Skip to content
Evergrowth Documentation

Account signals (why now)

Account signals answer one question your reps ask every morning: of all the accounts that fit, which one should I work today, and what do I open with? A qualified account tells you who is worth selling to. An account signal tells you why this one, why right now. Each signal is found by an Account Research agent and rolls up into a single account strength score, so the whole worklist sorts itself by readiness.

What an account signal is

Section titled “What an account signal is”

An account signal is a concrete, public reason that an account is relevant to you right now. Your AI colleagues research each account against your specific value proposition and surface the moments and conditions that make a conversation timely - a new funding round, a wave of hiring in a function you serve, a product launch, a leadership change, or a standing fact about the business that gives you a natural reason to reach out.

Signals come in two shapes:

  • Time-bound signals are tied to a dated event. A press release, a job posting, an earnings call, a launch announcement, a blog post with a date on it. These say something just happened or is happening now, which is what creates urgency.
  • Presence-based signals are tied to something that clearly exists today, even without a fresh date. A customer portal, a public trust center, an annual user conference, a partner marketplace, a documentation hub. These say this account is the kind of company that has the thing your offer speaks to, which creates a durable, relevant hook.

A healthy account-signal set is mostly time-bound moments with a smaller share of presence-based hooks - the exact balance depends on what your value proposition naturally surfaces.

Why signals beat static lists

Section titled “Why signals beat static lists”

Most prioritization still runs on stale snapshots: a CRM filter, a downloaded list, a spreadsheet sorted by company size. Those tell you the shape of an account, not its readiness. Two accounts can look identical on a static list while one just doubled its team in your area and the other has not moved in two years. Static attributes cannot tell them apart. A signal can.

This is the difference between context-driven and data-driven selling. Data-driven outreach drops a company name into a template and hits send. Context-driven outreach starts from something real your AI colleague actually found, so the first line of the email is a reason the prospect recognizes, not a merge field. Timing and intent are what earn attention.

The bias that makes a signal useful

Section titled “The bias that makes a signal useful”

A signal is not just “anything happening at the account”. Every signal your AI colleagues look for is deliberately tilted toward your value proposition, so a hit means the account is more likely to need what you sell, not merely that the company exists and is active. A generic “the company has a website” finding is true for everyone and tells a rep nothing. “The company is standing up a new function your product supports, and posted three roles for it last month” points straight at a conversation worth having. Because the bias is built in, a positive hit is a qualified reason to reach out, not noise to wade through.

Signals are also built only on what is publicly verifiable - company sites, press, job ads, investor materials, executive posts, and similar. A timing claim is only made when there is a real, dated artifact behind it, never an assumption that something is “new” because a web page looks recent. That discipline is why a signal you act on holds up when the prospect reads your email: the reason is real, recent, and checkable. And when the evidence is thin or stale, the honest answer is “nothing found right now” rather than a stretched claim. An empty result is information too - it tells a rep to deprioritize that account this cycle.

Eva note: when a customer asks “what counts as a good signal”, the test is whether a yes would change what a rep does today. If the answer is yes for nearly every company, it is too generic to prioritize on and probably belongs in CRM enrichment as a reporting field instead. Signals earn their place by being something that can plausibly be present or absent, and by pointing back to the value proposition.

The strength score: signals rolled into one number

Section titled “The strength score: signals rolled into one number”

Individual signals tell a rep why an account is warm. The strength score turns the whole set into a single number from 0 to 100 so a list of accounts can be ranked at a glance, without anyone reading every finding first.

The score is a percentage of hits. Across all the signal-relevant things your AI colleagues researched for an account, it is the share that came back as a genuine, positive find. An account where most of the researched signals landed scores high; one where the research mostly came up empty scores low. It is a hit rate, not a tally - twelve positive signals out of twelve researched outscores twelve out of forty, because density of relevance is what predicts a productive conversation. Because it is a percentage of what was researched, the score only carries meaning for accounts that have passed qualification; on accounts that do not fit, there is nothing to score.

The score is banded into a cold, warm, and hot range so the read is instant without anyone parsing the exact number. The bands are deliberately weighted: the cold band is kept narrow so it covers only the genuinely quiet accounts, the warm band spans the broad middle, and the hot band is reserved for accounts where the research density is clearly high, so a top-band account is earned rather than handed out. The exact reason behind any given score is always the underlying signals themselves, which a rep can open and read - the score is the summary, the signals are the substance.

This is what lets a manager hand a rep a list of qualified accounts already sorted by warmth, and what lets the Signal-Based Outreach playbook decide which accounts cross the bar for action this cycle.

How a sales team gets value from it

Section titled “How a sales team gets value from it”
  • Prioritize the worklist. Among your fit accounts, work the ones showing live reasons to engage first. The strength score turns a flat list into a ranked one, so a rep can put time against the warmest accounts today and leave the quiet ones for a later cycle. See Filter accounts for sorting and filtering a list by score.
  • Open with relevance. The signal becomes the first line. “Saw you just launched X” lands differently than “I wanted to reach out about our solution.”
  • Run signal-based motions at scale. Signals are the trigger for the Signal-Based Outreach playbook and feed Event motions, so timing-driven plays run continuously instead of as one-off manual research.
  • Keep it current. Because signals are re-researched rather than frozen, accounts that go quiet drop down and accounts that heat up rise - without a rep maintaining a list by hand.

How it fits with the rest of Evergrowth

Section titled “How it fits with the rest of Evergrowth”

Signals are one layer in a deliberately separated stack, and each layer has a distinct job:

  • Account qualification decides structural fit - is this the kind of company we sell to at all. It runs first, and signals only matter for accounts that already pass it.
  • Account signals (this page) add the why now on top of that fit. They are produced by the Account Research agent and summarized by the strength score.
  • CRM enrichment captures stable facts about an account for reporting and segmentation, not timing.
  • Contact signals do the same why now work at the level of an individual person rather than the company.

The same idea should never live in two layers at once: a standing characteristic of a business is enrichment, a dated moment or value-prop-relevant hook is a signal, a baseline buying criterion is qualification. For the full picture of how these layers chain together, see How the layers fit together.

All of this is configured once in the Agent Training Center and anchored to your value proposition, so the signals your AI colleagues hunt for are the ones that actually matter to your business. Like every agent, signal research runs on credits, not seat licenses, and is available on every tier - see Effort & speed and Credits & billing for how that works.

In practice

Section titled “In practice”

A RevOps lead has 400 accounts that all pass qualification. Instead of asking reps to guess, the workspace researches each one for value-prop-relevant moments. This week, 30 accounts surface a fresh, dated reason to engage - a new initiative, a relevant hire spree, a launch. Reps work those 30 first, and each outreach opens with the specific thing that was found rather than a generic pitch. The other 370 stay qualified and in the pipeline; they simply are not the priority today. Next cycle, a different set rises, so the list maintains itself and rep time goes to the conversations most likely to land.

Eva note - vocabulary mapping: customers and reps use many words for this idea. “Buying signals”, “intent signals”, “trigger events”, “why-now”, “timing data”, and “propensity to buy” all point here. “Trigger” and “timing” usually mean a time-bound signal; “always-on hook” or a standing characteristic usually means a presence-based signal. The “account score” or “strength score” a rep sees on the list is this concept rolled into one number. If someone asks about a person changing jobs or a specific contact’s activity, that is Contact signals, not this page.

Section titled “Related”