A fixed-day cadence sends your Day 3 email whether or not anything happened on Day 3. That is the actual bug, not your subject line. The fix is to stop scheduling touches by the calendar and start firing them off real buying signals, so the message lands while the account is genuinely in-market instead of three days after they moved on. Teams that rebuild even two or three steps of an existing Outreach, Salesloft, or Apollo sequence around visitor-behavior triggers consistently see faster replies, because the trigger and the outreach happen in the same window instead of on opposite sides of an arbitrary wait period.

Why calendar-based cadences are structurally broken

Most sequence tools were built around a simple assumption: you don't know when a prospect is paying attention, so you space touches out and hope one lands. Day 1 email, Day 3 call, Day 7 LinkedIn message, Day 12 breakup email. It works well enough as a floor. It fails as a strategy, because the day count has nothing to do with the account's actual behavior.

Across the implementations we've been part of, the pattern repeats: a website visitor identification tool identifies a visitor, a human-configured workflow decides whether to enroll them, and then the sequence tool takes over as a downstream, disconnected system that only knows what it was told at enrollment. It has no idea the same contact came back to the pricing page four hours later. It just keeps counting days.

What "trigger-based" actually means (it isn't the same as lead scoring)

Lead scoring assigns a number to an account. Triggering changes what happens next, in real time, because something specific occurred. They're related, but conflating them is why a lot of "signal-based outbound" rollouts stall at a dashboard nobody acts on.

A trigger is an event you've decided is worth interrupting the plan for. In a visitor-identification context, the events worth wiring up usually fall into a short list:

• A known account revisits a high-intent page (pricing, comparison, security/trust) within 48 hours of the last touch
• A new, previously unseen title from a target account shows up on the site (a buying-committee signal)
• An account that's been dormant for 30+ days suddenly returns
• A single session includes 3+ pageviews on product or use-case pages, an intent-depth signal most cadences never check for

None of these require replacing your sequence tool. They require deciding, in advance, which events should override the calendar.

The signal-to-action map

Once you have events worth watching, the next step is deciding how hard each one should push. We map it in tiers rather than a single score, because a repeat pricing-page visit and a first-time anonymous session shouldn't trigger the same response:

• 🔴 Repeat visit to a high-intent page within 48 hours - skip the next queued email step, trigger an immediate call or LinkedIn touch instead
• 🟠 New titled contact identified at a known target account - branch into a multithreading sequence rather than the standard single-contact cadence
• 🟡 Net-new ICP-matched company, single session - start the standard day-based cadence; there's not yet enough signal to justify acceleration
• 🟢 Return visit after 30+ days dormant - route to a re-engagement sequence, not the cadence they originally fell out of
• ⚪ Single anonymous session, no ICP match - hold for nurture only; don't spend a rep's time on it yet

This is the piece most teams skip. They wire up the identification and the alerting, then send every signal into the same generic "hot lead" sequence. A tiered response is what actually protects rep time.

How to retrofit a sequence you already run

You don't need a new tool to do this. You need to change what triggers a step in the one you have.

1. Audit your current cadence for real time-dependencies. Some steps genuinely need a gap ("give it 3 days before following up on a reply"). Most exist only because the tool needed a number. Mark the difference.
2. Pick two or three signal types to wire up first. Don't try to trigger off everything in week one. Repeat pricing visits and new buying-committee contacts are usually the highest-leverage place to start, based on what we outlined in the pricing page trigger framework.
3. Route the signal into your sequence tool's enrollment or step-acceleration API instead of hand-building new sequence logic. Knock2 pushes identified-visitor events over webhook or directly into Apollo, Outreach, Salesloft, Instantly, Smartlead, Lemlist, HeyReach, Attio, and HubSpot sequences, which is what makes this a configuration change instead of an engineering project.
4. Keep a calendar-based cadence as the floor. Not every account will fire a trigger. The base cadence still needs to run for everyone else, or you've just built a more complicated way to ignore most of your pipeline.

Once triggers are wired in, the operational question shifts from "what should I send today" to "what fired overnight that changes today's plan," which is a different rep workflow than most cadences are built for. We cover how that daily triage should actually run in the SDR daily schedule for working real-time alerts.

What we see across real implementations

The clearest evidence for this isn't a benchmark, it's how teams actually talk about the problem before they fix it. In one conversation about testing outbound with a growth agency, a B2B founder put it plainly: he wanted "something to be done about landing at the right times with the right folks, the right cadence... with a bit more managed oversight to it" rather than a blanket send schedule. The agency's answer was itself a trigger framework: pull the accounts engaging with competitor content, showing new hires, or spiking on web traffic, and "when they come into that window, email them, LinkedIn DM them." Nobody on that call called it "event-triggered outbound." They described it anyway, because it's the obvious fix once you can see the window.

We see the opposite mistake just as often: teams that jump straight to full automation before they have the volume to justify it. In one onboarding call with a mid-market SaaS team running roughly six qualified leads per rep per week, the guidance was the reverse of what you'd expect: keep a manual step in the sequence rather than automating the whole thing, because at that volume a timely, human touch beats a scripted one. Trigger-based doesn't mean fully automated. It means the right thing happens at the right moment, and sometimes the right thing is a human who noticed in time.

Common mistakes when switching to trigger-based cadences

• Triggering on everything. If every session fires an alert, reps stop trusting alerts, and you're back to working the calendar. Tier your signals before you wire up more than two or three.
• No cooldown logic. A repeat visitor within an hour of the last one isn't a new trigger, it's the same session. Without a cooldown window you'll re-enroll the same account into overlapping sequences, which is the exact failure mode we broke down in how to stop your GTM stack from hitting the same visitor twice.
• Ignoring what level of identification you actually have. A trigger built around "a named contact revisited pricing" only works if you're resolving to the person, not just the account. Knock2's own identification rates run 93%* at the account level and 62%* at the person level for US traffic, measured against engaged sessions. If your program sits toward the lower end of person-level resolution, build your highest-priority triggers around account-level events (a known company returning) rather than assuming you'll always have a name attached.
• Killing the base cadence entirely. Trigger-based accelerates the accounts showing signal. It doesn't replace working the rest of the list, covered in the BDR playbook for website visitors.

*Identification rates measured against engaged sessions. Results may vary by traffic profile, geography, and industry.

FAQ

What's the difference between lead scoring and event-triggered outbound?

Scoring assigns a number to an account over time. Triggering changes what happens next, immediately, because a specific event occurred. A high score can sit in a dashboard indefinitely. A trigger is designed to be acted on the same day.

How many trigger types should we start with?

Two or three. Repeat visits to a high-intent page and new titled contacts at a known account are usually the highest-leverage starting point. Adding more before reps trust the first ones just creates alert fatigue.

Will this replace our existing sequence tool?

No. It changes what enrolls or accelerates a contact inside the tool you already run. The retrofit is a configuration and routing change, not a platform migration.

Does this work if we only have account-level identification, not person-level?

Yes, build your triggers around account-level events (a known company returning, a new session from an existing target account) instead of person-level events until your person-level resolution improves.

How fast does a trigger need to fire to matter?

Same-day at minimum. Signal-triggered outreach loses most of its advantage past 24 to 48 hours, which is the whole reason to move off a fixed day-count cadence in the first place.

Ready to route your own visitor signals into the sequences you already run? See how Knock2 triggers Apollo, Outreach, and Salesloft sequences off real-time visitor behavior, no new tool required.