00:00manual lead research is a massive bottleneck we are going to build an automated system that finds
00:06leads evaluates them and writes the perfect pitch for you while you sleep take a raw list of linkedin
00:13profile links click a single button and watch a spreadsheet automatically fill with prospect
00:19scores and highly customized outreach messages this flow chart shows exactly how the data moves
00:25we start with a linkedin url which feeds into a scraper like air top or bright data that data moves
00:32into an ideal customer profile calculator to score the lead then passes to claude ai to write the
00:38message and finally lands neatly in google sheets to build this you need four specific tools first
00:45an automation platform like n8n or make second a blank google sheet third a scraping api to extract
00:52the profiles and fourth access to the anthropic api to power the copywriting by linking these
00:58platforms together you offload the repetitive research tasks to the machine this shifts lead
01:04generation from a time-intensive manual chore into an autonomous engine that acts as a 24 7 outbound
01:12sales team step one is creating a destination for all this data open a new google sheet and paste your
01:20target linkedin urls down a single column look at these column headers you need exact columns for
01:27row number linkedin url name icp score and icebreaker pay special attention to that row number column
01:35that row number acts as an anchor it tells the automation exactly where to send the enriched data
01:42back to after it finishes processing if you skip this mapping step the ai won't know where to place
01:49its generated copy you'll end up pitching the wrong message to the wrong prospect breaking the entire
01:55system now open your automation canvas in n8n or make add a google sheets module and set it to
02:03read row to pull in your list of links next add an extraction node using either air top or bright
02:10data
02:10paste your api key to connect your account and set the module to target those linkedin urls you just
02:16imported you must tell the extractor exactly what to pull from each profile configure it to grab
02:23three specific data points their seniority level their technical skills and their recent posts this
02:30diagram shows how to handle delays scraping this much data takes time and your workflow might time
02:36out prevent that by adding a weight or pulling logic note to pause the flow until you get a signal
02:42that the
02:42snapshot is fully ready pulling these exact parameters ensures you aren't guessing you upgrade your process
02:49to objectively grade leads based on their verified public data to evaluate these leads we need an ideal
02:55customer profile or icp score add a logic or math node to your canvas to handle the calculation
03:02this scoring matrix shows how we build the formula assign 25 points to senior titles and 30 for executives
03:09then evaluate technical depth award up to 35 points for advanced or expert skills bringing our maximum
03:18potential score to 100 establishing a strict mathematical threshold creates an automated gatekeeper your sales
03:27team will now only spend their time talking to highly qualified decision makers who actually match your ideal
03:34profile let's write the outreach add the anthropic claude node to your workflow and map the recent posts data
03:42from your scraper directly into the prompt field we use claude specifically for this task because its language
03:49model produces sharper copy with far fewer hallucinations than the alternatives look at this prompt structure on screen
03:57you need to assign claude the persona of a sales development rep inject the prospect's recent post into the text
04:04and ask it to write a highly personalized conversational icebreaker left alone ai tends to write overly long
04:12essays that scream spam you must include a strict instruction limiting the output to exactly two lines
04:20come nine real recent social data with a rigidly constrained ai prompt produces outreach that feels
04:28undeniably human allowing authentic personalization to scale across your entire list now we complete the
04:36circuit add a final google sheets module and set the action to update row take the newly generated icp score
04:45and the icebreaker claude just wrote and map them back to their specific columns matching everything up using
04:51that original row number this final mapping step closes the data circuit your workflow is no longer just
04:59a scraper it is a closed loop system that continuously prepares leads for the sales team hit run on the
05:07automation
05:07you will see the nodes firing in sequence while your empty google sheet simultaneously fills with fresh data
05:14look at one of these generated icebreakers it directly references a specific point the prospect made in
05:21their latest post reading exactly like a custom message from a real person doing this manually would
05:27take an hour of tedious scrolling and typing per lead this system delivers the same results with zero human
05:35intervention during the research phase the automation removes the friction of prospecting entirely it hands
05:42your sales team a spreadsheet of perfectly pitched leads giving them operational leverage so they can focus
05:48strictly on closing revenue now you have everything you need to book meetings on autopilot if you want
05:55more tutorials on building automated systems and generating passive income subscribe
Comments