Skip to content
SmartReach AI
The learning engine

Every booked call teaches the next message.

We measure the signal that triggered the outreach, the hook, subject line, tone, channel and send window, then the reply, the objection and the booking. The engine keeps what converts and drops what does not.

01What we measure

Four data layers, joined on one outcome.

Every message we send carries the signal that triggered it and the decisions that shaped it. When a call gets booked, we know exactly which combination produced it.

01Signal data
Follows, event attendance, departures, new hires, funding and headcount movement, tech stack change, all timestamped.
02Campaign metadata
Hook, subject line, word count, tone, personalization tokens, sequence step, channel, send hour and day.
03Response data
Opens, replies, reply sentiment, objection type, and the exact language that moved a thread forward.
04Outcome data
Slots offered, calls booked, attendance, no shows, and time from first signal to booked conversation.
02The learning loop

The engine only keeps what books calls.

  1. Step 1

    Detect

    A signal fires on an account we are watching. It is stored with its source and timestamp.

  2. Step 2

    Decide

    The engine selects the hook, subject shape, tone, channel and send window with the best history for that signal and role.

  3. Step 3

    Send

    The message goes out on the channel that account is most likely to answer on.

  4. Step 4

    Measure

    Every open, reply, objection and booking is written back against the exact decisions that produced it.

  5. Step 5

    Learn

    Weights update. Weak signals and losing phrasing lose volume, winning combinations get more of it.

03Signal performance

Not every signal is worth a message.

We rank signals by lift against our own baseline outreach. Relationship signals, a prospect following a target exec or a company page, outperform demographic filters by a wide margin.

Top signals

Lift vs baseline

  1. People followers

    Prospect follows a target exec or peer

    3.1x48 booked

  2. Company followers

    New follows on the client company page

    2.6x39 booked

  3. Event attendance

    Registered or attending a tracked event

    2.2x31 booked

  4. Departure

    Key contact left, budget owner changes

    1.9x24 booked

  5. New hire

    Fresh VP or director in an owning role

    1.7x21 booked

  6. Growth

    Headcount or funding expansion signal

    1.5x18 booked

Linguistic lift

Soft vs hard ask

2.4x

A soft interest ask out performs a hard link ask.

First touch replies default to soft interest phrasing across every segment.

Attrition alert

Departure window

12 accounts flagged where a finance owner departed and a replacement has not been named.

Outreach is prioritized inside fourteen days of the departure signal, while the budget is still unowned.

04Message decisions

The hook, the subject, the shape. All decided by data.

Top messaging hooks

Reply and booking rate

  1. “Congrats on the new role”

    New hire or promotion opener

    34% reply9.1% booked

  2. “Saw you at {event}”

    Event attendance reference

    29% reply7.8% booked

  3. “Noticed you follow {exec}”

    People follower warm hook

    27% reply7.2% booked

  4. “Since {departure} left”

    Budget owner change angle

    24% reply6.4% booked

  5. “Team looks like it is growing”

    Headcount or funding growth

    21% reply5.3% booked

  6. “Quick idea for {company}”

    Generic value opener, our baseline

    12% reply2.1% booked

Signal based hooks convert roughly three times better than a generic value opener.

Subject length

Shorter wins
  • 1 to 3 words28%
  • 4 to 5 words25%
  • 6 to 7 words17%
  • 8+ words10%

Three words or fewer replies at roughly three times the rate of an eight word subject. The lighter bar is booked calls.

Structure tests

With vs without

  • Personalized token

    {firstName} or {company} in the subject

    2.0x
    With28%Without14%

  • Ends in a question

    Question marks invite a reply

    1.6x
    With26%Without16%

  • All lowercase

    Reads peer to peer, not broadcast

    1.4x
    With24%Without17%

  • No sender company name

    Avoids looking like a pitch

    1.9x
    With23%Without12%

Top subject lines

Open, reply, book rate

  1. 1“quick question, {firstName}”

    3w · 26 charsCasualPersonalizedQuestion
    68% open31% reply8.4% booked

  2. 2“congrats on the new role”

    5w · 24 charsWarmGenericStatement
    64% open29% reply7.9% booked

  3. 3“{company} + follower overlap?”

    3w · 28 charsDirectPersonalizedQuestion
    61% open26% reply7.1% booked

  4. 4“saw you at {event}”

    4w · 19 charsCasualPersonalizedStatement
    59% open24% reply6.6% booked

  5. 5“worth a look?”

    3w · 13 charsDirectGenericQuestion
    54% open19% reply4.8% booked

  6. 6“Introducing our outbound intelligence platform”

    5w · 45 charsFormalGenericStatement
    38% open9% reply1.9% booked
05Channel and timing

Email or LinkedIn, and exactly when.

Channel performance

Email vs LinkedIn

Email

1,842 sent, 72% of volume

42.4%Reply rate7.7%Book rate

781 replies, 142 booked

LinkedIn

716 sent, 28% of volume

44.6%Reply rate7.4%Book rate

319 replies, 53 booked

LinkedIn replies slightly higher, email drives far more booked volume. So the decision is per prospect, not per campaign.

Timing

Send, book, attend
Best send time
01:00
33% booking rate
Best booking day
Monday
28% booking rate
Highest no show day
Tuesday
55% no show rate
Best reminder timing
1h before
58% attendance rate

Tone mix

Reply rate by tone
  • Casual28%41% of volume
  • Warm26%22% of volume
  • Direct23%27% of volume
  • Formal9%10% of volume

Formal phrasing under performs by a wide margin, so its volume moves to casual and warm.

06Objection handling

A reply is not a booking. The engine learns the turn after it.

Best reply pattern

Resolution rate
  1. Acknowledge, reframe, dated slot47%

    Highest performing objection structure, our default template.

  2. Direct rebuttal57%

    Only when the objection is factual. Contradicting outright underperforms.

  3. Peer benchmark, soft ask42%

    Strong second option when price is the blocker.

  4. Send asset only48%

    Never send an asset without a dated call to action attached.

Best tone

Resolution rate
  1. Empathetic55%

    Resolves objections best across every segment.

  2. Friendly40%

    Solid default, empathetic phrasing wins on price pushback.

  3. Formal38%

    Reserve for financial services and enterprise threads.

  4. Casual31%

    Loses credibility once an objection is raised.

Best next ask

Resolution rate
  1. Direct ask53%

    Pair the reframe with one specific dated slot.

  2. Social proof49%

    Naming a comparable customer lifts resolution.

  3. Urgency47%

    Urgency after an objection backfires, avoid it.

  4. Soft suggest40%

    Too passive after an objection, switch to a dated ask.

Resolution by segment

Objection to booking
  1. Professional services41%

    Anchor on billable hour recovery instead of seat cost.

  2. Healthcare38%

    Lead with privacy posture before any product claim.

  3. Financial services28%

    Address compliance up front, unaddressed risk stalls threads.

  4. SaaS and software25%

    Technical proof points resolve objections fastest.

  5. Manufacturing18%

    Longer cycles, expect two objection turns before a booking.

Use the intelligence with your team, or let ours run the outreach.

Either way, the same engine decides the signal, the hook, the subject, the channel and the timing, and every outcome is written back into it.

Book a Strategy Call