Steps in this guide
Step 1
Measure your baseline visibility first
Before making any changes, establish baseline data: which queries trigger your brand mention in which AI assistants, at what position, and with what citations. Without this baseline, you cannot measure whether your AEO efforts are working.
Step 2
Optimize your entity presence on high-citation sites
AI assistants pull from review sites, directory listings, and trusted comparison sources. Claim and fully complete your profiles on G2, Capterra, Product Hunt, and your category-specific directories. More complete profiles with more reviews correlate with higher citation rates.
Step 3
Create content that directly answers category discovery queries
Write blog posts and landing pages that match the exact queries AI users ask about your category. A post titled 'Best tools for SaaS subscription billing in 2026' that includes your product is more likely to be cited than a generic homepage.
Step 4
Improve your structured data and llms.txt
Add schema markup (SoftwareApplication, Product, Organization) to your marketing pages. Create a /llms.txt file that summarizes your product for AI crawlers. Some AI assistants explicitly use these signals when generating recommendations.
Step 5
Build mentions from high-authority sources
AI training data heavily weights mentions from high-domain-authority sites. Getting covered on TechCrunch, a popular HackerNews post, a well-read developer newsletter, or a featured Product Hunt launch improves your citation probability across all LLMs.
Step 6
Address your specific competitive gaps
Use your citation data to find where competitors are being recommended and you are not. Target those specific query types with new content and outreach to the sites that are being cited. Competitive gap analysis is the most efficient AEO improvement activity.
Step 7
Monitor and iterate
AEO is not a one-time fix. Track your visibility weekly, attribute changes to specific actions, and double down on what moves the needle. The brands that win in AI search are the ones treating it as an ongoing channel, not a checklist.
Skip the manual version of this guide
AskRank runs this exact workflow automatically across ChatGPT, Claude, Perplexity, and Gemini, on a schedule.
Answer Engine Optimization (AEO) is the practice of improving how often and how favorably AI assistants recommend your product when users ask relevant questions. It is not a replacement for SEO - it is a parallel channel that is growing fast and that most SaaS founders are not tracking yet.
This guide covers the concrete steps you can take to improve your AEO for ChatGPT, Perplexity, Claude, and Gemini.
What actually influences AI recommendations
Before diving into tactics, it helps to understand what drives AI recommendations. AI assistants generate answers based on:
- Training data - the text they were trained on, which includes reviews, documentation, blog posts, and comparisons across the web
- Retrieval augmented generation (RAG) - for AI assistants like Perplexity that search the live web before answering
- Entity recognition - how clearly an AI can identify your product as a distinct entity with known attributes
Each of these influence channels has corresponding tactics.
Tactic 1: Establish your baseline first
Trying to improve AEO without baseline data is like trying to improve your conversion rate without knowing what it is today.
Before making any changes, run your target queries across the AI assistants you care about, record the results, and document the baseline. This takes 1-2 hours to do manually for 20 queries, or you can use AskRank to automate it from day one.
Everything else in this guide is only meaningful if you can compare to a baseline.
Tactic 2: Review site presence
G2, Capterra, and Product Hunt are among the most commonly cited sources when AI assistants recommend software tools. The pattern is consistent across ChatGPT, Perplexity, and Gemini: products with more reviews, better ratings, and more complete profiles on these platforms appear more frequently.
The minimum viable review site strategy:
- Claim your G2 and Capterra profiles (if you have not)
- Fill in all product attributes, features, and integrations completely
- Reach out to existing happy customers and ask for reviews - a personal email works better than automated review request sequences
- Aim for at least 15-20 reviews on G2 as a baseline; 50+ puts you in a much better position
Product Hunt is a separate priority. A strong Product Hunt launch (top 5 of the day) creates a high-authority citation that all major LLMs reference when recommending tools in your category. If you have not launched on Product Hunt, plan one.
Tactic 3: Create category-specific comparison content
When someone asks ChatGPT “what is the best tool for X,” the AI frequently pulls from articles that explicitly compare tools for that use case. If there is a well-written, comprehensive comparison post that includes your product alongside the top alternatives, it becomes a candidate for citation.
Content that works for AEO:
- “Best [category] tools for [specific use case] in 2026”
- “Top alternatives to [market leader] for [target segment]”
- “How to choose a [category] tool: what to look for”
These are the queries your potential customers are already asking. Create the best answer to those questions and publish it on your own site. Then reach out to independent review sites and developer blogs to create similar content that features your product.
Tactic 4: Technical signals for AI crawlers
Add structured data to your marketing pages:
SoftwareApplicationschema with accurate feature and pricing dataProductschema with offersOrganizationschema with contact details and social links
Create a /llms.txt file at your domain root. This is a relatively new convention where you summarize your product in a format optimized for AI reading. Not all AI assistants use it yet, but it is becoming more common and costs nothing to implement.
See llms.txt in the glossary for the format.
Tactic 5: High-authority mentions
Training data weights mentions from high-domain-authority sources more than low-quality sites. A single mention in a TechCrunch article, a top HackerNews thread, or a popular indie hacker newsletter carries more AEO weight than 50 mentions in low-traffic blogs.
Prioritize:
- Press coverage from tech publications in your niche
- Features or comments in HackerNews discussions relevant to your category
- Guest posts on well-read developer or founder blogs
- Conference talks and podcasts that get published as text transcripts
You cannot manufacture all of these overnight, but systematically pursuing one or two per month compounds over time.
Tactic 6: Fix your specific competitive gaps
The most efficient AEO improvement comes from identifying specific queries where you are absent and competitors are present, then fixing those gaps.
Use citation data from Perplexity to find which sources are driving competitor recommendations. Target those exact sources. If a competitor is being cited from a specific comparison blog post that does not mention you, contact that site about adding your product, or create a better comparison piece yourself.
How long does AEO improvement take?
For Perplexity and other RAG-based AI assistants that crawl the live web: 4-8 weeks for content and review changes to affect citation frequency.
For ChatGPT and Claude (training-data based): 3-6 months for changes to training data to impact recommendations. Model updates and fine-tuning cycles determine when new information gets incorporated.
This is why tracking consistently from baseline is critical - AEO improvements are not instant, and you need historical data to see the trend.
See also: What is AEO, GEO vs AEO vs SEO explained, and How to measure AI share of voice.