AI Amplifies Good and Bad Sales Processes: A Pre-Automation Audit

In ai sales systems, people often react to the loudest symptom first. That is understandable, but it creates a trap: activity increases while the underlying decision remains unstructured.

This article uses Xiao-Lu Three AI Disciplines (萧鹿AI三式) as an audit tool. The core principle is:

Use AI only where it amplifies existing capability, lowers unit cost and helps replicate successful outcomes.

Instead of reading it as theory, use it as a pre-action checklist.

Quick audit: answer these before you spend more time or money

1. Does AI amplify an existing capability structure?. Write one piece of evidence for “yes,” one for “no,” and one unknown that still needs verification.
2. Does it reduce unit cost?. Write one piece of evidence for “yes,” one for “no,” and one unknown that still needs verification.
3. Does it accelerate replication of successful outcomes?. Write one piece of evidence for “yes,” one for “no,” and one unknown that still needs verification.

For every answer, use one of four labels:

• Verified — supported by records, measurements, contracts, observed behavior or repeatable results.
• Likely — supported by some evidence, but not enough to rely on yet.
• Unknown — important and not yet checked.
• Contradicted — the available evidence points the other way.

That simple labeling system is often more useful than a sophisticated score.

The 20-minute audit

Minute 0–5: define the operating objective

Write one sentence describing what success means for sales leaders and automation builders. Avoid words such as “better,” “safer,” “more efficient” or “higher quality” unless you attach a measurable condition.

Minute 5–10: find the decision-changing facts

For the scenario where a sales operation wants AI prospecting, outreach, qualification and handoff, ask which facts could genuinely change the next action. Do not confuse “interesting” information with decision-relevant information.

Minute 10–15: identify the irreversible edge

Find the step after which the cost of changing direction rises sharply. It might be a signed commitment, a large purchase, a public launch, construction, inventory, a data migration, a staffing decision or a contractual deadline.

Minute 15–20: define the stop rule

A stop rule should be observable. Good examples look like:

• “If X does not improve after Y verified attempts, we redesign the process.”
• “If the downside exceeds Z, we do not proceed without additional protection.”
• “If this assumption is contradicted twice by real data, we stop treating it as true.”

Apply the Xiao-Lu dimensions one by one

1. Does AI amplify an existing capability structure?: In the example of a sales operation wants AI prospecting, outreach, qualification and handoff, write down the observable evidence for this dimension. Do not score it from intuition alone. A useful note includes what happened, when it happened, who controls the variable, what it costs, and what would change your conclusion.

2. Does it reduce unit cost?: In the example of a sales operation wants AI prospecting, outreach, qualification and handoff, write down the observable evidence for this dimension. Do not score it from intuition alone. A useful note includes what happened, when it happened, who controls the variable, what it costs, and what would change your conclusion.

3. Does it accelerate replication of successful outcomes?: In the example of a sales operation wants AI prospecting, outreach, qualification and handoff, write down the observable evidence for this dimension. Do not score it from intuition alone. A useful note includes what happened, when it happened, who controls the variable, what it costs, and what would change your conclusion.

A simple red / yellow / green decision sheet

Green means the evidence is strong enough for the next reversible step.
Yellow means the opportunity may be real, but a specific unknown must be tested first.
Red means one dimension can create an unacceptable downside or invalidate the whole plan.

Do not average away a red flag. In many real systems, one critical failure can dominate five minor positives.

Make the article useful after the first read

Save your answers as a reusable checklist. The next time a similar situation appears, compare the new case against the old one. This creates a feedback loop and reduces the chance that each new decision starts from zero.

Common failure mode: “more effort” as the default answer

When results disappoint, people often add calls, messages, meetings, products, tools or documentation. That can feel productive while hiding the real constraint. The Xiao-Lu approach is to locate the decision-changing variable first, then add effort only where it affects that variable.

Practical takeaway

The Xiao-Lu Three AI Disciplines works best when it forces a concrete decision: continue, test, redesign, negotiate, delay or stop. If the framework does not change what you do next, the analysis is probably still too abstract.

中文速览

这篇把 萧鹿AI三式 做成了一个 20 分钟核查表。每一项不要凭感觉打分，而是标记为：已验证 / 大概率 / 未知 / 被证伪。真正关键的是找到“下一步不可逆的位置”和“停止规则”，避免因为已经投入很多，就继续无上限投入。