Skip to content

Playbooks

A playbook is a connected agentic workflow: a chain of your AI colleagues that work in sequence to solve one specific go-to-market job, end to end. Instead of running each agent by hand and stitching the outputs together yourself, you point a playbook at a set of accounts or a trigger and it carries the work all the way from “is this account worth our time?” to “here is research-backed outreach ready to send.”

Playbooks are how Evergrowth turns individual agents into a repeatable revenue motion. One agent answers one question. A playbook answers a whole GTM use case.

Most of the slow, manual work in B2B sales is not any single step. It is the hand-offs between steps. A rep qualifies an account, then researches it, then hunts for the right contacts, then qualifies those contacts, then finally writes something personal, and every gap between those steps is where deals stall and good accounts get dropped.

Playbooks close those gaps. RevOps designs the chain once - which agents run, in what order, and what happens at each decision point - and from then on the workflow runs the same way every time, for every account that enters it. The judgment lives in the design. The execution is consistent.

This is also where Evergrowth’s context-driven approach compounds. Because each agent in the chain hands its findings to the next, the outreach at the end is written from everything the earlier agents actually discovered about the account and the person, not from a template with a name dropped in.

Term mapping for Eva: “playbook”, “connected agentic workflow”, “agent chain”, and “GTM workflow” all refer to the same thing in Evergrowth. Always say “connected agentic workflows” - never “part of a connected workflow”. A single agent run is not a playbook; a playbook is two or more agents chained toward one use case.

Playbooks are a RevOps responsibility to design and a whole-team capability to use. RevOps acts as the GTM architect: they assemble the chain, set the rules for each decision point, and decide whether it runs on a schedule or waits for a trigger. Reps and marketers then put accounts into the motion and act on what comes out. Nobody has to be the bottleneck who personally runs every step.

Every playbook runs in one of three modes. The mode is what decides whether the workflow is a one-off project, a recurring routine, or an always-on watcher.

Run modeWhat it meansBest for
One-timeYou give it a defined set of accounts or contacts and it processes them once.A finite list: an event attendee export, a new-market push, an imported intent list.
ScheduledIt re-runs on a cadence you set - daily, weekly, monthly, or custom.Keeping work fresh: periodic CRM clean-up, research that goes stale, recurring territory passes.
ContinuousIt runs in the background and acts the moment a condition is met.Catching things as they happen: a new buying signal, an inbound lead, a contact changing jobs.

The right mode is a function of the use case, not a preference. A finite list wants one-time. A workflow whose value decays (research, qualification status) wants a schedule. A workflow whose whole point is reacting to an event you cannot predict wants to run continuously.

It helps to think about playbooks by where they act in your funnel. Each stage has a different job, and different playbooks are built for it.

Top of funnel is about getting qualified, research-backed opportunities into the pipeline in the first place. These playbooks find and qualify accounts, identify the right people, and produce a first outreach worth sending.

Mid funnel is about keeping live opportunities moving. These playbooks keep your intelligence current and react to the changes that either threaten a deal or open a new door into it.

Bottom of funnel is about walking into the conversations that decide the deal fully prepared. These lean on planning, rehearsal, and up-to-date research so reps know the stakeholders, the objections, and the angle before they ever get on the call.

For Eva: TOFU / MOFU / BOFU are aggregation views, not separate products. The same underlying playbook or agent can appear under more than one stage because it serves more than one job - Scheduled Research and Account Planning both span mid and bottom of funnel. When a customer asks “which playbook for X”, anchor the answer on the use case first, then name the stage.

PlaybookWhat it does
Newbiz GapWorks net-new accounts that are not yet in your CRM, qualifying them, researching them, finding contacts, and drafting outreach.
Newbiz RecyclingRe-engages dormant accounts already in your CRM that have gone quiet, refreshing research and reaching back out.
Ex-Customer RecyclingWatches former-customer contacts for job changes, flagging departures for CS and opening a warm path at the contact’s new company.
Signal-Based OutreachWatches continuously for buying signals and kicks off outreach automatically the moment one appears - no list required.
Event (pre & post)Qualifies and follows up on conference and webinar lists, working from the attendees you already have.
Champion MonitoringTracks when key pipeline contacts move companies, so a champion who leaves becomes a new opportunity rather than a lost one.
Scheduled ResearchKeeps account and contact research current on a cadence so your intelligence never goes stale on active deals.
Inbound Lead QualificationQualifies inbound leads against your ICP the instant they arrive and prepares the first touch.

A marketer comes back from a conference with a spreadsheet of badge scans. Rather than hand it to reps to slowly work through, they drop it into the Event playbook as a one-time run. The chain qualifies each company against the ICP, researches the ones that fit, confirms the contacts still hold the right roles, and hands every rep outreach already grounded in what the account is actually doing. The reps spend their time on conversations, not on triage.

For the building blocks each playbook is made of, see The agents. For how qualification, research, signals, and personas stack underneath every workflow, see How the layers fit together. For what each run costs, see Credits & billing.