saas.unbound is a podcast for and about founders who are working on scaling inspiring products that people love, brought to you by https://saas.group/, a serial acquirer of B2B SaaS companies.

In episode #40 of season 5, Anna Nadeina talks with Joran, founder of Reditus, affiliate management software tailored for SaaS businesses.

In this episode of saas.unbound I sat down with Benjamin Spinola, co-founder of SlidesGPT, to unpack how a tiny, product-first team built a profitable AI-powered presentation generator and carved out real traction in a crowded market. Benjamin’s journey—from mechanical engineering and the German manufacturing mindset to running a lean SaaS that automates PowerPoint creation—offers a clear, practical playbook for founders who want to win with AI without a large headcount or massive marketing budgets.

From engineering to AI product: the origin story

Benjamin started in mechanical engineering, captivated by German car manufacturing, before moving into marketing and then tech. After roles in a LA startup and as a CTO, he returned to Germany and co-founded SlidesGPT. The idea is deceptively simple: remove the tedious parts of building presentations so presenters can focus on the message, strategy, and storytelling.

SlidesGPT uses generative AI to create slide decks, charts, images and layouts in minutes. The goal isn’t to replace presenters; it’s to eliminate formatting, alignment and time-sink problems so users can focus on the content that matters.

Feature vs. product: should founders be afraid of being “wiped out”?

One of the recurring fears among niche AI founders is that a bigger company or ChatGPT itself could replicate your feature overnight. Benjamin’s perspective is frank and pragmatic:

“This happens not just for small startups—this also happens for bigger companies. When you look at… there are companies that just disappear because they are not relevant anymore.”

SlidesGPT’s defense isn’t magic IP; it’s diversification and depth. They operate with two legs—B2C and B2B—and see long-term value in B2B where power users need richer workflows, integrations, and data-driven slides. The strategy: double down on verticalized, enterprise-focused features that generalist models won’t provide out-of-the-box.

B2C vs B2B: different users, different expectations

  • B2C users: mostly students and teachers who want a presentation fast—five-minute workflows, simple UX and low pricing.
  • B2B users: heavy users who need accurate graphs, branded templates, and connections to company data—this is where premium features and higher ARPU live.

SlidesGPT started B2C but is shifting more toward B2B because that’s where the stickier, revenue-heavy use cases are. Benjamin believes OpenAI and other platform players will expand into workplace tools, but there’s still ample space for specialized SaaS solutions that match company-specific needs.

Naming, positioning and the SEO tailwind

SlidesGPT’s name (and the inclusion of “GPT”) was a deliberate growth lever early on. It boosted discoverability in Google searches and in the ChatGPT app store, creating a steady stream of trial users. But Benjamin acknowledges the branding trade-offs: a descriptive name can be hard to trademark and may create confusion about affiliation with OpenAI.

Key growth channels that worked for SlidesGPT:

  • GPT store and ChatGPT integrations (high discoverability)
  • Organic SEO (strong rankings for slide generator, PPT generator keywords)
  • Microsoft store (growing importance for B2B reach)

Paid acquisition was less attractive: competitors with large funding drove up CAC on platforms like Meta and Google, so SlidesGPT emphasized organic channels and product-led growth.

How a tiny team ships a lot: AI-first operations

SlidesGPT runs with two core engineers and a few freelancers. They scale output by adopting an “AI-first” approach across the company:

  • AI-assisted development tools (Cursor, GitHub Cloud) to speed up coding and PR generation.
  • Automated customer experience flows and proactive product fixes to reduce repetitive support demands.
  • Prioritizing automation so one person can manage customer support in roughly ten minutes a day.

Benjamin’s view on engineering talent is blunt: junior roles are getting compressed because AI can handle many routine tasks. The future center of value is senior architects who can orchestrate AI-generated code safely and avoid building long-term technical debt.

Customer-driven roadmap (but keep a vision)

Most of SlidesGPT’s short-term roadmap is informed by direct user feedback from multiple channels: ChatGPT ratings, cancellation feedback via Stripe, in-app tools like Churnkey, and support tickets. That gives a reliable signal for which features reduce churn and increase retention.

But Benjamin warns that product direction can’t be only reactive: “This is not like a vision. For this, you still need to understand the business you’re in… and you have to have a clear vision.”

For B2B especially, requests can be explicit (“build this or we leave”), forcing a trade-off between bespoke work and scalable features that benefit multiple customers.

Dealing with reviews, model updates and hallucinations

Marketplace rankings (GPT store, app stores) are influenced heavily by ratings, but Benjamin focuses on fixing the root causes in the product rather than firefighting each bad review. One important operational reality: changes in underlying LLMs (OpenAI rolling out model changes) can impact performance suddenly. SlidesGPT treats these as engineering incidents—investigate, patch, then monitor—rather than blame the user-facing symptom.

On hallucinations, the stance is honest: AIs hallucinate, and SlidesGPT avoids high-risk domains (medical, legal, military). Users are learning to expect imperfect output, and as base models improve, downstream products will benefit. Meanwhile the team focuses on improving prompts, filters and UX to minimize glaring errors.

Wins, regrets and what’s next

Wins:

  • Reaching seven-digit ARR with a tiny, profitable team.
  • Exceptional organic growth and SEO performance despite limited marketing spend.
  • Low CAC due to frictionless, no-login-first user experience.

Regrets and learning points:

  • Being too conservative early on—less risk-taking may have limited faster scaling.
  • Underinvesting in HR—Benjamin emphasizes that most problems boil down to people, and better hiring would have multiplied impact.
  • Not building stronger investor relationships while they were in a strong position (future negotiation leverage matters).

Founder hack: adopt an AI-first mindset and reduce friction

Benjamin’s practical advice for founders:

  1. Think AI-first: whenever you have a problem or task, ask whether AI can solve it or accelerate the solution.
  2. Keep barriers to value extremely low—SlidesGPT’s no-login-first approach drove fast user adoption and better SEO signals.
  3. Focus on product fundamentals. Fix root causes, not symptoms. A great product compounds over time—marketing can amplify it, but it can’t replace it.

Closing thoughts

SlidesGPT shows that you don’t need a massive team or deep pockets to build a thriving SaaS in the AI era. The combination of product-led growth, lean operations, prioritizing B2B depth, and leveraging AI tooling for both development and operations creates disproportionate leverage.

If you’re building in the AI space, think hard about where specialization beats general platforms, keep experimentation fast, and don’t underestimate the power of removing friction for your users.

Connect with Benjamin on LinkedIn: https://www.linkedin.com/in/benjamin-spinola-691250a8/

Try SlidesGPT: https://slidesgpt.com

Head of Growth, saas.group