Cookies managing
Cookie Settings
Cookies necessary for the correct operation of the site are always enabled.
Other cookies are configurable.
Other cookies are configurable.
Trusted by Industry Leaders
While running SEO for 60+ tech brands, we created a 4-direction method that shows correlation with organic growth both in Google and LLMs.
LLMs bring not just visibility, but users and revenue
To decode how AI engines evaluate and cite software brands we:
- reverse-engineered thousands of LLM queries
- analysed 300+ sources from the leading AI SEO researchers and data providers
- gained practical insights from our 60+ SaaS, App and Tech projects.
That research formed the foundation of our 4-direction GEO framework, which has already turned organic AI citations into thousands of monthly sign-ups and driven up to 20-30% of total ARR (Rask.AI case study)
After tracking 12 months of Rask AI data to find the four levers driving their traffic spikes, we turned them into a GEO framework now proven across 15+ tech products.
Our ESA 4-Direction GEO Framework
1
Brand Authority (Brand SEO, GEO)
LLMs don't recommend brands based on backlinks. They evaluate your digital footprint across trusted third-party sources. We inject your product into the ecosystem AI models scrape to dominate buyer consideration sets.
What you get:
- Buyers find you instantly: Dominate high-intent lists ("Best [Category] Tools", "Alternatives to [Leader]") with 20–30 strategic placements every quarter.
- LLMs naturally recommend you: Build unlinked brand mentions across the exact sources AI models trust to form their answers.
- Steal competitor market share: Force AI engines to put your brand in the same consideration set as the category leaders.
- Establish permanent authority: Publish proprietary research that journalists quote and LLMs scrape as absolute industry truth.
2
Topical Authority (Semantic SEO, AEO)
AI search engines recommend entities, not keywords. We establish your domain as the primary source in your category's Knowledge Graph through structured, high-density coverage.
What you get:
- Market demand mapping: A content strategy aligned with live query data from ChatGPT, Perplexity, and AI Overviews.
- Structured topical authority: Pillar pages connected to focused 3–7 article clusters that establish clear depth.
- Optimized site architecture: An audit and removal of low-value, irrelevant content to strengthen your domain's semantic footprint.
3
Free Tools (Programmatic SEO)
Giving away part of your product's functionality for free as top-of-funnel entry. It's how Canva scaled with template pages, Wise with currency converters, Lovable with component galleries, and Webflow with template marketplaces. Each utility becomes a landing page that captures informational traffic and funnels users into your core product.
What you get:
- Job-to-be-Done Mapping: Identification of the 5–10 most frequent jobs your users have around your core product.
- Functional Free Utilities: Built-for-purpose calculators, generators, converters, quizzes, or templates.
- Conversion-Optimized Landing Pages: Every tool page structured with an H1, problem framing, the live tool, results, and a direct path to your main product.
4
Brand Authority (Brand SEO)
LLMs and Google love fresh content, not new, but recently updated. Stanford NLP papers from 2024-2025 and independent SEO experiments converge on the same finding: if a piece of content has been updated within the last 30 days, the probability of being cited in an AI answer goes up by roughly 25%.
What you get:
- Automated 30-Day Update Cycle: Regular optimization applied to 5–10% of your key pages every month.
- Fresh Data and Signals: Updated statistics, current examples, and modified timestamps across your schema, sitemaps, and visible page dates.
- Persistent AI Visibility: Consistent citation retention in AI answers, preventing the citation decay competitors experience within months.
How the 4 directions
work together
work together
None of these four directions on its own delivers a systemic result. Topical authority alone, without brand mentions, gets out-ranked. Brand mentions alone, without topical coverage, don't get cited (nothing to take from). Programmatic without refresh disappears in 6-9 months. Refresh alone, without topical coverage, has nothing to refresh. All four work as a single flywheel that compounds with every cycle.
Who We Work With
We work with two types of companies,
funded startups and scale-ups at Seed to Series B stage:
funded startups and scale-ups at Seed to Series B stage:
Apps
Mobile and web apps with strong indexation opportunities through ChatGPT, app store surfaces and help centre content
SaaS Brands
B2B and B2C SaaS across CRM, marketing, HR, dev tools, analytics, localisation and more, including AI-native products built on top of LLMs
Apps have one edge we've been actively using: they get strong indexation through ChatGPT, app stores and their help centres. That's an insight we found and now build into every app engagement. The core ESA 4-Direction GEO Framework, Topical Authority, Brand Authority, Programmatic SEO and Content Refresh, works for both categories
Traditional SEO vs Generative Engine Optimization (GEO)
Your Journey to AI Search Visibility
It doesn't start with content or links. It starts with seeing the picture in numbers. Here's our 4-step process:
Step 1
SEO & GEO ROI Forecast (Week 1)
SEO & GEO ROI Forecast (Week 1)
The starting point. We analyse your own visibility (branded vs non-branded), benchmark 5+ competitors on their branded search, LLM visibility and LLM-to-Google ratio, map content opportunities across topical clusters, and build a 1-2-3-year ROI projection with three package tiers.
→ You leave with a full picture of where you stand and what growth is realistic.
→ You leave with a full picture of where you stand and what growth is realistic.
Step 2
Setup / Foundation (Month 1)
Setup / Foundation (Month 1)
Once you approve the direction, we execute all technical and strategic setup: technical foundation (schema, LLMs.txt, bot access), Topical map build, first Brand SEO placements, Programmatic SEO planning, page refresh cadence setup.
→ By end of month 1, the machine is ready to run.
→ By end of month 1, the machine is ready to run.
Step 3
Monthly Paid Delivery
(Month 3-6)
Monthly Paid Delivery
(Month 3-6)
The full ESA 4-Direction GEO Framework running in parallel every month. New pages, new placements, refresh cycles, prompt tracking, competitor monitoring. Ownership split with your team is clear: what we deliver, what your team owns. Monthly reporting with LLM visibility, citations, and revenue attribution.
→ Tests with the team velocity and growth speed.
→ Tests with the team velocity and growth speed.
Step 4
Scaling & Compounding
(Month 6+)
Scaling & Compounding
(Month 6+)
The system runs on its own, every new page improves everything else, every new brand mention lifts existing rankings. We expand into new markets and languages, optimise the LLM-to-Google ratio toward 1:2 (benchmark for mature systems), and pull new AI search engines into your visibility as they launch.
→ Scalable growth with users and revenue month over month.
→ Scalable growth with users and revenue month over month.
Results from Our GEO Services
Rask AI,
AI Video Translation SaaS
❻ 1M+ organic users, $1M ARR in 149 days
❺ 5x LLM traffic growth, 12,100 LLM visits/month
❹ SEO + LLM = 70% of all signups
Enterprise CRM,
Top 10 Global Category
❻ 5.9X Google AI Overviews brand share in 4 months
❺ 20,000 AI visits/month, 100,000 monthly impressions
❹ Top 8 recommended CRMs in target clusters
❺ 20,000 AI visits/month, 100,000 monthly impressions
❹ Top 8 recommended CRMs in target clusters
Noji,
Mobile App for Flashcards & Language Learning
❻ 4x organic traffic, 17x registrations per month
❺ Web channel = 20% of $4M ARR
❹ Exit to DelightRoom, May 2026
Common Questions About Generative Engine Optimisation
A: Almost every time 'it didn't work' means the old approach. Teams mass-produced content, usually AI content, didn't work on Brand SEO, bought links, and now they're wondering why everything is sliding. This isn't 'SEO doesn't work', it's a specific set of old-school mistakes. Modern GEO needs all 4 directions running in parallel: topical authority, brand mentions, programmatic pages and 30-day refresh.
A: It depends on the starting point. If your product already has decent brand authority, we start with Programmatic SEO pages, they index very fast (ChatGPT crawls almost daily). With the right approach, you can see the first result within a week. We had a case with 3.5x LLM traffic growth in just 3 months using Programmatic + Brand SEO combination.
A: You don't need all four at once. After the ROI Forecast, it's always clear where the biggest potential lies for your specific product, that's where you start. For a brand-new product, we usually start with Brand SEO (budget-friendly tactics: guerrilla PR, Reddit, Quora, YouTube). Once that's moving, we layer in Semantic SEO. Programmatic SEO is a maximum-boost lever once the foundation is in place. Refresh is insurance.
A: The surface changes: interfaces, features, SERP format. The foundation doesn't change: entities, topical authority, brand mentions, behavioural signals, freshness. That's what Google and every modern LLM are built on, and every new AI search engine will be built on. Strategy is built around the foundation, not around specific AI search features that might change next quarter.
A: Traditional SEO optimises for search result rankings (getting to position #1 on Google). GEO optimises for citations (getting your brand named as a source when ChatGPT, Claude, Gemini, AI Mode or AI Overviews answer a user's question). The core difference is entities vs keywords: AI systems work with entities in the Knowledge Graph and connections between them, not with keyword clusters.
A: All major AI search engines: ChatGPT (OpenAI), Claude (Anthropic), Gemini and AI Overviews (Google), Perplexity, Microsoft Copilot, AI Mode. Each has different citation logic, and we track visibility across all of them separately, with a target LLM-to-Google ratio of 1:2 for mature systems.
A: First AI citations typically appear within 6-12 weeks after we deploy the technical foundation and start LLM seeding. Full framework impact (5X+ growth in AI Overviews brand share, cited across all major LLMs) takes 4-6 months, based on our client data.
Working with ESA Digital has been fantastic! They bring deep industry knowledge and used it to build a strategy that delivered exceptional outcomes for our mobile app and our wider expansion. Their data-driven approach and excellent client service mean they consistently go above and beyond to meet our specific requirements. We've achieved so much together and couldn't be happier with the results!
Andrew Bondarenko
CPO & Co-founder, Noji (Exit, DelightRoom)
CPO & Co-founder, Noji (Exit, DelightRoom)
“I met Dennis and his Google team, when they pitched their startup Smartsub. We accepted them to our accelerator program in 2017, Batch 9 and they successfully passed through it. Within a short period of time they managed to develop MVP of their product, design and implement go-to-market strategy launching the product on the US market. Their broad digital business experience and data-driven approach allowed them to make things done.”
Cristobal Alonso (www.startupwiseguys.com)
CEO at SWG (funded Smartsub in 2017)
CEO at SWG (funded Smartsub in 2017)
ESA team exceeded my expectations while working on several of our key projects. They delivered 2X organic traffic growth across 12 websites by coordinating technical SEO, content and link-building teams. They improved websites architecture, navigation, technical SEO and implementing conversion rate optimisation.”
Maria Kristalinskaya
Head of Retail Media at eBay (GE), Techstars Startup Mentor, ex-BBDO
Head of Retail Media at eBay (GE), Techstars Startup Mentor, ex-BBDO