企业AI 服务第一单怎么卖?先带员工干活,再谈系统开发

原文链接: https://x.com/lukepierceops/status/2115189554785452224 来源: x.com 发现于: 小互 · AI 解读站 · 2026/10/29 说明: 本文由 AIHot 自动聚合(原文即中文,未翻译)。正文取自小互 AI 解读


全文 Markdown · 会员

作者:Luke Pierce10-29 · 7分钟🎧 听播客版0:00 / --:--

视频/音频 0:00 / --:--

企业请人做 AI 培训,真正值得付费的成果,是员工学会用它处理自己的工作,并且在顾问离开后仍能继续使用。对提供企业 AI 服务的团队来说,这也可以成为第一笔合作:先带员工解决眼前的问题,再从一起做事的过程中,找到值得进一步开发的项目。

Boom Automations CEO Luke Pierce 分享的 AI Enablement,可以理解为围绕岗位真实工作的 AI 实操培训。它把教员工、了解业务和讨论后续开发连在一起:客户先获得能用起来的技能,服务商也在现场看清工作怎样运转,双方再决定哪些问题需要系统开发来解决。

先把培训做成一项值得付费的服务

这类培训的目标很具体:员工知道 AI 能帮自己做什么,会把它用到岗位任务里,并能在没有顾问指导时继续使用。形式可以是现场工作坊、分部门实操、录播课程、工具操作规程(SOP),或让部门围绕实际任务动手尝试的活动。

Luke 自述,他曾因提案里缺少培训层而输掉一次竞标,对手提供了培训。从那以后,他的每份提案都会加入这一层。客户期待的收获也因此超出了软件本身:外部服务结束后,团队仍能把学到的方法用在日常工作里。

因此,培训包要有独立的交付价值。即使客户最后不买定制开发,员工也应当学会处理一些实际任务,留下可以继续使用的操作方法。后续开发是进一步的合作选择,不能成为前一笔培训费是否花得值的唯一条件。

按 Luke 的观察,实操培训能让客户在第一次课就得到具体帮助,第一周便感受到这笔投入的用途。企业即使尚未决定是否开发定制系统,也可以先为团队解决眼前的问题付费。员工开始受益时,顾问也开始了解真实业务,第一笔合作便有了明确的交付理由。

这一入口也回应了原来的一项销售摩擦。按 Luke 的介绍,独立的业务流程诊断通常需要两到四周,客户先付数千美元买报告,却还不知道后续开发要花多少钱;即使诊断费可以抵扣开发费,对未知总价的抵扣也未必有吸引力。服务商容易陷入反复解释文档价值、不断压低诊断费的局面。先交付团队能用起来的技能,让第一笔钱的用途更容易理解。

Luke 的团队仍在做流程诊断,也在并行测试培训入口;已经能够顺利成交的诊断服务,可以继续保留。

让员工带着真实工作来上课

培训和业务了解能够同时发生,关键在于实操用的是员工自己的工作:每周一都要重新整理的报表、让人头疼的收件箱,或者在三种工具之间反复复制粘贴的流程。顾问教员工处理这些任务,也就有机会看见任务背后的实际步骤。

假设一位运营同事每周都要做一份业务周报:从客户管理系统导出跟进记录,从财务软件导出回款信息,在表格里匹配后,再把异常情况整理成给同事看的说明。

顾问带他做这项工作时,可以辅导他用现有 AI 工具整理文字或表格,并把可复用的操作步骤留下来。同一过程中,顾问也看见了数据来自哪里、哪些字段需要人工整理、在哪些工具间搬运,以及结果还要由谁检查。员工练习怎样完成任务,服务商则理解了这个任务为什么繁琐。

这些现场看到的步骤与卡点,正是业务调研要找的内容。把课上的实际操作记录下来,再请员工补充最想交给 AI 的任务,就能把零散抱怨变成有具体对象的需求。

实操还提供了另一类信息:谁愿意尝试新工具,谁有顾虑,哪些流程是员工真正愿意改变的。它帮助服务商更早了解使用意愿与潜在阻力。找到技术上可以改的地方之后,还要看改出来的工具能否接进团队的工作习惯。

从员工自己的工作开始

同一场培训,两类收获

带着周报、邮件或跨工具搬运的任务来实操

员工学会做事

练习怎样用 AI 处理手头任务 ↓

留下能继续使用的步骤与操作规程 不买后续开发,也有培训收获

服务商看清工作

观察输入、操作步骤和业务卡点 ↓

收集想交给 AI 的任务和使用意愿 获得有具体对象的需求材料

↓

整理机会地图

把现场记录变成可讨论的改善方向

现有工具可以改善

继续练习并完善操作方法 需要进一步评估开发

核对接口与数据条件,再讨论项目

进入开发合作后,继续安排每月培训,保留录播与操作规程。

培训交付能力,也提供业务线索;机会地图仍需整理,项目仍需评估。

从团队培训到系统开发的五步方法

1. 先卖一项范围明确的培训包

先为管理层安排启动工作坊,再按部门开展动手实操,按照训练好团队的价值定价。服务包要说清培训对象、要练的工作以及会留下什么,便于客户判断这项服务本身是否值得购买。范围明确,并不意味着所有客户必须采用同一个报价。

2. 每次课都收集员工最想交给 AI 的任务

在每场实操结束时,让团队列出:如果明天能把一部分工作交给 AI,最希望它接手什么?参与者要包括以后真正使用工具的人,不能只听企业负责人的想法。课上的操作记录和课后的任务清单,一起构成后续梳理需求的材料。

3. 把实操记录整理成机会地图

将各部门暴露的重复劳动、跨工具搬运和流程卡点汇总,整理成一份机会地图。它承担的是业务诊断成果的作用:让客户看见哪些地方值得改善,以及可能采取什么办法。

例如,周报文字整理可能通过员工掌握现有工具来改善;如果问题出在多个系统之间持续搬运数据,就需要进一步评估接口和数据条件,再判断是否值得定制开发。这样整理之后,培训中发现的任务才变成可讨论的项目选择。并非每个痛点都要卖成一个开发项目。

4. 在最后一次课上约好当面讲方案

Luke 的具体建议是:在最后一次培训时就约好后续汇报,不要会前先把报告单独发给客户,而是在现场或实时通话中一起过方案。把发现的工作卡点、改善建议和下一步选择放在同一场讨论里,客户可以就具体问题继续追问。

5. 把持续培训放进开发合作

提出系统开发方案时,将团队培训纳入合作内容:合作期间每月开展实时培训,并提供客户可以保留的录播课程和操作规程。此前一起处理工作的经历已经建立了信任,后续方案也能接回团队实际提出的需求。

开始开发后,培训继续服务于实际使用

前期培训让员工学会用 AI 处理工作,也让服务商了解需求。进入开发阶段后,同一批使用者仍需要跟着新系统学习和调整工作方式。因此,培训要贯穿合作过程,和系统交付一起安排。

每月的实时培训可以用来讨论使用中的问题;录播与操作规程则留下可反复查阅的做法,帮助团队在顾问不在场时继续完成工作。目标是让能力留在客户手里,让员工能使用已交付的系统,也能继续在自己的岗位上用 AI。

这套方法把第一笔合作做成了具体的工作帮助:员工带着任务来,带着方法回去;服务商从实操中收集业务线索,再整理为可以讨论的改进方向。客户即使停在培训这一阶段,也有独立的收获;需要进一步开发时,双方则已经有了共同处理过的任务和更清楚的需求。

来源AI Audits Are Dead. Sell This Instead (Full Guide)Luke Pierce·查看主材料延伸Boom Automations 官网本站说明依据完整 X Article 及作者公司官网公开背景。 AI Enablement · 业务实操赋能

以实操培训为入口:让员工留下技能,让服务商看清业务

围绕员工自己的工作开展 AI 实操:让团队学会方法并持续使用,同时了解业务,再判断哪些改善需要系统开发。 先看培训的独立价值,再看同一课堂的双向收获与后续五步方法。

一、第一笔合作,先交付员工能用的能力

员工带着真实任务来,带着可复用的方法回去。即使企业暂时不决定做定制开发,培训也应当解决眼前的一些工作问题,留下操作规程,帮助员工在顾问离开后继续使用。

Boom Automations CEO Luke Pierce 自述,曾因提案缺少培训而输掉竞标。按他的观察,实操培训在第一次课便能提供具体帮助,客户第一周就能感受到投入的用途。

培训包本身 先说明谁学、练什么、留下什么 管理层启动、分部门实操,围绕真实岗位任务定价。后续是否开发,不决定这笔培训费是否值得。

回应购买顾虑 让第一笔钱的用途更明确 Luke 认为,2—4周、数千美元的独立诊断报告可能面临总开发价未知的顾虑;已能顺利成交的诊断服务仍可保留。

二、核心机制:同一实操现场的双向收益

让员工拿周报、收件箱和跨工具搬运数据来实操。顾问辅导员工时,能同时观察数据、步骤和检查环节,了解繁琐工作的实际原因。

Core Engine · 核心机制 现场实操的双向收益流动 同一项真实任务,让员工学会处理工作,也让服务商看清业务;后续开发仍需再判断。

共同起点:员工带着真实工作上课 每周一要整理的报表、CRM与财务软件间搬运的数据、需要处理的收件箱。

面向客户员工 员工学会可复用的方法

  • 练习用现有AI工具处理眼前任务
  • 留下可继续使用的操作规程(SOP)
  • 顾问不在时也能继续应用

面向服务商团队 服务商理解工作与需求

  • 看清数据来源、搬运和人工检查环节
  • 了解员工使用意愿与潜在顾虑
  • 记录实际卡点,收集最想交给AI的任务

把实操记录整理为机会地图,再区分:哪些问题可用现有工具改善,哪些需要评估接口、数据条件与定制开发。

实操能让服务商更早了解员工的使用意愿。技术上可以改善,还需要看新工具能否接进团队的工作习惯;观察到卡点并不等于已经完成开发评估。

三、落地路径:从团队培训到系统开发的五步递进

将培训期间的记录整理成可讨论的改善选择。周报文字整理可能通过现有工具改善;跨系统持续搬运数据,则要进一步评估接口和数据条件,才能决定是否值得开发。

01

先卖范围明确的培训包 为管理层开启动会,按部门展开实操,说清训练对象、练习任务与交付产物,按团队被赋能的价值独立定价。

02

每次课收集员工任务清单 实操结束时让真正使用工具的一线人员列出最想交给 AI 的工作,把零散抱怨转化为带具体场景的需求原料。

03

提炼实操记录为“机会地图”(核心枢纽) 汇总重复劳动与流程卡点,区分现有工具可改善的任务和需要进一步评估开发的任务,承担业务诊断成果作用。

04

最后一次课上约好后续汇报 不要会前单独发报告;在约定的现场或实时通话中,一起过卡点、建议与下一步选择。

05

把持续培训放进开发合同 开发合作期间安排每月实时培训,并留下录播与SOP,让前期形成的需求理解与信任接到实际使用。

四、全周期闭环:让能力真正留在客户手中

进入开发阶段后,同一批员工还要学习新系统、调整工作方式。培训与系统交付一起安排,让方法和能力留在客户手里。

第 1 周 Luke的观察:首课有具体帮助,首周感受到用途

机会地图 把现场记录整理成可讨论的改善选择

每月实操 开发合作期持续学习,留下录播与操作规程

培训一直服务于团队的实际使用 客户停在培训阶段,也有独立收获;继续开发时,双方已有共同处理过的任务和更清楚的需求。每月培训、录播和操作规程,帮助团队在顾问不在场时继续使用。

原文版权归原作者,这里只供对照阅读。 看原网页 ↗

AI Audits Are Dead. Sell This Instead (Full Guide)

I’ve been one of the loudest voices for audit-first selling since April 2025. It’s how we built automation and AI systems for a ton of companies. Now I’m starting to lead with something else, because it gets the same discovery and clients say yes to it faster. Bookmark this before your next sales call.

If you run an AI agency, you've probably felt this. A prospect loves your discovery call, you pitch the audit, and they go quiet. Or they ask why they'd pay a few thousand dollars for a report. Maybe you've already cut the audit price so far that it's basically free discovery.

Mapping the business before you build is still the right instinct, and a well-run audit still closes deals. Talk to anyone who has delivered for more than a handful of clients with high adoption and they'll tell you the same thing. The operators at the top all map first because it's the only way to get results. What I'm noticing is that the audit as a standalone paid offer is getting harder to sell, and AI enablement might be the better front door.

AI enablement means training a client's team to be AI native. Done right, it gets you every piece of information an audit would, while delivering something the client feels in week one.

Where the audit gets stuck

There's nothing wrong with the audit. Buyers are bought into it. For a while you had to educate every prospect on why you map the business before you build anything. You barely have to do that anymore, and a lot of owners now expect it.

We still lead with the audit, and I know dozens of operators who are closing audits right now. Enablement is something we're testing alongside it. If your audits are already closing, keep running them and keep this in your back pocket.

But if you're running into friction on the way to a signed audit, it probably looks like this.

A real audit takes two to four weeks. You spend a week or so interviewing the team and dissecting workflows, then another week building the deliverable. Done right, it's one of the most valuable things an agency can do.

The problem is what the owner is buying. They're paying a few thousand dollars for a report, and they still have no idea what the build will cost. You can credit the audit fee toward the build, but that barely helps. Credit toward an unknown number doesn't feel like a deal.

So you end up in one of two places. You discount the audit until it's basically a free discovery call, or you spend your sales calls defending why a document is worth paying for. Neither is a great place to start a relationship.

That's where enablement has the edge right now. Clients see it as a higher-value purchase than a report, and it's an easier yes to sign.

What I mean by enablement

AI enablement means making the client's team AI native. Everyone understands what AI is, knows how to use it in their own job, and actually uses it without you in the room. That can look like live workshops, a recorded video library, SOPs for their tools, or a department hackathon.

This is separate from a build. The goal is a team that gets better at their own jobs, whether or not you ever build them anything.

Buyers want this right now. I lost a competitive deal a few months ago because my proposal didn't have a training layer and the other agency's did. Since then, every proposal we send includes one. Owners want certainty that their spend goes beyond software, and they want a team that keeps getting better after you leave.

Enablement can double as the audit

This is what's extremely interesting we've noticed running enablement sessions. When you sit with a team and teach them to use AI in their work, they bring you their actual work. They show you the reports they rebuild every Monday, the inbox they hate, the three tools they copy and paste between. They tell you what they'd love to never do again.

That's everything you'd collect in an audit. Same interviews, same workflow mapping, same list of opportunities. You just gathered it while delivering something the client could feel in the first week.

You also get information an audit usually misses. You see who on the team is curious and who is resistant. You see which workflows people will actually adopt a tool for. Adoption is the biggest measure of whether a build succeeds, and you learn it before you've built anything.

The buying decision gets easier too. An audit asks the owner to pay for a diagnosis before they know what the fix will cost. Enablement gives them something useful in the first session, so saying yes feels low risk even if they aren't ready to commit to a build.

How I'd structure it

  1. Sell a defined enablement package as the entry offer. A kickoff workshop for leadership, then hands-on sessions by department. Price it on the value of a trained team.
  2. Build discovery into every session. Each one ends with the team listing the tasks they'd hand to AI tomorrow if they could. Make sure that includes the people who would use a build, beyond the owner.
  3. Turn what you collected into an opportunity map. It's the audit deliverable, produced as a side effect of the training.
  4. Walk through it live. Never send it ahead of time. Book that call while you're still in the last session.
  5. Propose the build with enablement baked in. Monthly live sessions with their team during the engagement, plus a recorded course and SOPs they keep. The client already trusts you with their people, so the build is the obvious next step.

The audit still happens. It just doesn't have to be the thing you ask them to pay for up front.

Where I think this is heading

Understanding the business before you build is still everything, and audits aren't going anywhere. I just think more owners are going to want their first dollar with you to make the team better right away, and they'll happily let you learn the business while you do it.

If your audits are closing, keep running them. If they keep stalling at the proposal stage, test leading with enablement on your next deal and see if the conversation changes.

If you want the full system

Everything in this article comes from what's working inside the AI Agency Inner Circle, including the full AI Enablement Offer Playbook: how we price it, pitch it and deliver it.

Inside, you license the exact acquisition and delivery systems our members are running right now. You get the offers, scripts, proposals and playbooks, and you can copy and run them in your own agency. You also get coaches who are ahead of you and a network of 100+ operators building the same kind of business.

If it doesn't work for you, you get every dollar back.

DM me "INNER CIRCLE" and I'll get you set up to book a call.

↑ 正文结束 ✓ 本篇全文免费开放

会员可复制、下载全文 Markdown,还能让你的 AI 订阅和提问。 订年卡 $59/年看权益对比 →

阅读原文 ↗

📮 新解读每晚进邮箱

读完觉得有用?留个邮箱,每晚一封当天的新解读,随时退订。