If you listen to the marketing trade press, performance marketing is now entirely run by robots. Agencies compete to talk about their "AI-native autopilot systems" and proprietary algorithms that budget, bid, and write creative on the fly. It sounds highly sophisticated, and it's a great narrative to present to clients.
But if you look inside the actual ad accounts, that "autopilot" is usually just the default automated bidding settings that Meta and Google give you. Autopilot settings are engineered to spend your money as quickly as possible. At PopLab, we treat AI with a heavy dose of pragmatism: we use it for speed and leverage, but we never let it drive the bus.
Where AI Buys Us Speed
We are an AI-native studio, which means we integrate LLMs and script automation to bypass the slow administrative work of marketing. Here is where AI actually adds value in our workflow:
1. Speed in Customer Research
Before launching any ad campaign, we need to know the customer. Instead of manually reading thousands of Amazon reviews, Reddit threads, or YouTube comments, we dump them into an LLM and ask it to cluster pain points, vocabulary, and specific emotional angles. What used to take three days of desk research takes 15 minutes of structured analysis.
2. Creative Script Iteration
AI is an excellent brainstorming partner. Once we write the main hook and concept for a founder-led video ad, we use LLMs to script ten alternative hooks, body angles, or transitions. This gives us raw material to test and shoot, which is crucial for fighting ad fatigue in highly active accounts.
3. Landing Page Copy Formatting
We write our core landing page copy by hand because voice and positioning matter. But once we have the core structure, we use AI to write variant headlines, bullet points, and subheaders for A/B testing. This reduces the friction of launching multiple layout options.
Where AI Stays Out of Our Accounts
While AI is great for leverage, there are specific areas where we actively refuse to use it:
- Strategy & Positioning: AI copy is generic. It defaults to the average of the internet. It uses empty buzzwords and lacks the raw voice that moves high-ticket buyers. The strategic concept must come from human operators.
- Budget Allocation: We do not use third-party "AI budget optimizers." We scale budgets manually because ad platforms have algorithmic limits. Bidding needs a human operator who watches the client's actual CRM show-up rates, not just the platform's pixel indicators.
- Creative Production (Visuals): AI images look fake. They carry an synthetic, glossy look that screams "ad." High-value buyers trust authenticity. We prefer raw, founder-led video or slightly blurred dashboard screenshots over beautiful, AI-generated placeholders.
AI doesn't replace the marketer. It gives a single marketer the speed of a five-person team. By using it where it fits (leverage and formatting) and leaving it out where it hurts (strategy and budget control), we deliver faster results without sacrificing quality.