immediately prioritized.