Latest Research
2022
Year AI research began
3 years
Researching AI search and GEO
I809962
Taiwan invention patent (co-inventor)
Dozens
Technical books translated and published
Three Working Principles
What separates SEOBAR from a typical search marketing service.
Technical work handled directly
Structured data, llms.txt, index management and load performance are engineering work at heart. SEOBAR has a development background, so these can be inspected and changed directly instead of waiting on a separate engineering team.
- Schema.org structured data and entity recognition
- llms.txt and readability for AI agents
- Index coverage, canonicals and site architecture
- Load performance and Core Web Vitals
Measurement tools built in-house
Limitations stated up front
Why Start with Measurement
What makes search engine optimization hard is that most outcomes cannot be traced back to a cause. A ranking change can come from an algorithm update, a competitor's changes, seasonal shifts in demand or something on the site itself, and without baseline data and repeated observation any attribution is a guess.
Generative engine optimization (GEO) is even less certain. Answers from generative systems are partly random, and asking the same question several times can return different lists, so a single observation is not enough to judge anything. In our own measurements, when the same set of questions was asked again twenty-eight days later, only about a quarter of the businesses recommended the first time were still on the list.
SEOBAR therefore sets up measurement first and optimizes second:
- Get a baseline: Search Console impressions and average position, real keyword search volume, and which sources and businesses AI engines currently cite and mention
- Observe repeatedly: sample the same question many times and compare across periods to tell trend from noise
- State the limits: sample size, number of observations and what could not be measured
All the tools used for this are built in-house. AI citation tracking samples generative engines with the kinds of questions ordinary consumers ask and records the full answer, the cited sources and the businesses mentioned; the keyword tool puts real search volume next to existing Search Console impressions and positions. Building the tools ourselves means they can be rerun across periods, producing time series instead of one-off snapshots.
This process does not guarantee any particular ranking. What it does guarantee is that every change has observations behind it, so whether it worked can be checked.
From Long-Tail Keywords to Head Terms
The order decides how efficiently resources are spent.
Frequently Asked Questions
Q.How long does SEO take to show results?
It depends on how competitive the keyword is. For specific, less contested questions, changes in impressions usually show within a few weeks to two months; high-volume head terms usually take six months or more, and no specific position can be guaranteed. Trends in impressions and average position are a more reliable signal than traffic alone.
Q.How is GEO different from SEO?
SEO aims for positions in search results, while GEO aims to have content cited inside answers from generative systems. The technical foundations overlap heavily, since content has to be retrievable, clearly structured and accurate in both cases. They differ in how they are measured, because GEO has no rankings to look at and can only be tracked by sampling the same questions repeatedly and counting how often a brand is mentioned or cited.
Q.Does my current website need to be rebuilt before optimization?
In most cases, no. Technical issues such as structured data, index settings, load performance and internal linking can usually be fixed on the existing site. A rebuild is needed mainly when the site architecture itself blocks content from growing, or when the platform cannot output the necessary technical markup, and this is confirmed during the baseline audit.
Q.Can you guarantee keyword rankings?
No. Rankings are decided by search engines and generative systems, and no third party controls them. What can be promised is that every change has observations behind it and that each stage comes with measurement results you can check.
SEO / GEO Consulting Inquiries
Get SEOBAR Research Updates
Occasional emails with new SEO, GEO and AI search data and findings.