B2B Sales On Autopilot - Automating a Real Business Use Case (Robotic Process Automation)

Summary

这段视频讲解了如何使用机器人流程自动化(RPA)技术来获得更多B2B销售线索。作者分享了利用微软Power Automate工具自动化LinkedIn上的流程,从而查找发布招聘启事的公司,并锁定其CEO或CMO等高管进行营销邀约。这种基于公司招聘需求锁定高意向客户的方法,比传统的大规模冷邮件更有针对性和效率。作者还介绍了通过谷歌搜索运算符和在线教师平台获取RPA知识的技巧。总的来说,这个视频提供了一种实用的自动化方式来提高B2B销售的成功率。

Takeaways

• 🤖 视频介绍了使用机器人流程自动化（RPA）工具，如Microsoft Power Automate，来增加B2B销售的方法。
• 🔗 具体案例是在LinkedIn上自动化寻找潜在客户的流程，包括自动发送喜欢和直接消息。
• 🛠️ 强调了无代码工具的使用灵活性，提及了除了使用专有工具外，也可以通过Python等编程语言来构建自动化流程。
• 🎯 描述了一个详细的工作流程，包括筛选特定职位的LinkedIn职位发布、公司规模和关键决策者信息，以及如何通过自动化与他们互动。
• 💡 强调了在构建自动化时，了解所使用软件的内置功能的重要性，以减少对额外工具的依赖。
• 🚀 分享了通过优化和使用如Apollo这样的先进工具来提高效率和效果的个人经验。
• 🎓 推荐了使用Preply和Upwork等平台来寻找专业人士和导师，以提升构建自动化的技能。
• 🔍 强调了Google搜索操作符的使用，以便在寻找特定技能的导师时更精确地过滤搜索结果。
• 📊 介绍了自动化配置和输出文件的结构，包括如何设置目标、调整偏好设置和记录结果。
• 🧠 讨论了使自动化尽可能接近人类行为的策略，包括添加随机延迟和控制交互频率。
• 📈 强调了在B2B销售中，通过自动化定位高意向买家的重要性，以提高外展效果。

Q & A

RPA工具的全称是什么？

RPA工具的全称是机器人流程自动化（Robotic Process Automation）。

视频中提到的自动化流程主要是利用哪个工具实现的？

视频中提到的自动化流程主要是利用Microsoft Power Automate工具实现的。

如何提高LinkedIn自动化流程的效率？

提高效率的方法包括理解你正在使用的软件内置的功能，并尝试以不增加更多工具的方式使用特定的工作流程。

Apollo在自动化流程中起了什么作用？

Apollo作为一个领先源数据库，可以根据职位发布筛选潜在客户，例如筛选最近发布CMO职位的公司的CEO，从而简化了寻找目标客户的过程。

视频中建议使用什么方法来让自动化看起来更像人工操作？

建议使用随机延迟和等待，例如在发送一定数量的点赞后等待30分钟，并且在每次操作间添加随机延迟，以使自动化看起来更像人工操作。

如何在Preply平台上找到特定技能的导师？

可以通过在Preply平台搜索具体的课程或技能，如Python或RPA，并利用Google搜索操作符在Preply.com网站内搜索特定关键词来找到特定技能的导师。

视频中的自动化流程的目标是什么？

自动化流程的目标是查找LinkedIn上发布特定职位的公司，然后识别并接触这些公司的高层管理人员（如CMO或CEO），最终发送自动化的直接消息。

为什么视频中的自动化流程包括了对公司职位发布的筛选？

筛选公司职位发布是为了确保目标客户具有购买意向，比如最近发布了媒介购买者职位的公司更有可能对媒介购买服务感兴趣。

视频提到了如何使用Upwork找到帮助？

视频中提到，如果需要在特定领域提高技能，可以通过在Upwork发布工作需求来寻找专业人士的帮助。

视频中如何强调自动化与人性化操作之间的平衡？

视频强调了在自动化操作中加入随机延迟和等待时间，以及量化但质量并重的策略，旨在使自动化操作尽可能地模仿人类行为，同时维持效率。

Outlines

00:00

😀 B2B销售自动化流程概述

本段介绍了作者使用Microsoft Power Automate等RPA工具实现B2B销售自动化流程的方法。该流程包括搜索招聘信息、识别公司、寻找公司高管、给高管的最新动态点赞以预热关系,最后自动发送营销信息。作者还提到了后来发现Apollo等服务也提供了类似的功能,但当时并不知情,强调了要不断扩展知识以优化自动化流程。

05:01

😃 寻找优秀导师的两种方式

本段作者分享了他寻找合格导师以提升自动化技能的两种方式。一种是在Preply这个平台上按需求查找导师,作者还介绍了如何使用Google高级搜索操作获得更精准的结果。另一种方式是在Upwork上发布工作招聘岗位进行人才挖掘。这说明了作者愿意投资提升自身技能水平的决心。

10:01

😄 自动化工作流配置详解

本段详细介绍了自动化流程中的配置文件,包括指定工作机器人的名称、感兴趣的招聘信息链接、保存文件路径、工作时间段、是否启用点赞功能等参数设置。此外还说明了限制爬虫频率、点赞次数等参数,以及在输出文件中记录爬取和操作信息的情况。可见作者非常注重自动化流程的细节控制和数据记录。

15:04

😃 核心工作流程及实现步骤

本段作者开始详细阐述了自动化工作流程的具体实现步骤。包括配置初始文件、打开并抓取招聘信息页面、滚动直至页面底部、抓取公司信息写入文件、查看公司规模人数、识别并前往目标高管个人主页、滚动至最新动态进行点赞操作。作者还提到了使用第三方软件进行自动私信发送的部分。整个流程条理清晰,层层递进,逻辑严密。

20:05

🙂 总结与致谢

最后一段作者总结了整个自动化流程,表示如果读者感兴趣会在后续深入分享实现细节。他还特别感谢了一位专家对流程的优化贡献,并邀请读者对改进和优化提出建议。作者自己也总结了更注重工作效率快速迭代的理念。最后鼓励了B2B营销人员尽量锁定高需求用户群体进行营销的观点。

Keywords

💡B2B销售

B2B是"企业对企业"的英文缩写,指的是企业之间的商业交易行为。在视频中,作者分享了一种使用机器人流程自动化(RPA)工具来提高B2B销售效率的工作流程。该工作流程旨在自动化在LinkedIn上寻找并接触潜在客户的过程,为企业获取更多的B2B销售机会。

💡机器人流程自动化(RPA)

机器人流程自动化(RPA)是一种通过配置计算机软件或"机器人"来捕获和解释现有应用程序的操作,从而实现自动化流程的技术。在本视频中,作者使用微软的Power Automate这个RPA工具来建立自动化工作流,执行在LinkedIn上找到潜在客户、点赞他们的帖子并发送自动化私信的过程。

💡LinkedIn自动化

LinkedIn是一个职业社交网络平台,用于建立商业联系。在视频中,作者介绍了如何使用RPA工具在LinkedIn上自动执行寻找潜在客户、与之互动(如点赞帖子)并发送自动化私信等操作,从而提高了B2B销售的效率。该自动化流程可以根据特定条件(如公司规模、工作职位等)筛选潜在客户,并与之建立联系。

💡潜在客户发现

潜在客户发现是指识别和找到可能对公司的产品或服务感兴趣的人或企业。在视频中,作者分享了利用LinkedIn工作岗位发布来识别具有购买意向的潜在客户的方法。例如,如果一家公司发布了招聘媒体买手的职位,那么该公司就很可能对媒体购买服务有需求,成为潜在客户。

💡个人导师

个人导师是指为学习者或学徒提供专业指导的人。在视频中,作者提到他寻找了一位RPA专家作为个人导师,帮助他以更专业和强大的方式构建RPA自动化工作流程。个人导师可以传授专业知识和技能,提高工作效率并建立更加稳健和优化的系统。

💡谷歌搜索运算符

谷歌搜索运算符是一些特殊的关键词和符号,可以让谷歌搜索更加精准和高级。在视频中,作者提到使用site:运算符来在特定网站中搜索特定关键词,以找到相关的个人导师。了解和使用这些搜索运算符可以帮助我们更有效地查找所需的信息。

💡配置文件

配置文件是指用于设置自动化程序各种参数和选项的文件。在视频中展示的自动化工作流程中,作者使用了一个配置文件来设置机器人名称、工作位置URLs、目标职位、公司规模限制等参数,从而定制和控制自动化过程。配置文件让程序更加灵活和可定制。

💡LinkedIn个人资料URL

LinkedIn个人资料URL是指每个LinkedIn用户个人主页的网址。在自动化流程中,作者通过抓取公司的招聘信息,获取到公司高管(如CMO、COO等)的个人主页URL,然后程序自动导航到这些个人主页,对其最新动态执行点赞等操作。掌握个人资料URL的模式和结构对自动化互动很重要。

💡网络滑流

网络滑流是一种通过模拟人工操作来实现自动滚动网页的技术。在视频中,作者提到他们的自动化程序会自动滑动到招聘信息页面的底部,以确保抓取到所有相关的职位信息。这样可以避免由于分页等原因而错过部分数据。自动化任务时添加类似的"人性化"操作可以使程序行为看起来更自然。

💡自动化优化

自动化优化是指在建立自动化流程或系统后,持续改进和优化它们以提高效率和效果。视频作者提到,随着时间推移,他发现了一些方法来优化和简化原有的 LinkedIn自动化工作流程,如利用现有的专业工具替代部分手动步骤。优化自动化系统可以节省时间和资源,提高工作效率。

Highlights

The speaker describes an automated workflow using Microsoft Power Automate (an RPA tool) to find relevant job postings, identify the company and key decision makers (CMOs/CEOs), and then engage with them on LinkedIn by liking their recent posts before sending a direct message.

The speaker emphasizes the importance of understanding the full functionality of the software used for automation to optimize and streamline the workflow, reducing the need for additional tools.

The speaker discovered that the lead sourcing database Apollo has a feature to filter prospects based on job postings, which could have simplified the initial workflow.

The workflow aims to find job postings based on specific criteria (location, position, experience level) and then engage with the CMO or CEO of the company that posted the job by liking their recent LinkedIn posts before sending a direct message.

The speaker highlights the importance of making the automation look as human as possible by adding random delays and varying the frequency of actions to avoid detection.

To find help for building the automation, the speaker recommends using platforms like Preply to find tutors or experts in specific fields, or using Upwork to post job requests.

The speaker explains the configuration file, which includes parameters like job posting URLs, file paths, bot start/end times, and settings for the number of profiles to scrape, likes to send, and delays between actions.

The output file includes information about the companies, whether likes were sent, the number of employees (to filter out larger companies), and a summary of the actions taken.

The workflow involves scraping job postings, scrolling to the bottom of the page to capture all results, extracting company names, visiting company pages to scrape employee counts, identifying the CMO/CEO profile, and then liking their recent LinkedIn posts.

The speaker highlights the importance of understanding patterns in the user interface to automate actions effectively, such as navigating to the recent posts section of a LinkedIn profile.

The speaker used a separate software (wxy) to send direct messages after the liking process, as the Power Automate workflow focused on the initial engagement.

The speaker emphasizes the value of targeting prospects with high buying intent, such as those who have recently posted job openings related to the services being offered, rather than cold outreach.

The speaker encourages thinking outside the box to find customers with high buying intent and automating the outreach process accordingly, while also welcoming feedback and suggestions for improvement.

The speaker credits a collaborator (Michael) for helping build a robust and state-of-the-art automation workflow, as the speaker typically prioritizes speed over robustness for personal use.

The speaker has built many other use cases and automations using RPA and is willing to dive deeper into the technical details of this specific workflow if there is interest from the audience.

Transcripts

getting more B2B sales with robotic

process

Automation in today's video I'm going to

share with you a workflow that I have

been using for my clients and also for

myself for attaining more clients it's

an Automation in LinkedIn um is

something that I've built a while ago

and it has been performing very well and

I will probably dive deep into the all

the steps in the workflow to share with

you like my insight and my thought

process in this specific

case I build a workflow with Microsoft

power automate which is an RPA tool RPA

stands for robotic process Automation

and basically what you can do with RPA

tools is think about the

process

and then you just build the process in a

no code tool or not or you can also use

Python

obviously and eventually you can use

robots or code in order to repeat this

process using triggers or in any

interval that you would like

specifically in this case I'm using

Microsoft power automate again and this

I triggered this workflow on daily basis

whenever I wanted to get um more

leads one thing to men mention is that

this

workflow was built like a year ago and

maybe less not so crucial but since I've

built this workflow I have figured out

ways to make it even more efficient and

more streamlined and this relates to a

video that I uploaded a few days

ago and one of

the key takeaways or my lessons that I

shared with you guys whoever watched the

video a few days ago is the fact that

when building automations you should

always

understand the functionalities that are

built in to the software that you're

using and you should always strive to

understand if you can

use certain workflows in a way that you

don't use or include more tools and

specifically in these examples the

workflow goal was to find people who are

posting jobs on

LinkedIn after I found obviously

specific job so let's say I want to be a

look for positions of functional CMOS

remote position of functional CMOS or if

I want to look for outreach experts job

job post so I'm looking for this type of

job post

then I scraped the specific job post I

looked up the company that posted the

job post I

looked at the amount of employees that

they have because I'm not targeting

companies that have more than let's say

50

employees I then went to the LinkedIn

page to the company page and I sorted

out not sorted out but search for the

sea level let's say I'm looking to

discuss with CMOS or CEOs and then I

automatically sent likes to the recent

post in different intervals and only

after sending a few likes and warming

them up I used to send an automated

DM now as I said a while after I started

using this automation I figured out that

Apollo which is a lead sourcing database

let me share this with

you

Apollo they also have this functionality

in which you

can filter out

prospects based on a job post so let's

say you can go to Apollo look for only

for CEOs who just in companies that just

posted a job post looking for

CMOS so basically if I knew about this

feature before I didn't have to build

this whole automation on my

own and this just goes to show you that

the more you expand your knowledge and

know and become aware of all the

solutions that different uh

service providers are offering it gives

you a better understanding of how you

can optimize the workflow and make it

more effective and

efficient anyway

um basically the process that was

looking for over here is let's say I

want to find jobs as a creative

director only job posted in the past

week or so only in the

US

only remote

positions and only experience

level mid senior director and

executive so all of these job post might

be relevant for me assuming that I'm

selling um creative director assuming

that I'm a creative director and what I

I wanted to do what I wanted to achieve

with this automation is just go into

each of

these job post find the company in this

case Robert

Half then come

here look for the

CMO make sure that the company isn't too

big in this case it's too big it's more

than 10K employees it's irrelevant for

me and reach out to the specific CMO

obviously also like his recent post

automatically just to warm him up and

after liking his recent post for like

three days in a row or in different

intervals keeping them random so it

won't look

automated only afterwards I would send

the

DM and this whole process I automated

the whole process a few things that um I

try to do in every automation process

before we dive into the automation is

obviously we want to make it look as

human as possible so we add random

delays also um in the automation with

regards to if we're using LinkedIn or

Facebook so we add random delays so it

won't look as if we automated the

process and also add like weights like

every 15 um likes that I was sending I

sent like 30 minutes of

wait um and also in between each like I

added a random DeLay So it it shouldn't

be too fast and it also shouldn't be

like

too the frequency

the delay should should be like there

there should be some sort of

variability in the amount of delays just

to make it look more

human

um now let me dive into the

configuration H before I dive into the

configuration files and show you like

under the hood of how I built this in

this specific instance I use someone to

help me because I was building

automations and I was achieving pretty

good results but I wanted someone to

upgrade my skills and help me

build the automations in a more robust

manner so whenever I'm looking for help

in a certain field that I want to

improve my game I go to two places the

first place and this is not

straightforward people overlook this is

preply preply is a platform for finding

H personal

tutors so you can find over here tutors

to teach you um languages Spanish

English whatever you feel like and also

you can find tutors in specific H for

specific um business

related stuff so for example if you want

to Learn Python you can come here let

me make this bigger you want to learn

python or

computer science

accounting web development whatever you

want you just set the price that you're

willing to pay per lesson the country of

birth of the

teacher and that's pretty much it so

let's say we want to look for

python so we have over here a list of of

python teachers some of them charges

only like five bucks per hour so this is

a total no-brainer so I highly recommend

checking out preply if you want

something

specific but you can't find it over here

because here they have like this drop

down which is pretty limited you can

just go to

Google and you use Google search operant

so site preply.com

and let's say we want something specific

let's say

RPA so not everyone like they don't they

don't offer the option of RPA right out

of the

gate in these

Alternatives because you can't write RPA

here it will lead you to nothing so you

can just come

here and do like this

RPA so I don't know if you're aware of

Google search operan but it's very

valuable and important knowing the

alternative that you can use in order

to add like an advanced layer of filter

T in or searching in Google but in this

case we are looking for stuff only

specifically in the website preply.com

and only the world RPA now if we write

this we now find a list of people a list

of tutors in preply that mentioned the

word RPA in their um

profiles so now we can find tutors

specific for our

needs or let's say um selenium so

selenium is

a is a library for

automating uh testing and also for

scrapping stuff and we could find python

tutors who teach selenium but if you

want to go directly to selenium to

tutors we can just look it up like

this by the way I highly recommend that

you check out afterwards um Google

search operons

Google

search let's just show you it right now

operators so we find Google searches and

over here you have a list of all the

Google searches that you can do you have

like in specific sites you can do

exclusions obviously dates let's take a

look over here what else you can just

use this by the way you can just use

this

one just fill out what whatever you want

specific words exact words any of the

words specific languages last update

there is a ton of options and it's just

something that is good to know if you're

looking

for to refine your search so this is the

first alternative prep and second

alternative is obviously going to upoke

and looking for tutor you can just go to

Appo over here search

for I normally I wouldn't search for a

talent I just would

upload a specific job post you once you

log in you upload a job post you you say

looking for an RPA tutor and this is how

I found the guy who help me with this

specific

automation now let's go back to the

automation so as I

mentioned there are a few parameters

that I wanted

to add at the beginning so this is the

bot name this is the configuration file

this is

a cell in which I added up

front all the job post you all the the

the job looking URLs that I'm interested

in so for

example I wanted to find out job post in

the US and UK and Germany

and Netherlands and Israel Etc and

obviously you can't look all of these

locations in the specific I mean here

you can you need always to State the job

position and the state country so what I

did I created over here I just copied

this URL let's say this is in the United

States I copied this URL and I added it

to the

Excel then I wrote

Canada

Canada and again I copy this

URL and I pasted all the URLs over here

separated by comma because what I want

the automation to do is go into each one

of this in each one of these um URLs and

look for the specific position that I

wanted and then find the CEO or CMO

whatever and then start liking his post

the next cell is where is this going to

be saved over here we have the bot

starting time so just a trigger if we

want to automatically have the B start

in specific times and the B ending

time and over here I defined if I wanted

to send likes I can turn off the liking

of recent post over here in the

configuration file over here we set like

some boundaries with regards to how many

profiles will the bot scrap how many

profiles will scrape per hour the amount

of likes it will send per

profile and how many seconds we will

wait between each

[Music]

iteration and what else is interesting

over here and this is the position that

we were looking for so in every company

that posted a job

post I had to Define up front who would

I like to reach out to because I'm not

interested in reaching out to the guy

who posted the job post it might be an

HR or just a team leader but I always

like going to the top of the food chain

and in this case it's the COO or the CMO

so this is the configuration

file over

here we have the output

file which as you can see you have we

have the

link we have when did we like the recent

post the amount of like that we

sent if we added any comments which I

didn't do because I still don't trust

uh GPT to create good enough comments

and send them on my behalf but as you

can see we have a ton this specifically

I after

after this is an old file so we don't

have many of these but as you can see

over here we have like

a summary if we sent any likes if the

company has more employees than the

criteria it also notifies

me yes yes so you get the idea now let's

dive into the automation itself and I'm

not sure how deep should I dive over

here

because um it might be too Technical and

boring but on a high level what we did

is like I mentioned

before we start we configure all the

files you can see over here we have the

configuration file and then we have the

main

flow and then we open

each job posting um

URL we

fetch all the job post we also made it a

point to scroll until the B bot of the

page because

sometimes as you can see over here

sometimes you

have this specifically if you don't have

if you don't have premium and if you use

this wait wait a second I'll show you a

better example United

States not enough results let me do like

any time 70 results so we also created

the automation that it was called down

until the bottom and then move to the

next page because obviously if there are

more um results we want to grab all of

them and after grabbing all the results

we place them in the Excel file and the

next step would be entering each

company page because we scraped this

which is the name of the company in each

of the job

posts and then we went to the company

page we scraped the amount of employees

added them to the file and the next step

would be going to the

people writing over here coo or CMO

whatever and

fetching the CEO his uh LinkedIn profile

URL so we will be able to go to his

profile URL afterwards go to the Post

section let's as an example let's go to

my

profile

over

here you can go to the

activity post and it's always the same

it's always this is always the pattern

of recent

posts so this is very valuable if you

automate and you understand a specific

pattern you just make sure that whatever

profile that

you scraped the URL of the profile that

you script instead of going to the

profile and clicking a few buttons I I

just added

automatically this um slug or

continuation of the URL I don't even

know how you call it in English and

automatically we could have arrived to

the recent post of the Prospect and like

his recent

post The Next Step um yeah I guess that

was the last step in the automation

afterwards I used the different software

to send the DMS because that software

which is called wxy I also did a video

about wxy a while ago and it didn't get

too many views I don't know why if you

guys don't find LinkedIn automation

interesting uh obviously let me know but

anyway this is on a high level the

workflow if you guys are interested in

in that I will dive deeper into how we

build this I'm I'd be happy to um share

another video about like all the

nitty-gritty stuff because we build this

very nicely obviously uh I must give the

credit to Michael who helped me build

this because he's a real professional

when it comes to

RPA he's a perfectionist I prefer moving

faster and building stuff breaking stuff

but just putting in a ton of volume and

I don't care so much about the

robustness unless I give the automation

to a client that doesn't know how to

handle RP and then obviously I built

robust stuff but when I build stuff on

my own I don't care so much about the

robustness I just wanted to work and as

fast as possible because automations

constantly change automations constantly

break so I don't emphasize building obot

stuff but this thing that Michael helped

me build is really a state of art um and

if you

want me to dive deeper into what we

built over here just let me know and I

can do a followup

video um let me take a sip of water and

think if there's anything else that we

didn't

cover um no I think that's it for

today you guys wanted me to share more

use cases I have a ton of use cases that

I've built using RPA

this is one of the most interesting and

complex ones that I built and I also

believe it's very useful most of the

people in the B2B space and prospecting

space um I I wouldn't say most but a lot

of people are just pouring in volume I'm

a big believer in volume I think it

makes sense to put in a lot of volume

but on the other hand all the platforms

LinkedIn Facebook also called emailing

they all impose limit ation so it's

better to P to pour in volume with

quality this is why I think this

workflow is specifically very smart

because we are only targeting people

with high intent of buying so if someone

posted a jaw post a few days ago that

they are looking for media buyers

obviously it make more sense to reach

out to them and offer them a media

buying media buying Services just

scraping a ton of CMOS in different

different companies in the US and just

sending them a cold Outreach and

pitching them with a with with media

buying services so this is why I think

this

is a very valuable workflow for anyone

and I highly recommend that you guys

think try to think outside of the box

how you can figure if you're doing B2B

prospecting try to figure out how you

can find customers with high intent of

buying and only based on the high intent

create the automation or do the manual

reach out obviously this can all be done

manually but I mean what's the point of

doing stuff manually if you can automate

them um yes let's start to conclude this

H video for today if you guys have any

questions or comments regarding this

video any questions regarding the

workflow regarding RPA regarding my

train of thought when I build this leave

them in the comment section below

obviously if you guys have ideas how I

can improve the

videos optimize the video make it more

interesting make it more useful just

leave them in the comment section below

I'd be happy to optimize I welcome any

type of

criticism besides that if you think this

video was valuable obviously please

leave a like below and make sure that

you subscribe and if you can share this

video with someone that you think might

find it useful that would be

awesome yes I guess that's it until next

time keep on

automating