5D Library › Conversational Closing Automata, Objection-Handling Information Gain and High-Ticket Sales Kinetics · 中文

Chapter 6. Trust-Threshold Stopping Rules and Planning the Human Desk

Keep talking, ask for payment, or hand over

After each turn the AI has three choices: keep talking; ask for payment; hand over to a human closer. This is an optimal stopping problem: when to stop gathering information and take an irreversible action. The AI’s belief is the probability pp that the buyer is ready. Each further turn, 10% of buyers leave, a buyer who is not ready is persuaded and becomes ready with probability 0.04, a ready buyer shows a buying signal with probability 0.6 and one not ready with only 0.15. Asked for payment, a ready buyer closes with probability 0.7 (RM 9,800); one not ready refuses, and half of those leave. A human hand-over has a cost per hand-over; the human closes a ready buyer with probability 0.9 and one not ready with 0.2.

Calculation

# Chapter 6. p11: When should the AI ask for payment, hand over to a human,
# or keep talking? The buyer is ready to buy or not; the AI holds the
# belief p = P(ready). Each further turn: 10% of buyers drift away, a
# buyer who is not ready becomes ready with probability 0.04 (questions
# and proof do their work), and the buyer shows a buying signal with
# probability 0.6 if ready, 0.15 if not. Asking for payment closes a ready
# buyer with probability 0.7 (RM 9,800); a buyer who is not ready refuses,
# and half of them leave. A human closer's time costs c per hand-over;
# the human closes a ready buyer with probability 0.9 and one not ready
# with 0.2. Value iteration on a grid of 401 beliefs.
V, N = 9800.0, 400
DRIFT, NURT, SIG, ASK, LEAVE, HUMAN = 0.10, 0.04, (0.6, 0.15), 0.7, 0.5, \
    (0.9, 0.2)

def interp(f, p):
    x = p * N
    i = min(int(x), N - 1)
    return f[i] + (f[i + 1] - f[i]) * (x - i)

def solve(cost):
    f = [0.0] * (N + 1)
    for _ in range(5000):
        g, act = [], []
        for i in range(N + 1):
            p = i / N
            q = p + (1 - p) * NURT
            cont = 0.0
            for sy, sn in ((SIG[0], SIG[1]), (1 - SIG[0], 1 - SIG[1])):
                pz = q * sy + (1 - q) * sn
                cont += pz * interp(f, q * sy / pz)
            cont *= 1 - DRIFT
            stay = p * (1 - ASK) + (1 - p) * (1 - LEAVE)
            close = p * ASK * V + stay * interp(f, p * (1 - ASK) / stay)
            esc = (p * HUMAN[0] + (1 - p) * HUMAN[1]) * V - cost
            best = max((cont, "keep talking"), (close, "ask for payment"),
                       (esc, "hand to a human"))
            g.append(best[0])
            act.append(best[1])
        done = max(abs(a - b) for a, b in zip(f, g)) < 1e-7
        f = g
        if done:
            break
    return f, act

if __name__ == "__main__":
    for cost in (250.0, 500.0):
        f, act = solve(cost)
        print("   human time worth RM {:.0f} a hand-over:".format(cost))
        prev = None
        for i in range(N + 1):
            if act[i] != prev:
                print("     from P(ready) {:.3f}: {}".format(i / N, act[i]))
                prev = act[i]
        if cost == 250.0:
            for p in (0.1, 0.3, 0.6):
                i = int(p * N)
                print("     value at {:.1f}: RM {:,.0f} (ask now RM {:,.0f})"
                      .format(p, f[i], p * ASK * V))
#    human time worth RM 250 a hand-over:
#      from P(ready) 0.000: keep talking
#      from P(ready) 0.138: hand to a human
#      from P(ready) 0.500: ask for payment
#      value at 0.1: RM 2,454 (ask now RM 686)
#      value at 0.3: RM 3,768 (ask now RM 2,058)
#      value at 0.6: RM 6,001 (ask now RM 4,116)
#    human time worth RM 500 a hand-over:
#      from P(ready) 0.000: keep talking
#      from P(ready) 0.295: hand to a human
#      from P(ready) 0.415: ask for payment

With human time costed at RM 250 a hand-over, the best rule has three zones: below a belief of 0.138, keep talking; between 0.138 and 0.500, hand over to a human; above 0.500, the AI asks for payment itself. This is the trust-threshold stopping rule: the AI should not ask for money while its belief is low, nor hand to a human the sales it is sure of. At a belief of 0.3 the conversation is worth RM 3,768 under the best rule, while asking for payment at once is worth only RM 2,058. When human time costs more (RM 500 a hand-over), the hand-over zone shrinks to between 0.295 and 0.415.

Value of the conversation against the belief, and the three best actions (human hand-over at RM 250, p11).

Note

The thresholds depend on the shadow price of human time. The tighter the staffing, the narrower the hand-over zone; the Gittins indices and uplift ranking of Chapter 5 decide which conversations within the zone go first.

How many human closers behind the AI

Hand-overs arrive at random, and a human call lasts 25 minutes on average. A buyer who has to wait may drift off: still there after ww minutes with probability e−w/10e^{-w/10}. The Erlang C formula gives the probability of having to wait (Gans et al. 2003).

Calculation

# Chapter 6. p12: How many human closers stand behind the AI? Hand-overs
# arrive at random (Poisson) at lam per hour; a human call lasts 25
# minutes on average (exponential). Erlang C gives the chance a buyer
# must wait and P(wait > t) = Pw * exp(-(c mu - lam) t). A waiting buyer
# is still there after w minutes with probability exp(-w / 10), so the
# share lost while waiting is Pw * (1/10) / (gap + 1/10), where
# gap = (c mu - lam) per minute.
from math import exp, factorial

MU = 60 / 25.0                       # calls per closer per hour

def erlang_c(lam, c):
    a = lam / MU
    if a >= c:
        return 1.0
    s = sum(a ** k / factorial(k) for k in range(c))
    top = a ** c / factorial(c) * c / (c - a)
    return top / (s + top)

def row(lam, c):
    pw = erlang_c(lam, c)
    gap = (c * MU - lam) / 60          # per minute
    late = pw * exp(-gap * 5)
    lost = pw * (1 / 10) / (gap + 1 / 10)
    return pw, late, lost

if __name__ == "__main__":
    print("   hand-overs/h  closers  P(wait)  P(wait > 5 min)  lost waiting")
    for lam in (3.0, 6.0, 9.0):
        c0 = int(lam / MU) + 1
        for c in range(c0, c0 + 3):
            pw, late, lost = row(lam, c)
            print("   {:12.0f}  {:7d}  {:7.3f}  {:15.3f}  {:12.3f}".format(
                lam, c, pw, late, lost))
#    hand-overs/h  closers  P(wait)  P(wait > 5 min)  lost waiting
#               3        2    0.481            0.414         0.370
#               3        3    0.155            0.110         0.091
#               3        4    0.042            0.024         0.020
#               6        3    0.702            0.635         0.585
#               6        4    0.320            0.237         0.200
#               6        5    0.130            0.079         0.065
#               9        4    0.865            0.823         0.786
#               9        5    0.462            0.360         0.308
#               9        6    0.227            0.145         0.120

With 3 hand-overs an hour, 2 closers leave 0.481 of buyers waiting and lose 0.370 while they wait; 3 closers cut that to 0.155 and 0.091; 4 to 0.042 and 0.020. With 6 hand-overs an hour, 4 closers still lose 0.200 while waiting, and 5 cut it to 0.065. With 9 an hour, 5 closers lose 0.308 and 6 cut it to 0.120. A human desk staffed to the average load loses many of the most valuable buyers while they wait; it should have one or two closers more than the load.

Hypothesis · not established

The author’s hypothesis (not yet tested): in private-channel sales where AI and humans work together, teams that hand over by a belief threshold (rather than when the buyer asks or by warmth) bring in at least 20% more sales value per human call. What would show it wrong: in an experiment switching the hand-over rule at random by week, the sales value per human call differs by less than 20%.

Summary

The best stopping rule has three zones: keep talking when the belief is low, hand over in the middle, and let the AI ask for payment when it is high; the dearer human time is, the narrower the hand-over zone. The human desk must be staffed as a queue: staffed to the average load, losses while waiting can exceed four in ten.

References

Gans, Noah, Ger Koole, and Avishai Mandelbaum. 2003. “Telephone Call Centers: Tutorial, Review, and Research Prospects.” Manufacturing & Service Operations Management 5 (2): 79–141.

Fan, Y. (2026). Conversational Closing Automata, Objection-Handling Information Gain and High-Ticket Sales Kinetics: Diagnostic Question Trees, Trust-Threshold Stopping Rules and AI Closer Escalation Protocols [对话式成交自动机、异议消解信息增益与高客单销售动力学] (Version 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.24090221 · Licensed under Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0).