After 15+ years in quality engineering — eight of them at Accenture, then leading QA and building validation frameworks for AI systems — Somya is a first-time founder building Stellarcast in the open from India. In this conversation with host Jeremy Rivera, she breaks down why AI citations are a different game from search rankings, why "AI is confidently wrong" about so many brands, and how a continuous monitor-diagnose-execute-prove loop beats the one-time audit.

In this episode

• The monitor → diagnose → execute → prove flywheel — and why most tools "stop at the dashboard"
• Why citations aren't rankings: AI answers overlap Google's top ten by only ~14%
• AEO in plain English — make your brand easy for AI to find, understand, and read
• Why "AI is confidently wrong", citing two-year-old pricing and reviews even after fresh data lands
• Treating LLM answers as a "liquid surface" — freshness, lag, and query-deserves-freshness in AI
• CMS integrations (WordPress, Shopify, Webflow) that turn diagnosed gaps into approved, shipped fixes
• Somya's leap from 15+ years in QA to first-time founder — and the exposure of shipping under your own name
• Market fit for SMBs and agencies, with the human always in the loop

Chapters

• (00:00) Meet Somya Goyal & what Stellarcast does — a "system of AI records"
• (02:30) The four-agent loop: monitor, diagnose, execute, prove
• (05:45) CMS integrations & building in the open with pilot design partners
• (08:10) How the models behave differently — and why citations aren't rankings (~14% overlap)
• (12:20) Freshness, lag, and the "liquid surface" of LLM answers
• (16:05) AI as your least-trained support rep — the rogue knowledge base
• (19:15) Somya's background: fifteen years in QA to first-time founder
• (22:40) Market fit: SMBs, agencies, and the human-in-the-loop
• (25:30) Early access & pricing

Notable quotes

"Visibility in AI is simple. Make your brand easy for AI to find, understand, and read. That is the new name of AEO, in very simple English."
— Somya Goyal"Ranking is different and citation is different. AI citations overlap Google's top-ten links by only fourteen percent."
— Somya Goyal"Most brands have never read what AI says about them, and sometimes AI is confidently wrong. It cites data that's maybe two years old — the pricing, the reviews — even after a lot of fresh data."
— Somya Goyal

Resources & links

• Stellarcast — AI brand-visibility platform (monitor, diagnose, execute, prove)
• Somya Goyal on LinkedIn — where she's building in the open
• When AI is "confidently wrong" the fix is filling the gaps with human-certified, expert-sourced content
• Monitoring your brand's AI footprint and entities is exactly where SEO advisory earns its keep
• Jeremy's fastest AI-citation win came from local link building through community cleanup events

About the show

Unscripted SaaS is hosted by Jeremy Rivera. Catch more episodes at unscriptedsaas.com

Key takeaways

• The differentiator is what incumbents treat as an edge case. Not accuracy, but a product decision to drop what it does not recognise.
• Africa first is a constraint, not a slogan. Limited funding cannot cover many languages at once, so splitting attention means doing none of them well.
• Self-taught to shipping in roughly two years. YouTube, a first paid client, then a role where being the only developer forced the pace.
• He refuses features users ask for. Explicitly so the product never becomes a jack of all trades.
• First version shipped in about a week. The scope was small because the problem was his own.

On this page

• The gap the incumbents left open
• Africa first, and why it is a constraint
• Self-taught from Lagos
• Refusing the features people ask for
• Chapters and timestamps
• People, ideas and sources mentioned
• Questions this episode answers
• Go deeper

The gap the incumbents left open

Tangible things were said and their machine did not capture that. And I thought that since I am a software developer, maybe I could build something that would solve this problem.

— Tobiloba Sulaimon

Mainstream tools filter unrecognised languages out, or map them onto the nearest English word. The output looks like a transcription failure. It is closer to a design choice, and it is the kind of choice a large company can make for years without anyone internally noticing.

That is the general lesson worth taking from this episode even if you never touch transcription. Find the thing your category's incumbents have classified as an acceptable failure.

Africa first, and why it is a constraint

We are African first. We are focusing on ourselves first, so I’m trying to solve this problem within Africa first before we move on.

— Tobiloba Sulaimon

This is not positioning. Limited funding cannot cover the continent's language count, so attempting breadth would mean doing all of it badly. The narrowness is what makes the quality possible, and the quality is the entire product.

Self-taught from Lagos

Everything I know today I learnt on my own. I am a self-taught software developer.

— Tobiloba Sulaimon

HTML, CSS and JavaScript from YouTube after high school. A first paid client build. Then a front-end role at a game studio where being the only developer meant learning fast or shipping nothing. The first version of AuTrans took about a week.

Refusing the features people ask for

With a team and outside consulting, he now rolls features out step by step and turns down most requests, so AuTrans does not become a jack of all trades. Jeremy anchors the same lesson in Raven Tools, which expanded horizontally until it was competing with everyone and beating no one.

Around one hundred users and the first paying customers at the time of recording, with mobile apps, more languages and API access on the roadmap.

Chapters

People, ideas and sources mentioned

Questions this episode answers

• Why do mainstream transcription tools fail on African languages?
• Should I focus on one market before expanding internationally?
• How do I avoid scope creep when users keep requesting features?
• Can you become a working developer self-taught, and how long does it take?
• How do I find a differentiator an incumbent will not simply copy?
• How long should a first version take to build?
• How do I decide which feature requests to refuse?

Go deeper

Every link above goes somewhere different. These are the ones not already mentioned above.

• Malith Gamage on why survival is the whole strategy, arguing the same narrowness from a bootstrapped European position.
• Tom Girgash on building for a group you actually belong to, which is the founder-market-fit version of the same instinct.
• The machine-readable version of this episode, carrying the complete transcript with speaker labels on every turn. Written to be pasted into an LLM.

Find Tobiloba Sulaimon

autrans.online

Cite this episode

Sulaimon, Tobiloba. Interviewed by Jeremy Rivera. “Africa First.” Unscripted SaaS, 2026. https://unscriptedsaas.com/africa-first-tobiloba-sulaimon-on-focus/

Key takeaways

• Survival is the strategy, not a precondition for one. No in-house hires, infrastructure scaled only when needed, and a year of AWS credits used as a cash-flow cushion.
• White labelling sits on every plan, not the top tier. Agencies have to look more put-together than a client’s internal marketing team, so branding is the product rather than an upsell.
• He ships AI visibility reporting and openly doubts it. Ask the same question five times and get five answers pulled from five places.
• Domain authority did not predict who got cited. His own unpromoted, zero-authority domain was the source models pulled from, while the promoted one was not.
• A lifetime deal run as a feedback engine. The value was buyers who kept sending feature requests, not the one-off revenue.

On this page

• Do not die
• Your AI is only as smart as your schema
• Shipping a feature he does not fully trust
• White labelling, and a lifetime deal as a feedback engine
• Chapters and timestamps
• People, ideas and sources mentioned
• Questions this episode answers
• Go deeper

Do not die

Some main advice is don’t die. Like as long as you can stay for like three, four years alive, then you’re good.

— Malith Gamage

Unromantic, and more useful than most strategy. The failure mode for a bootstrapped tool is almost never being out-competed on features. It is running out of runway during the years when nothing is visibly working.

So the discipline is structural: no in-house hires, infrastructure scaled only when demand actually arrives, and a year of cloud credits treated as a cash-flow buffer rather than free compute to burn.

Your AI is only as smart as your schema

Your AI is smart as your schema. As long as you give the correct data and the correct description, then you will get good answers back.

— Malith Gamage

He draws the parallel to MCP without prompting, and it is the reason he pipes context in before asking anything: pull the Analytics data first, describe it properly, then ask the question. The quality ceiling is set at the data-description step, not at the model.

Shipping a feature he does not fully trust

I don’t even know sometimes if you ask the same question five times you get five different answers they pull out from five different places.

— Malith Gamage

Zapdigits aggregates brand visibility across several models and averages it into a score for agency clients. He will tell you the underlying signal is unstable while the feature is live, which is rarer than it should be.

His sharpest example is his own: a throwaway landing-page domain with no authority and no promotion behind it was what the models surfaced when asked about his brand. The domain he had actually promoted was not.

White labelling, and a lifetime deal as a feedback engine

Competitors gate white labelling behind the most expensive plan. Zapdigits puts it on every plan, and the reasoning is about who the customer is rather than about pricing psychology. An agency has to look more together than the client’s internal marketing team, so its own branding on the dashboard is not a premium feature, it is the thing being sold.

The lifetime deal was run the same way. Buyers came in through live webinars and stayed in a private group, on a public roadmap where they add feature requests directly. Many bought partly to back a founder who was starting out, and then kept sending ideas.

Which is the actual return: people who tell you something useful, rather than people who tell you it was good but too expensive.

Chapters

People, ideas and sources mentioned

Questions this episode answers

• What is the single most important thing for an early-stage SaaS?
• Should I bootstrap or raise venture capital?
• Is AI visibility reporting reliable yet?
• Why does domain authority not seem to matter for AI citations?
• Should white labelling be a premium feature?
• How do I get useful product feedback rather than complaints about price?
• How do I keep infrastructure costs down as a two-person team?

Go deeper

Every link above goes somewhere different. These are the ones not already mentioned above.

• the same narrowness argued from Lagos, where the funding constraint is what makes solving one market properly possible at all.
• building a tool for the operator you already are, which is founder-market fit doing the job a roadmap usually does.
• the full transcript of this episode in machine-readable form, speaker-labelled on every turn and written to be pasted into an assistant.

Find Malith Gamage

zapdigits.com.

Cite this episode

Gamage, Malith. Interviewed by Jeremy Rivera. “Do Not Die: On SaaS Survival.” Unscripted SaaS, 4 August 2026. https://unscriptedsaas.com/do-not-die-malith-gamage-on-saas-survival/

Kevin Urrutia is a software engineer turned serial entrepreneur with 20+ years of experience building tech products. He's worked at Silicon Valley companies including Mint.com and Zaarly, and founded Voy Media, a digital marketing agency that has generated over $50 million in revenue. His latest venture, Magic Rinku, tackles one of SEO's most persistent challenges: automating internal link building at scale.

Connect with Kevin:

• Email: kevin@magicrinku.com
• Twitter: @danest
• Website: Magic Rinku

Episode Summary

In this unscripted conversation, Kevin shares the journey of building Magic Rinku from personal frustration to SaaS solution. We dive deep into the technical challenges of WordPress integration, the strategic use of AI versus traditional algorithms, and unconventional marketing approaches that are actually working in 2025.

Kevin reveals how he discovered 350 out of 400 articles on his own site had zero internal links, the technical nightmare of supporting multiple WordPress page builders, and why cold email and Reddit still outperform traditional SaaS marketing for developer tools.

Key Topics Covered

The Genesis Story

• How Kevin's personal pain with internal linking across 8-10 affiliate sites led to Magic Rinku
• The validation process through Reddit and direct customer conversations
• Why most SEO professionals are missing huge linking opportunities due to broken manual processes

Technical Deep Dive

• WordPress ecosystem challenges: Supporting Elementor, Beaver Builder, Gutenberg, and Classic Editor
• The 30-second delay problem and why bulk operations are complex in WordPress
• Database structure differences across page builders and how they affect plugin development
• Error handling and retry systems for WordPress API integration

AI Implementation Philosophy

• Strategic use of AI vs. traditional algorithms (RAKE, ENG tagger)
• Two specific AI use cases: contextual understanding and smart sentence generation
• Why 30% of AI suggestions are perfect, 50% need tweaking, and 20% are unusable
• The explainability problem and showing confidence scores to users

Advanced SEO Features

• Silo construction from both technical and strategic perspectives
• Tree rebalancing algorithms for hierarchical site structures
• Why circular silos were pulled from production despite working code
• The decision framework for shipping features vs. keeping them in development

Unconventional Marketing Strategy

• Cold email targeting WordPress agencies 
• Reddit as a B2B channel for finding people actively complaining about the problem
• Why traditional SaaS marketing (content, paid ads) doesn't work for unknown problem categories
• The two-email rule and why persistence beyond that is counterproductive

Company Philosophy

• Intentionally staying at 2-3 person team size
• Lifestyle business approach vs. venture scale ambitions
• The hiring challenge: finding people who understand both WordPress and SEO
• Using contractors vs. full-time employees at current scale

Quotable Moments

"Even for someone like me who knows SEO, who's been doing this for years, I had this massive blind spot. When I ran my prototype, it found 350 articles with zero internal links out of 400 total articles."

"Our biggest competitor is Excel spreadsheets and manual processes. There are other tools, but most are part of $200-500/month SEO suites. We do one thing really well."

"My rule is pretty simple - use AI for problems that require understanding context and nuance, use traditional algorithms for problems that have clear mathematical solutions."

"The freedom is the best part of being a founder, but also the worst part because you have infinite options every day."

"If a feature works for 80% of use cases and doesn't break anything, I ship it. I can iterate based on real user feedback rather than trying to anticipate every edge case."

Resources Mentioned

Tools & Platforms

• Magic Rinku - AI-powered internal link building
• WordPress page builders: Elementor, Beaver Builder, Gutenberg, Divi
• Convert SEO - for finding WordPress agencies
• Screaming Frog - SEO crawling tool
• Ahrefs - SEO platform mentioned for potential integration

Books & Learning Resources

• "The Mom Test" by Rob Fitzpatrick - customer development
• SEO blogs and communities for understanding user problems
• Reddit communities for market research and validation

Marketing Channels

• Cold email outreach to WordPress agencies
• Reddit for finding users with the specific problem
• Word of mouth within the SEO community
• Organic search for "WordPress internal linking plugin"

Key Takeaways for SaaS Founders

Product Development

1. Build something you actually use daily - Customer development is easier when you are the customer
2. Ship at 80% completeness - Iterate based on real feedback rather than anticipating edge cases
3. Use AI strategically - Apply it for context/nuance problems, traditional algorithms for mathematical solutions
4. WordPress is complex but necessary - 40% market share makes technical challenges worthwhile

Marketing & Sales

1. Go where people complain about problems - Don't wait for them to search for solutions
2. Cold email still works - When targeting the right problem with the right audience
3. Two-email maximum rule - More persistence often backfires
4. Reddit can be B2B - Focus on DMs to people explicitly stating problems

Business Strategy

1. Focus beats feature creep - Most SaaS tools fail from doing too much, not too little
2. Lifestyle business is valid - Don't scale into a job you hate
3. Small teams move faster - Less meeting overhead, closer to customers

Technical Decisions

1. Platform integration vs. standalone - User experience often trumps technical simplicity
2. Error handling is crucial - WordPress ecosystem requires robust retry systems
3. Multiple complexity levels - Serve both power users and beginners in the same tool
4. Documentation for key person risk - Critical when running lean teams

Angshuman shares insights on customer needs, market trends, and the importance of sales feedback in shaping product development. He's been heading up the creation and product management of their new TapDataMax product, and shared a lot of detail on the way Tapclicks views the "movement" of data. The discussion also touches on the evolving terminology in marketing technology, the role of AI in data management, and the future roadmap for TapClicks, including new features and integrations.

• Tool Desk started with a focus on home services and lead qualification.
• The initial product was inspired by a need for ringless voicemail systems.
• Pricing is strategically set at $149/month for the Jobber plug-in.
• Customer support is highly personalized, with direct communication from founders.
• The company is 100% bootstrapped, allowing for thoughtful financial decisions.
• Cold outbound marketing has been effective for customer acquisition.
• Building a product that fills gaps in existing platforms is crucial.
• Platform risk is a significant concern for businesses relying on third-party services.
• Future-proofing involves diversifying product offerings and reducing dependency on a single platform.
• The founders prioritize building something people want and scaling thoughtfully. The value proposition of Tool Desk AI is to fill marketing automation gaps.
• Platform dependency can pose significant risks for SaaS businesses.
• Tool Desk AI acts as a 24/7 sales representative for capturing leads.
• AI technology is being utilized to enhance lead qualification processes.
• Onboarding for home service businesses is designed to be frictionless and quick.
• The AI is trained on each customer's website to provide tailored interactions.
• Even if you're doing B2B, like saving commercial kitchens on cost, customer feedback is crucial for success.
• Marketing strategies include cold outreach and SEO optimization.
• Building a SaaS product requires a strong understanding of coding or a capable partner.
• Entrepreneurs should only pursue SaaS if they have a clear market need and potential customer interest

• Vooz aims to innovate in the anonymous video chat space with AI moderation.
• User engagement features like screen sharing and games are key to Vooz's strategy.
• Bootstrapping has been essential for Vooz's development and team formation.
• Finding motivated co-founders can be challenging but crucial for success.
• SEO and PPC are primary marketing strategies for user acquisition.
• Understanding how pricing strategy impacts SaaS marketing decisions
• Influencer partnerships can drive initial traffic but should be complemented by SEO efforts.
• Monetization opportunities are abundant once a user base is established.
• Starting with design helps in attracting talent and building a vision.
• AI moderation is a significant technical achievement for Vooz.
• Taking action and starting somewhere is vital for aspiring entrepreneurs.

• Mark Williams-Cook: Director of Kanda (formerly Candour SEO agency)
• Creator of Also Asked, a search intent research tool
• Publisher of Core Updates Newsletter
• Host of SEO Patent Podcast
• 20+ years of SEO experience

Key Topics Covered

Current Google Search "Meta"

• Google's algorithm favors brand-like signals over traditional relevance
• Example: The Vatican website ranking for CBD gummies after being hacked
• Mark describes this as "a bandaid for the shortfalls that have been magnified in Google's algorithm because of Gen.ai"

Content Creation Strategy

• Video-first approach: "Start at video and work backwards" using transcription tools
• Particularly effective for client information: "If you are having to get information, especially from clients, sometimes trying to get clients to write an article is like getting blood out of a stone"
• Allows natural expertise to flow: "They will happily talk your ear off"
• Tools turn video into written content without "regurgitating stuff"

User Knowledge Levels

• Google has patents for understanding user expertise through search patterns
• Demonstrated in People Also Ask (PAA) results showing different questions based on perceived knowledge
• Example: Car dealership knowledge levels
  • Novice: "Red car under $20k near me"
  • Intermediate: "2024 Honda Civic MPG vs Toyota Corolla"
  • Expert: "Civic Type R torque curve comparison"

Google Query Classification

• Google classifies queries into distinct semantic categories
• These classifications determine which SERP features appear
• Additional classifiers like "local intent" and "explicit local intent" work alongside these categories
• Mark is developing a tool to predict which category a query falls into

Evolution of SERP Features

• AI Overviews: Likely to eventually replace featured snippets
  • More dynamic and nuanced answers
  • Google reported 90% cost reduction in generating them
• Featured Snippets: Currently coexist with AI overviews
• People Also Ask (PAA): Critical for Google's query understanding
  • Mentioned in DOJ trial as part of Google's "induction loop"
  • Allow Google to understand different interpretations of the same query
  • Localization feature coming to Also Asked tool

The Future of Search

• Prediction: Shift from users doing research to AI agents doing it for them
• "I think we are going to have a fundamental shift of rather than people doing their own research..."
• Advantage to companies that own hardware (Google, Apple, Microsoft)
• Privacy concerns will diminish as convenience increases

SEO Tools Mentioned

• Also Asked: Mark's tool for analyzing PAA data
• Screaming Frog: For automated content analysis
• ChatGPT API: For content gap analysis
• Perplexity: AI search tool Mark recommends
• 11 Labs: Voice cloning technology (mentioned in context of potential misuse)
• Storm: Stanford's agentic model for scientific paper creation

Technical SEO Insights

• Google uses binary trust systems for many technical signals
• "If you get them all wrong, it's not like Google will trust you a bit less. They just have a very simple thing of, well, you don't seem to know what you're doing."
• Site reputation abuse handled manually rather than algorithmically
• Example of PageRank evolution: still used but significantly modified from original patent

People Referenced

• Mordy Oberstein: Mentioned regarding brand discussions
• Gary Illyes ("Garyish"): Google representative
• Danny Sullivan: Google Search Liaison
• Darren Shaw: Discussed query categories with Mark
• Arnoud Hellman: Provided examples of localized PAAs
• Bill Slawski ("Bill Flosky"): Known for patent analysis
• Olaf Kopp: Current patent analysis work
• Lily Ray: Mentioned regarding Reddit in SERPs
• Michael Buckby of Noah Toa
• Matt Brooks of SEOteric
• Commercial restaurant cost savings company:Save Fry Oil 

Resources

• Search Engine Land guide: Mark's guide on using Screaming Frog with Also Asked
• learningseo.io: Referenced for learning about SERP features

Live Test Examples

• "Jam sandwiches" query test showing AI overview replacing featured snippet
• Demonstration of entity analysis with non-existent URL showing AI hallucination issues

Upcoming Developments

• Query classification tool (planned release before Brighton SEO in April)
• Also Asked localization feature allowing city-specific PAA monitoring
• Location-specific search results using longitude/latitude data

SEO Conferences Mentioned

• Brighton SEO
• SERPConf

Contact Information

• LinkedIn: Mark Williams-Cook
• Bluesky: Mark Williams-Cook (previously on Twitter/X)
• Newsletter: CoreUpdates.com