💼 RetailCRM to align sales, marketing, and service around a unified customer story
RetailCRM has announced Release 12, titled 'Retail with AI,' which will take place on October 29, 2026, in Moscow at 'Soglasie Hall' (Profsoyuznaya, 36s1). The key innovation is the 'Three Atoms Model,' which integrates sales, marketing, and service into a single CRM centered on a shared customer history. Sales teams can view open inquiries before product recommendations are made, marketing automatically suppresses promotional campaigns for customers with unresolved complaints, and service teams receive full context within a single ticket. The release will include AI agents for routine inquiries that provide ready-made responses to specialists, leaving the final decision to humans, as well as a new customer service module that aggregates inquiries alongside purchase and campaign history. The program features a panel with representatives from Hoff, Askona, Familia, HENDERSON, Lemana PRO, and Geltek, along with presentations by Dmitry Borodin (CEO, RetailCRM), Ilya Salikhov (CTO), Evgeny Butman (ECS Group), and Yuliana Gordon (Ivengo).
🌍 For the retail industry, this represents a concrete mechanism for embedding AI agents into CRM processes: not a standalone chatbot, but a unified contextual data layer where the agent's utility is determined by the completeness of the customer history as much as by model quality. The human-in-the-loop scheme—where the agent proposes a response and the specialist makes the decision—reduces the risk of erroneous automation, while the automatic suppression of promotional campaigns during unresolved complaints minimizes losses from departmental desynchronization. However, the announcement declares the approach without confirming its operational viability: metrics such as the percentage of accepted drafts, processing time, and error rates are currently unavailable. Another open question is that consolidating purchases, inquiries, and campaigns into a single profile expands access to customer data within the company, yet the mechanism for access control is not described in the announcement.
👤 If you work in retail, e-commerce, or marketing, this is an opportunity to observe how AI agents are integrated into CRM processes and where the human role remains: specialists receive a draft response but make the final decision themselves. For those implementing agents in their services, it is useful to study RetailCRM's approach to human-in-the-loop and data integration from different departments, and to verify access control permissions in advance before consolidating customer data into a single profile.
Source 1: https://www.retailcrm.ru/events/release-12