Something shifted in how people find information in 2024 and 2025. A growing share of searches — especially informational queries — now end with a generated answer from an AI engine rather than a click to a website. ChatGPT, Perplexity, Google's AI Overviews, and Bing Copilot are pulling information from across the web and synthesising answers, citing sources in the process.
The businesses that appear in those citations get visibility, credibility, and traffic at a new level. The ones that don't are invisible to a growing portion of queries. This is Answer Engine Optimisation (AEO), and it requires a slightly different approach than traditional SEO.
What AI engines actually look for
When an AI engine generates an answer, it's pulling from sources it considers authoritative, well-structured, and clearly topically relevant. Several factors influence whether your site gets cited:
- Structured data (schema markup) — JSON-LD markup in your HTML that explicitly tells machines what your content is, who created it, what it's about, and what entity it describes. This is the single most direct signal you can send to AI engines.
- Clear, direct answers — content that answers a specific question in the first 100 words, then expands, is more likely to be pulled than content that buries the answer five paragraphs in.
- Entity definition — AI engines build knowledge graphs. If your website clearly defines your organisation (name, location, founder, services, contact details) in machine-readable format, you become a known entity rather than an anonymous web page.
- Authoritativeness — external citations (links and mentions from credible sources), consistent information across the web, and a demonstrable track record all contribute to how AI engines weight your content.
The most important schema types for Kenyan businesses
LocalBusiness / ProfessionalService — the entity definition for your organisation. Includes your name, address, telephone, geo coordinates, opening hours, area served, and services offered. This is what gets you cited when someone asks "who is the best web agency in Nairobi?"
FAQPage — a list of questions and answers about your business or topic. This is one of the highest-value schema types for AEO because it's literally the format AI engines use to synthesise answers. Every FAQ section on your site should have this markup.
HowTo — step-by-step process schemas. When someone asks "how do I integrate M-Pesa into my website?", a properly marked-up HowTo article has a strong chance of being cited.
Article — for blog content, including author markup (linking to the author's entity), publisher, publication date, and word count. This helps AI engines assess content freshness and credibility.
The double-layer approach
The most effective AEO implementation uses two layers of structured data simultaneously: JSON-LD in the <head> (for machine parsing) and microdata attributes (itemscope, itemprop) on the visible HTML elements. This redundancy means both crawler types — those that parse JSON-LD and those that read inline microdata — extract your information correctly.
Content strategy for AI citation
Every piece of content you publish should be written with a specific question in mind. What question is this answering? Write the direct answer in the opening paragraph. Then support, expand, and contextualise. Structure with clear headings that are themselves questions or answer fragments. Use lists for multi-part answers.
At Frank Dex Devs, we implement comprehensive schema markup as a standard part of every website build — covering LocalBusiness, FAQPage, BreadcrumbList, and Article types across the full site. If your current site lacks this, our SEO service includes a full structured data audit and implementation.
Discussion (0)
No comments yet!