Follow-up is an implementation intention
Abstract. Most positive replies arrive after the touch most reps never send. This review connects the behavioral science of implementation intentions to sequence design: why pre-deciding when and how follow-up happens beats relying on daily discipline, and what changes when the sequence executes itself and the human failure mode flips from quitting early to not reviewing at all.
The question
The persistence gap is one of the oldest findings in outbound: across published vendor datasets, positive replies concentrate in mid-sequence touches, while most reps stop after one or two attempts. The gap has survived a decade of training courses telling reps to follow up more. If exhortation worked, it would have worked by now. The interesting question is why it fails, and what works instead.
The evidence
Gollwitzer (1999) distinguished goal intentions (“I will follow up more”) from implementation intentions (“when X happens, I will do Y at time Z”). Across hundreds of studies meta-analyzed in Gollwitzer and Sheeran (2006), implementation intentions produced medium-to-large improvements in follow-through, in domains from health behavior to studying. The mechanism is delegation of control to the situation: the decision is made once, in advance, and the trigger executes it, so the moment’s mood and competing demands lose their vote.
A sales sequence is an implementation intention in industrial form. The decision to send touch five was made on design day, rationally, with the evidence in view. The rep who quits at touch two is not lazy; they are making a fresh decision under discouragement, which is exactly the condition implementation intentions exist to remove.
The mechanism
Designed once, the sequence runs on its trigger structure: day offsets, channel switches, conditional stops. Modern orchestration completes the logic by removing the human from execution entirely. The system never gets discouraged at touch two.
But automation moves the failure, it does not remove it. Parasuraman and Riley (1997) documented the drift: when a system runs itself, human attention decays. The sequence-era failure mode is not quitting early; it is the unread reply, the unreviewed exception, the sequence still running on an account that answered. And reply latency is expensive: the lead-response literature (Oldroyd et al. 2011) shows engagement decays within hours.
Implications for practice
Design the sequence as the commitment device it is: six to ten touches for a tier-1 account, spacing that widens over time, every touch adding something new, a breakup touch that closes with dignity. Then let the system execute it, and move the human discipline to the two places it still matters: the daily reply sweep with a response-time standard measured in hours, and the weekly readout where the weakest touch gets redesigned.
In the curriculum this paper underpins the motion design units of PA 110 and the operate loop of PA 120. The maturity ladder restates it: at Manual, follow-up is willpower; at Orchestrated, it is architecture.