SMEs and Marketing: AI Assists Without Replacing Strategy

Three female leaders of small businesses explain how they use ChatGPT, Claude, or Meta's tools to accelerate analysis and production while maintaining control over key decisions. They discuss learning costs, the importance of measuring the entire conversion funnel, and their skepticism regarding the automatic recommendations from platforms.

Costs, Creative Variables, and Decisions Not to Delegate

Andrea Garcia Fry warns that paid marketing can become very expensive quickly, particularly because the learning phase of Meta's algorithms requires time and budget without guaranteeing the expected results. For Arquia, finding the right customer on the platform has been challenging with a limited budget, and strengthening the brand along with organic marketing has yielded better results than paid campaigns. Bonnie Gurry emphasizes that it is difficult to know precisely why an ad works and that isolating creative variables to understand performance, without wasting capital, requires constant balance. She specifies that she would not delegate capital allocation or brand positioning decisions to a tool without verifying that its recommendations align with the company's objectives. She also cautions against quick clicks that can drain a marketing budget without return on investment if the leads are not qualified, recommending not to entrust strategy to an AI, especially if it is developed by the platform selling the ads.

Winning Procedures: Starting from Organic, Testing, and Tracking the Funnel

To kick off advertising campaigns, Andrea Garcia Fry first selects content that has already been successful organically, then adapts or promotes it through paid campaigns. She consults ChatGPT or Claude to analyze the setup, examine the data, and identify testing opportunities, while retaining the final say on decisions. Catalina Parker adopts an experimentation-focused approach: she identifies key indicators, tracks them over time, conducts modification trials, and then decides based on the results obtained. She stresses the importance of analyzing the entire conversion funnel, rather than simply increasing the budget because leads seem cheap, and recommends tracking each step after exposure to the ad, from signing up for a lead magnet to making a purchase. She cites a successful, simple, text-based campaign that has been running for over two years with very few adjustments.

Tool Choices: Welcome Assistance, Interpretation to Control

Andrea Garcia Fry prefers using ChatGPT or Claude because she can provide them with detailed context and consider them as thought partners. Bonnie Gurry believes that tools like Meta AI are useful for speeding up diagnostics, spotting creative trends, and quickly generating text variations. Catalina Parker, who has tested recommendations from Meta Ads Manager that she found poorly suited to her business, reminds that Meta benefits when advertisers increase their ad spending and prefers independent analyses covering her entire marketing ecosystem. Andrea Garcia Fry thinks that Meta's tool would be truly useful if it learns deeply how the business operates, rather than simply offering more automation or budget.

The Framework: New AI Functions from Meta and Common Uses of Generative AI

In August, Meta announced new AI features aimed at small business owners, designed to audit organic posts and ads on Facebook and Instagram, compare a business's performance to that of similar brands, and provide strategic suggestions to enhance marketing efforts. Many small business leaders also use generative AIs like ChatGPT, Claude, and Gemini for marketing tasks such as prospecting and content ideation. Among them, Andrea Garcia Fry (Arquia), Catalina Parker (Relatable Nonprofit), and Bonnie Gurry (GreenPortfolio) detail their practices. Andrea Garcia Fry reminds us that effective paid marketing requires a solid understanding of brand positioning, tone, business objectives, and the customer journey.