Comment on page
Generate leads from posts
Tags: #linkedin #post #comments #naas_drivers #content #snippet #dataframe
Last update: 2022-01-09 (Created: 2022-01-09)
Description: This notebook provides a guide to leveraging LinkedIn posts to generate leads for your business.
Disclaimer:
This code is in no way affiliated with, authorized, maintained, sponsored or endorsed by Linkedin or any of its affiliates or subsidiaries. It uses an independent and unofficial API. Use at your own risk.
This project violates Linkedin's User Agreement Section 8.2, and because of this, Linkedin may (and will) temporarily or permanently ban your account. We are not responsible for your account being banned.
from naas_drivers import linkedin, hubspot
import pandas as pd
import numpy as np
import naas
from datetime import datetime, timedelta
import requests
import json
# Lindekin cookies
LI_AT = "AQEFALsBAAAAAAbjCtIAAAF_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
JSESSIONID = "ajax:42778xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx"
# Enter your profile URL
PROFILE_URL = "<YOUR_LINKEDIN_PROFILE_URL>"
# HubSpot API key
HS_API_TOKEN = "<YOUR_HUBSPOT_API_TOKEN>"
# Your company located after "reports-dashboard/" when you are connected in HubSpot https://app.hubspot.com/reports-dashboard/2474088/view/244930
HS_COMPANY_ID = "<YOUR_HUBSPOT_COMPANY_ID>"
# Hubspot contact URL
HS_CONTACT_URL = "https://app.hubspot.com/contacts/" + HS_COMPANY_ID + "/contact/"
# Contact owner => contact id to whom task needs to be assigned
HS_OWNER_ID = "<YOUR_HUBSPOT_API_TOKEN>" # remove double quotes from owner id (to be added as integer value)
# Time delay to set due date for tasks in days
time_delay = 10
# Calc timestamp
tstampobj = datetime.now() + timedelta(days=time_delay)
tstamp = tstampobj.timestamp() * 1000
hs = hubspot.connect(HS_API_TOKEN)
SEND_EMAIL_TO = "<YOUR_EMAIL_ID>"
df_posts = linkedin.connect(LI_AT, JSESSIONID).profile.get_posts_feed(
PROFILE_URL, count=100
)
df_posts.head()
def get_likes(df_posts):
DF_all_post_likes = pd.DataFrame()
for index, row in df_posts.iterrows():
df = linkedin.connect(LI_AT, JSESSIONID).post.get_likes(row["POST_URL"])
DF_all_post_likes = DF_all_post_likes.append(df)
get_likes(df_posts)
def count_likes(df):
to_group = ["PROFILE_URN", "PROFILE_ID", "FIRSTNAME", "LASTNAME"]
# df = df.groupby(to_group, as_index=False).agg({"PROFILE_ID": "count"}).reset_index(name='count')
# df = df.sort_values(by="PROFILE_ID", ascending=False)
df2 = (
df.groupby(["PROFILE_ID"])
.size()
.sort_values(ascending=False)
.reset_index(name="LIKE_COUNT")
)
return df2
df_counts = count_likes(DF_all_post_likes)
df_counts
- Grouping the like count and classify the leads as per levels i.e. L1, L2, L3, L4. ( from highest to lowest counts of likes) with column name "POTENTIAL LEAD"
- Group posts as per the the column "POTENTIAL LEAD"
max_likes = df_counts["LIKE_COUNT"].max()
cluster_1 = round(max_likes * 0.1, 0)
cluster_2 = round(max_likes * 0.5, 0)
cluster_3 = round(max_likes * 0.8, 0)
conditions = [
(df_counts["LIKE_COUNT"] <= cluster_1),
(df_counts["LIKE_COUNT"] > cluster_1) & (df_counts["LIKE_COUNT"] <= cluster_2),
(df_counts["LIKE_COUNT"] > cluster_2) & (df_counts["LIKE_COUNT"] <= cluster_3),
(df_counts["LIKE_COUNT"] > cluster_3),
]
values = ["L4", "L3", "L2", "L1"]
df_counts["POTENTIAL_LEAD"] = np.select(conditions, values)
df_counts
- filter rows from "grouped_profile_posts" dataframe based on L1, L2, & L3 levels and add data to dataframe "df_leads"
- Iterate through dataframe "df_leads" and add items to leads_list
- Extract EMAIL, FIRSTNAME, LASTNAME, PHONE_NUMBER, OCCUPATION from profile URL
df_leads = df_counts.loc[df_counts["POTENTIAL_LEAD"].isin(["L1", "L2"])]
df_leads
leads_list = []
r_count = 1
for index, row in df_leads.iterrows():
profileid = row["PROFILE_ID"]
profileurl = "https://www.linkedin.com/in/" + profileid + "/"
PROFILECONTACTS = linkedin.connect(LI_AT, JSESSIONID).profile.get_contact(
profileurl
)
PROFILEIDENTITY = linkedin.connect(LI_AT, JSESSIONID).profile.get_identity(
profileurl
)
profileemail = PROFILECONTACTS.at[0, "EMAIL"]
profilephoneno = PROFILECONTACTS.at[0, "PHONENUMBER"]
profilename = (
PROFILEIDENTITY.at[0, "FIRSTNAME"] + " " + PROFILEIDENTITY.at[0, "LASTNAME"]
)
profilefirstname = PROFILEIDENTITY.at[0, "FIRSTNAME"]
profilelastname = PROFILEIDENTITY.at[0, "LASTNAME"]
profileoccupation = PROFILEIDENTITY.at[0, "OCCUPATION"]
leads_list.append(
[
profilename,
profilefirstname,
profilelastname,
profileemail,
profilephoneno,
profileoccupation,
profileurl,
]
)
r_count = r_count + 1
HS_CONTACT_ID_LIST
- The list to store the contact id recently created and the linkedin profile URL of associated contact
# to store the resulting output of create contact method
contact_id = ""
HS_CONTACT_ID_LIST = []
for i in leads_list:
profilename = i[0]
profilefirstname = i[1]
profilelastname = i[2]
profileemail = i[3]
profilephoneno = i[4]
profileoccupation = i[5]
profileurl = i[6]
# With send method
data = {
"properties": {
"linkedinbio": profileurl,
"firstname": profilefirstname,
"lastname": profilelastname,
"jobtitle": profileoccupation,
"email": profileemail,
"phone": profilephoneno,
"hubspot_owner_id": 111111086,
}
}
# write data to CRM ( create new contact in CRM)
contact_id = hs.contacts.send(data)
HS_CONTACT_ID_LIST.append([contact_id, profileurl])
Build Email Template
- This includes the hubspot contact_id, contact URL and LinkedIN URL
table_header = f'<table style="border-collapse:collapse;border-spacing:0;font-family:Arial"><thead><tr><th style="border-color:black;border-style:solid;border-width:1px;padding:6px 10px;text-align:left">Contact ID</th><th style="border-color:black;border-style:solid;border-width:1px;padding:6px 10px;text-align:left">Hubspot URL</th><th style="border-color:black;border-style:solid;border-width:1px;padding:6px 10px;text-align:left">LinkedIN URL</th></tr></thead><tbody>'
table_body = ""
table_footer = "</tbody></table>"
for i in HS_CONTACT_ID_LIST:
hs_contact_url = HS_CONTACT_URL + i[0]
linkedin_url = i[1]
tablerow = (
'<tr><td style="border-color:black;border-style:solid;border-width:1px;padding:6px 10px;">'
+ i[0]
+ '</td><td style="border-color:black;border-style:solid;border-width:1px;padding:6px 10px;">'
+ hs_contact_url
+ '</td><td style="border-color:black;border-style:solid;border-width:1px;padding:6px 10px;">'
+ linkedin_url
+ "</td></tr>"
)
table_body = table_body + tablerow
table = table_header + table_body + table_footer
email_body = (
f'<p style="font-family:Arial">Hi there,</p><br/><p style="font-family:Arial">Following new task(s) has been created for you :</p><br/>'
+ table
)
def create_task(
owner_id,
tstamp,
contact_id,
contact_props,
contact_linkedin_url,
hs_contact_url,
engagement="TASK",
):
"""owner_id=HS_OWNER_ID, tstamp=tstamp, subject=subject, body=body, status=status, engagement="TASK"
Engagement type = TASK | NOTE | EMAIL | MEETING | CALL
"""
payload = json.dumps(
{
"engagement": {
"active": "true",
"ownerId": 111111086,
"type": "TASK",
"timestamp": tstamp,
},
"associations": {
"contactIds": [1551],
"companyIds": [],
"dealIds": [],
"ownerIds": [],
},
"metadata": {
"body": "Hi there, you need to contact following user & task is already assigned to you.<br/>"
+ "Name :"
+ contact_props["firstname"]
+ contact_props["lastname"]
+ " Contact URL : "
+ hs_contact_url,
"subject": "Task created for Contact ID :" + contact_id,
"status": "NOT_STARTED",
"forObjectType": "CONTACT",
},
}
)
url = "https://api.hubapi.com/engagements/v1/engagements"
params = {"hapikey": HS_API_TOKEN}
headers = {"Content-Type": "application/json"}
# Post requests
res = requests.post(url, data=payload, headers=headers, params=params)
# Check requests
try:
res.raise_for_status()
except requests.HTTPError as e:
raise (e)
res_json = res.json()
# Fetch the task id of the current task created
task_id = res_json.get("engagement").get("id")
print("🎆 Task created successfully: ", task_id)
return res_json
TASK_ID_LIST = []
for i in HS_CONTACT_ID_LIST:
contact_id = i[0]
contact = hs.contacts.get(contact_id)
contact_props = contact.get("properties")
contact_linkedin_url = i[1]
hs_contact_url = HS_CONTACT_URL + contact_id
# print(hs_contact_url)
result = create_task(
owner_id=HS_OWNER_ID,
tstamp=tstamp,
contact_id=contact_id,
contact_props=contact_props,
contact_linkedin_url=contact_linkedin_url,
hs_contact_url=hs_contact_url,
engagement="TASK",
)
TASK_ID_LIST.append(result.get("engagement").get("id"))
if not TASK_ID_LIST:
print("No tasks created")
else:
email_to = SEND_EMAIL_TO # to send the report
subject = "LinkedIN Leads Alert"
content = email_body
naas.notification.send(email_to=email_to, subject=subject, html=content)
Last modified 24d ago