Rise of the Revenue Robots
Automation and AI combined form the greatest force in modern business.
AI has become the brain and nervous system of successful companies, constantly collecting and comparing data.
Sensor readings are compared to forecasts.
Performance fine tuned constantly.
Causal modeling creates maps and forecasts that the AI uses to asses different interventions.
Predicting the future.
Once an optimal decision is detected the AI engages its muscle - automation.
Visualize AI as a super brain in a glass jar.
Endlessly capable, but impotent.
Isolated and inert.
But when you put that Brain inside a titanium robot now suddenly you've got an engine ready to execute work.
And the exciting thing is - AI can analyze its own performance and recursively improve.
This means the work is:
Continuous —> no breaks, no Holidays, no sick days…
Observable —> realtime sensor data vs emails/Slacks to your colleagues asking for updates
Reliable —> no variance in output rates or quality
There are physical robots manipulating and interacting in the real world - as well as digital robots, frequently called Agents.
The cutting-edge companies are layering physical and digital robots into a revenue engine.
Amazon has spent years integrating physical and digital robots into their work streams.
The complexity and power of AI rises geometrically with the number of Agents.
Automation evolves from integration into orchestration when you layer multiple systems, teams & protocols.
AI and Automation combined create perfect workers that never skip a day, miss an email or misbehave.
Python is the language that brings AI and Automation to life and combines them.
Python can significantly enhance your automation capabilities by providing various libraries and functions that enable efficient data processing, analysis, and reporting - let's explore a few of the most useful automations:
Data extraction and manipulation
Revenue intelligence report generation and transmission
Revenue system synchronization
Python-powered automation not only executes the commands of the AI, it also captures and processes the sensor data that feeds the AI-decision-making process.
Python Libraries
Side Note: I am going to write dedicated Substack Notes on most of libraries we work with - here are a few favorites we’ll be using and re-visiting frequently:
Beautiful Soup
for web scraping and parsing HTML and XML documents
Requests
for making HTTP requests
Pandas
for data manipulation and analysis
Matplotlib
for creating static, animated, and interactive visualizations
ReportLab
for creating PDF reports with custom formatting and graphics
SQLAlchemy
SQL toolkit & Object-Relational Mapping library for Python
PyMySQL
to connect and interact with MySQL databases
Scikit-learn
for predictive modeling and data analysis
Statsmodels
for statistical modeling and hypothesis testing.
smtplib
for sending emails using the Simple Mail Transfer Protocol
email
for generating, parsing, and manipulating email messages
APScheduler
for scheduling and managing tasks, such as sending emails at specified times
Let’s explore a specific example of intelligence and automation from my two worlds:
Family Office Investing
Revenue Operations
In order to succeed you need timely and accurate information.
You need to understand the dynamics of the customer base because that's where revenue comes from.
Time series analysis is a great tool to study changes in revenue dynamics.
It can also be a key sensor reading that informs AI’s automation decision-making.
When the McDonagh Family Office is evaluating an investment we utilize time series and other exploratory analysis.
For subscription businesses it is vital to understand survivorship of customer contracts… do you sign a lot of deals and also lose a lot of customers?
Data to the rescue - we can connect Salesforce to a Python notebook, conduct a time series analysis and even automate this to run at intervals in the future.
The Code:
# Import necessary libraries
import pandas as pd
from datetime import datetime
from lifelines import KaplanMeierFitter
# Load the data
df = pd.read_csv('salesforce_accounts_time_series.csv')
# Print all column names
print(df.columns)
# Rename the columns to remove any spaces
df.columns = df.columns.str.replace(' ', '_')
# Convert 'Activation_Date' to datetime format
df['Activation_Date__c'] = pd.to_datetime(df['Activation_Date__c'])
# Create 'Event' column: 1 if 'Active_ARR' is 0 (account deactivated), 0 otherwise
df['Event'] = (df['Active_ARR__c'] == 0).astype(int)
# Compute 'Duration' as the number of days from 'Activation_Date' to today's date
df['Duration'] = (datetime.today() - df['Activation_Date__c']).dt.days
# Drop rows with NaN values in 'Duration' and 'Event'
df = df.dropna(subset=['Duration', 'Event'])
# Initialize the KaplanMeierFitter model
kmf = KaplanMeierFitter()
# Fit the model on the data ('Duration' and 'Event' columns)
kmf.fit(df['Duration'], df['Event'])
# Create an estimate of the survival function
estimate = kmf.survival_function_
# Print the estimate
print(estimate)
# Plot the survival function
kmf.plot_survival_function()The Output:
The initial results are represented in a uninteresting way.. but the information is priceless.
The timeline is measured in days, and this time series tells me survivorship drops off significantly from day 13 to day 1,865.
…but that’s 5-years out!
Most SaaS companies and other recurring revenue model companies are thrilled to see 22% of a cohort still contributing ARR 5-years after their contract activation date.
This is the power of visualization - we can bring this table to life:
The Point:
When this value drops under an alert threshold, we receive Slack / e-mail even SMS.
Python can “talk” to the outside world.
Where is the automation, you may ask?
Everywhere.
When you layer 20 to 30 of these different analytic reports together and design alerts triggered by values and values-vs-controls you have:
an early warning system for problems
an opportunity heatmap
Most importantly.. you have time.
Time to implement counter-measures.
Or time to align, amplify, optimize and monetize.
AI will continue to grow in power while simultaneously simplifying user interfaces and experiences.
Automation is rising rapidly with the tide of the AI.
Combined they are transforming the landscape of commerce and creativity.
Small companies have the capacity to quickly deploy powerful asymmetric capabilities.
If these are well-timed, and well-aimed…
David will fell a lot more Goliaths in the coming months and years.
That is Life in the Singularity.
Accelerating capabilities.
Connected cognification.
Incredible opportunities for people and firms invested in learning & applying these learnings to building revenue robots and more.
Subscribe now and spread this so we can learn and grow together!