AI agency is a TRAP - I'm shutting it down

Shutting Down an AI Agency: Lessons Learned

Introduction to the AI Agency Experience

• The speaker built an AI agency that generated $80K a month in revenue but is now shutting it down, emphasizing that gross revenue does not equate to net profit.

• The speaker will discuss two main reasons why the AI agency model is broken and share insights on their current projects with improved margins.

Background of the Speaker's Journey

• With over four years of experience in building AI solutions, the speaker began as a freelancer before the rise of ChatGPT, initially using GPT-3 and custom models.

• Following the launch of ChatGPT, demand surged, prompting the speaker to hire more staff and develop various custom solutions until focusing on AI agents by late 2023.

Financial Realities of Running an AI Agency

• Despite reaching $80K monthly revenue with 15 active clients, actual profits were significantly lower due to high operational costs.

• Less than 30% of revenue was profit after accounting for contractor payments and business expenses; this discrepancy highlights hidden challenges in service delivery.

Challenges with Custom Solutions

• Building custom AI solutions lacks scalability; unlike other agencies where outputs are similar, each client project varies greatly in requirements.

• An example illustrates how a client's evolving demands turned a simple project into a complex software product requiring extensive management.

Issues with Scaling and Profitability

• As client numbers increase, so do complexities and issues; scaling does not inherently improve profit margins in this model.

• Many successful figures can charge premium rates due to their established brands or additional income from courses rather than direct client work.

Future Outlook for AI Agencies

• The second major reason for discontinuing the agency is that as AI technology improves, traditional agency roles may become obsolete by 2026.

• Automation tools allow businesses to handle tasks independently without needing developers; this shift diminishes demand for traditional agency services.

Evolving Business Models

• Current trends show many agencies are merely rebranding basic prompt engineering services instead of offering substantial value.

• Some YouTubers propose new consulting models focused on strategy rather than implementation; however, these may struggle as knowledge workers increasingly adopt essential AI skills.

This structured summary captures key insights from the transcript while providing timestamps for easy reference.

Appointment Booking Strategy

New Pitch and Pricing Model

• The new pitch focuses on booking qualified appointments directly onto clients' calendars, with payment only required for confirmed appointments. If a meeting ends early due to unqualified leads, the appointment is replaced at no cost.

• This outcome-based pricing model simplifies sales processes, making it easier to close deals with potential clients.

AI Personalization in Sales

• Personalized Loom videos are created using AI, enhancing engagement by addressing prospects by name and showcasing their websites. This approach significantly boosts conversion rates.

Metrics Tracking and Revenue Generation

• After launching campaigns, metrics are tracked daily via Google Sheets to ensure performance aligns with KPIs. Recently onboarded clients generated four appointments worth $1,000 in revenue at a cost of $250 per appointment.

• The overall costs associated with API and infrastructure amount to approximately $11 per booked appointment, maintaining profit margins above 70%.

Shifting Business Focus

Transition from Service Selling to System Building

• The primary goal should shift from selling AI agency services to building robust AI systems that deliver consistent results. Clients prefer proven workflows over experimental solutions.

Internal Client Acquisition Pipeline

• Initially developed as an internal solution for client acquisition challenges, this pipeline proved so effective that it became more beneficial to sell outcomes rather than individual services.

Branding Strategy

• Agencies should rebrand themselves not as AI-focused but as traditional service providers (e.g., marketing or accounting agencies), while operating like an AI startup behind the scenes.

Automation and Scalability

Outcome-Oriented Approach

• Emphasizing measurable outcomes (like increased revenue or saved time), allows for automation of service delivery through AI, minimizing manual work post-onboarding.

Case Study: Successful Agency Model

• A podcast example highlights an agency valued at over $5 million that operates efficiently without heavily branding itself around AI despite its automated processes.

Challenges and Opportunities in Automation

Time Commitment for Results

• While this model requires upfront investment in identifying valuable processes for automation, the potential returns justify the initial effort due to faster testing capabilities enabled by modern AI tools.

Idea Generation Using AI Tools

• Utilizing tools like ChatGPT can help identify time-consuming processes within specific industries (e.g., solar companies), providing insights into areas ripe for automation.

Detailed Process Overview

Appointify Workflow Breakdown

• The process begins with onboarding clients who commit to a minimum number of appointments. It includes generating essential documents (ICP and offer documents), crucial for effective marketing strategies.

Stages of the Process:

1. Onboarding: Initial client commitment triggers onboarding steps.

2. Execution & Optimization: Continuous improvement based on feedback loops.

3. Client Feedback Loop: Ensures ongoing adjustments align with client needs and expectations.

How to Automate Client Onboarding and Campaign Management

Client Document Creation and Mailbox Setup

• The team assists clients in creating essential documents, enhancing their market positioning while simultaneously setting up mailboxes for client accounts.

• Mailbox warm-up is initiated by instructing an agent, which updates all accounts in the automation software and flags any setup errors.

Lead Scraping Process

• Leads are scraped from various sources like LinkedIn Sales Navigator and Google Maps, utilizing a previously generated Ideal Customer Profile (ICP) document.

• A LinkedIn search is generated based on the ICP to find leads, which are then automatically imported into Clay for data enrichment.

Data Enrichment and Voice Cloning

• Clay scrapes detailed information about leads, including company details and personal information, using an email finder tool.

• Before sending data to the backend, voice cloning is performed using 11 Labs; this process automates voice generation compared to other platforms like Senspark.

Campaign Launch Preparation

• Clients record personalized Loom videos with AI-generated scripts based on earlier documents. If preferred, standard voices can be used instead of the client's own.

• Once everything is set up correctly, launching the campaign involves verifying all parameters through an agent that manages campaign execution.

Ongoing Optimization and Performance Tracking

• An integrated agent tracks daily campaign performance metrics via scheduled triggers that log data into Google Sheets.

• Adjustments are made if KPIs are not met by modifying email copy or lead sources while ensuring clients focus on sales calls.

Automation Efficiency in Client Onboarding

• The entire three-stage process operates largely autonomously without needing additional personnel for onboarding new clients.

Steps to Start a Service-Based Business Using AI

Week-by-week Strategy Overview

• In week one, select a service-based industry with existing demand (e.g., accounting or SEO), engaging with experts to understand workflows.

Manual Selling Approach

• During week two, sell outcomes manually as a traditional agency while considering incorporating "AI" into branding temporarily.

Building Automated Systems

• In week three, begin developing systems for repetitive tasks after securing initial client agreements; ensure value exists even without AI involvement.

Iteration Based on User Feedback

• Week four focuses on gathering user feedback aggressively to refine the system until consistent outputs are achieved.

Potential Impact of AI Integration

• Successfully building such automated systems within a month can significantly transform business operations leveraging AI technology.