Custom GPTs for SEO: what Victoria Olsina built live on webinar
On December 15, 2025, SEO and AI content consultant Victoria Olsina joined Emanuel Petrescu, founder of How About Some Marketing?, for a live hands-on webinar titled “Using Custom GPTs to Create AI Agents for SEO and Marketing.” The session was built around working demos, not slides, and the core argument was simple: you can automate large chunks of your content and SEO workflow without writing a single line of code.
Olsina has worked with clients including Polkadot, Bankless, Consensys, NEAR, and Aztec, and was nominated for Best SEO in Europe 2024 by the Latam SEO Awards. The webinar drew from that hands-on operator experience.
The GEO Shift: Why Standard SEO Is Not Enough
The session opened with a framing that will be familiar to anyone tracking AI search: traditional SEO is losing ground to what Olsina calls Generative Engine Optimization (GEO). The idea is that platforms like ChatGPT, Perplexity, and Gemini are becoming discovery channels alongside Google, and content needs to be structured for AI comprehension, not just crawler indexing.
The three-part GEO requirement Olsina laid out:
- Content structured for AI comprehension
- An entity-driven content strategy
- Real-time adaptability through AI-powered systems
Custom GPTs, trained on your own brand content, are the practical tool she recommends for bridging that gap.
️ Four Workflows Demoed Live
Olsina walked through four specific workflows during the session, each built and run in real time.
1. GPTs trained on your own content
Upload your existing brand content and configure a GPT to mirror your tone, values, and writing style. The output reflects your voice rather than generic AI output.
2. Keyword to landing page in under 10 minutes
The workflow moves from keyword research to a fully formatted landing page, including metadata, CTAs, and social content, in under 10 minutes. That is the specific time claim Olsina demonstrated on screen.
3. Social content generation at scale
A dedicated GPT agent writes tweets, LinkedIn posts, Instagram captions, and blog outlines while staying consistent with your brand voice across platforms.
4. The C.L.E.A.R. prompting framework
To handle vague or incomplete briefs, Olsina applied her own C.L.E.A.R. method for turning unclear inputs into structured, usable prompts. The framework was not spelled out in full in the recap, but it was presented as a core input quality fix.
The Time Savings Reported
Olsina cited that the GPTs she demoed are already saving marketers 10 to 20 hours per week, freeing up capacity for higher-level strategy work. That figure comes from the presenter, not an independent study.
The Four GPTs She Shared
Four working agents were referenced in the webinar:
- SEO Keyword Research Assistant
- Social Media Content Generator
- Assistant for Writing X (Twitter) Posts
- Full-Service Marketing Assistant GPT
The Transferable Pattern
The core idea from this session is not about any specific GPT. It is about the configuration approach: train the agent on your own content first, constrain the scope to one workflow at a time, and apply a structured prompting method to handle messy real-world inputs.
That pattern holds whether you are building a keyword research tool, a social copy generator, or a client-facing content brief assistant. The no-code framing means the barrier is not technical. It is structural: knowing what the agent should and should not do before you build it.
The webinar replay and presentation are available through the How About Some Marketing? website.
I’m founder of BizStack at Cagri Sarigoz LLC and a passionate advocate for entrepreneurs.
With over 14 years in tech, marketing, and AI, including my role as Head of SEO at CitizenShipper and co-founder of TaleBot at Intale AI, I’m dedicated to sharing genuine, useful product insights and tips.
At BizStack, I aim to cut through the digital noise to provide clear, actionable advice.
And more than all else, I’m a father to a (always) little girl and a husband.
Contact me at [email protected] for assistance.
