ABM Персона

Account base

One context per company: from registry data to touch history and deal potential.

• CSV/XLSX import with remembered column mapping; Bitrix24 via an inbound webhook; amoCRM on request
• Account card: summary, touch timeline, duplicates and merge, brief and dossier, decision makers found
• Enrichment by tax ID from the open data of the Federal Tax Service and the SME registry
• Buying committee: which roles are found — and who is missing

Three priority axes

Computed on your data, without purchased intent sources.

• RFM — recency, frequency and monetary value on a 1–5 scale
• Lifecycle stage: new → active → at risk → dormant → reactivated
• Engagement — a weighted sum of events with a threshold tuned to your base
• Segments with readable names, a rule builder, a segment from a plain-language description, membership history

Scoring

ICP fit — similarity to won deals; the qualification moment; lookalikes of buyers.

• Recalculated on schedule and on demand
• The score is explained in the card — the weights are visible

Touch sequences

A playbook of steps with delays and conditions runs on a whole segment.

• Channels: email, SMS and Telegram through your own provider account (UniSender Go, MTS Exolve, a Telegram bot)
• A playbook library, email templates, A/B comparison of a step with traffic moving to the winner
• Approval before launch, a delivery log with the reason for every skip
• Events create touches; replies over IMAP are matched to the campaign
• Event evaluation on the Goal → Experience → Change → Result method: a metrics plan, approval, a verdict, a report and an evaluation portfolio

Results and attribution

What every channel and every step contributed — by touches, not by the last click.

• Plan and actual revenue by segment and playbook
• Attribution: first touch, last touch and the Shapley model
• The three R's of ABM, an event feed, activity surges

Tasks and recommendations

A task list for managers and AI recommendations for the next step with a useful / not-useful signal.

Behaviour and analytics

A first-party site tracker with a consent banner: visits, pages, sources and identified companies on the Site visits screen. Yandex.Metrica write-back of segments and goals.

AI agents

Nine agents behind one gateway: a budget, a call journal, schema-checked answers, no personal data sent to the model.

Personal data and 152-FZ

Contacts are never stored in the clear; tenants are isolated at the database level; a journal with a chain of signatures; erasure on request; separate consents.

Administration

Workspaces and five roles, including Sales: a salesperson handles tasks and touches and opens a contact with a button that is logged. API keys, signed webhooks, notifications, billing; in a customer installation — the Licence and AI providers screens.