Agentic Outreach Infrastructure

Why Outreach Needs Infrastructure Now

The outreach landscape has changed. Spam filters are more aggressive, inboxes are more crowded, and recipients have less tolerance for generic automation. The organizations that still get replies are the ones that lead with a concrete finding — a visibility gap, a misclassification risk, a competitor advantage — not a pitch.

BackTier builds outreach systems that earn attention by showing recipients something they did not know about how they are being found, understood, or excluded by AI systems. The reason for contact is always a diagnostic finding, not a promotional message.

Market and Target Mapping

Before any email is drafted, the system maps the target universe. BackTier agents search by city, state, industry, vertical, organization type, campaign type, civic category, and professional service category. Every target is collected with organization name, website, contact email, contact person, role, city, state, category, vertical, and source URL.

Deduplication and suppression run automatically. Before any lead enters the system, it is checked against existing records, do-not-contact lists, unsubscribed lists, bounced lists, and previously replied lists. If any suppression match exists, the lead is excluded.

Lead and Entity Research

Every qualified lead receives a visibility snapshot before any email is drafted. The snapshot identifies what the organization appears to do, what market or issue it serves, the likely AI and search visibility problem, the likely entity clarity problem, competitor or category gaps, the best outreach angle, and the recommended offer.

Leads are scored from 0 to 100 across ten factors: category fit, website quality, AI visibility opportunity, local or market relevance, likely budget, decision-maker clarity, public-facing visibility need, urgency signal, source confidence, and email quality. Each lead receives a score explanation in plain language.

Diagnostic Outreach

Email drafts are generated using the visibility snapshot. The default style is direct, serious, concise, and diagnostic. No fake urgency. No hype. No manipulative language. No guaranteed results. The CTA is always a low-friction ask: a free AI visibility evaluation that shows how the organization appears across major AI search tools, what is missing or unclear, and what should be fixed first.

Four base templates cover general business, political and civic organizations, regulated and high-trust industries, and agencies and consultants. Each template is personalized with the organization name, contact name, service category, and market.

Approval Queues and Governance

By default, every email requires manual approval before it is sent. Campaign settings control auto-send rules, maximum emails per day, maximum emails per week, allowed sending days, allowed sending time windows, excluded domains, excluded categories, excluded cities and states, and excluded keywords.

Default settings: manual approval required, maximum 50 emails per day, sending Monday through Friday, send window 9:00 AM to 3:00 PM Eastern. These defaults can be adjusted per campaign.

Sequence Stop Rules

The system stops automatically when a reply is received, when a lead unsubscribes, when an email bounces, or when a lead is manually marked do-not-contact. No sequence continues past any of these events under any circumstances.

The four-step follow-up sequence runs at Day 0, Day 3, Day 7, and Day 14. Each step uses a distinct message that adds new context rather than simply repeating the original ask. The Day 14 message closes the loop cleanly without pressure.

Reply Intelligence

Every reply is classified into one of fourteen categories: Interested, Wants Audit, Booked Call, Not Now, Not Interested, Wrong Person, Referral, Unsubscribe, Angry or Negative, Bounce, Out of Office, Partnership Interest, Media Interest, and Political or Civic Interest.

Each classification triggers a specific action. Interested leads and audit requests stop the sequence and create an audit task. Booked calls update the lead status and notify Jason. Referrals create a new lead record linked to the original. Unsubscribes and angry replies mark the lead do-not-contact permanently.

Opportunity Escalation

When a lead requests the free evaluation, an audit task is created with the full context: organization name, website, contact name, contact email, category, market, source URL, original email thread, requested date, priority, notes, and recommended audit type.

Audit types include Business AI Visibility Audit, Local AI Visibility Audit, Political and Civic Visibility Audit, Regulated Market Visibility Audit, Agency and Partner Audit, and Reputation and Narrative Audit.

Weekly Intelligence Briefs

Every Friday, the system generates a weekly intelligence report covering total sent, total replies, positive and negative replies, unsubscribes, bounces, booked calls, audits requested, best-performing verticals, weakest verticals, best subject lines, best CTAs, best markets, and recommended changes for the following week.

Daily summaries are sent at the end of each sending day with a full breakdown of leads found, deduplicated, suppressed, drafted, approved, sent, failed, replies received, interested replies, unsubscribes, bounces, top verticals, top cities, organizations contacted, and next scheduled follow-ups.

Map

Build the target universe by category, geography, and vertical.

Diagnose

Research every target and generate a visibility snapshot.

Draft

Generate personalized diagnostic outreach using the snapshot.

Approve

Review and approve every email before it is sent.

Deploy

Send within daily volume limits and time window controls.

Classify

Classify every reply and trigger the correct action.

Escalate

Route opportunities to audit tasks and call bookings.

Report

Deliver daily summaries and weekly intelligence briefs.