[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["缺少我需要的資訊","missingTheInformationINeed","thumb-down"],["過於複雜/步驟過多","tooComplicatedTooManySteps","thumb-down"],["過時","outOfDate","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["示例/程式碼問題","samplesCodeIssue","thumb-down"],["其他","otherDown","thumb-down"]],[],[],[],null,["[](/showcase) \nShare\nAUG 29, 2025 \n\nPassionfroot scales creator marketing \\~5x faster using Gemini 2.5 Flash \nVishal Dharmadhikari\n\nProduct Solutions Engineer \nLorenzo De Nobili\n\nCo-Founder and CTO \n\nCreator marketing in sectors like B2B SaaS, AI, and fintech is effective but difficult to scale. Unlike performance marketing, managing creator partnerships often involves manual processes and hands-on relationship management.\n\n\n\u003cbr /\u003e\n\n\nPassionfroot, a creator marketplace connecting brands with creators in emerging tech industries, addresses this complexity by streamlining discovery, outreach, booking, payments and results reporting. To help marketers design and launch effective creator marketing campaigns, Passionfroot developed an AI Campaign Strategist powered by Gemini 2.5 Flash.\n\nAutomating campaign strategy with an AI agent \n\nMarketers often face a \"blank page problem\" when initiating creator campaigns: identifying suitable creators, defining ad types, and allocating budgets. Passionfroot developed its AI Campaign Strategist to automate this process and help marketers scale their efforts.\n\n\n\u003cbr /\u003e\n\n\nThe AI agent uses Gemini models to generate tailored creator marketing campaigns based on specific audience and growth targets. It provides recommendations on creators, budget allocation, content placements and estimates expected CPM and engagement.\n\n\n\u003cbr /\u003e\n\n\nPassionfroot needed a model capable of understanding the nuances of effective creator marketing to guide the agent's strategy.\n\nOptimizing for latency and instruction following with Gemini 2.5 Flash \n\nWhen developing the Strategist, Passionfroot required a model that offered low latency, precise instruction following, and cost-effective performance. Their workflow involves multiple sequential prompts to fulfill a single user request, making speed critical.\n\n\n\u003cbr /\u003e\n\n\nThe team evaluated several advanced models, including Gemini 2.5 Pro. After prompt engineering, they determined that Gemini 2.5 Flash provided exceptional results while significantly reducing latency.\n\n\n\u003cbr /\u003e\n\n\n\"There is no other model that gets even close to the intelligence/latency of Gemini 2.5 Flash,\" said Lorenzo De Nobili, Co-Founder and CTO of Passionfroot.\n\n\n\u003cbr /\u003e\n\n\nThe implementation uses the Vercel AI SDK and relies heavily on structured output (generateObject). Instead of a chat-based interface, Gemini's structured responses directly control the application's user interface.\n\n\n\u003cbr /\u003e\n\n\nThe AI Campaign Strategist workflow involves three key steps:\n\n- **Strategy generation:** Gemini identifies ideal channels, placement types, and budget allocations based on the marketer's input.\n- **Creator matching:** Agentic retrieval-augmented generation (RAG) retrieves and evaluates the most relevant creators based on topical relevance.\n- **Inventory analysis:** The inventory of selected creators is analyzed, and the AI agent selects the top-performing product based on expected ROI.\n\n\u003cbr /\u003e\n\nLaunching creator partnerships \\~5x faster \n\nThe implementation of the Gemini-powered AI Campaign Strategist has delivered measurable improvements in efficiency and growth for Passionfroot's customers.\n\n\n\u003cbr /\u003e\n\n\nBrands using the AI strategist launch partnerships \\~5x faster than brands who manage partnerships manually. This reduction in friction has enabled Passionfroot's marketing partners to triple their growth within the first three months of running creator marketing on the platform.\n\n\n\u003cbr /\u003e\n\n\n\"Our campaign strategist would have simply not been possible without AI,\" De Nobili noted. \"It's incredible to be able to teach a computer about creator marketing so that it can act as a creator marketing expert to provide 1-1 advice to our users.\"\n\nFuture development \n\nPassionfroot plans to enhance the AI Campaign Strategist by performing deeper analysis of creator content, including assessing tone of voice and delivery style. They are also integrating ROI data from collaborations to further refine the agent's recommendations. By combining Gemini's capabilities with proprietary data, Passionfroot aims to provide sophisticated, data-driven insights for creator marketing.\n\n\n\u003cbr /\u003e\n\n\nTo explore how Gemini models can streamline your workflows, review our [API documentation](https://ai.google.dev/gemini-api/docs).\n\n\n\u003cbr /\u003e\n\n\n*Performance metrics were provided by developers and not independently confirmed by Google.* \n\nRelated case studies \n[Instalily\nInstalily is building an agentic search platform designed to help companies automate and run complex workflows in sales, service and operations.](/showcase/instalily) [Volley\nVolley uses Gemini 2.5 Pro to accelerate its game development process, building immersive voice-AI dungeon adventures with visuals from Gemini 2.0 Flash and Veo 2.](/showcase/volley-2) [Viggle\nExperimenting with Gemini 2.0 to create virtual characters and audio narration for their AI powered video platform](/showcase/viggle)"]]