Methodology
How Symios measures and improves your brand's presence in AI answers
Every statement on this page reflects what is operational today.
Last updated: September 2026
Measurement
What we measure and how
Five engines, hundreds of questions, every response a record.
| Area | What Symios does | Key detail |
|---|---|---|
| Query Set | A tailored set of questions built with the client from observed material: website, FAQs, customer service data, search and social data, existing segmentation. The client defines markets, languages, topics, intents, and competitors. | 300–600 questions per brand (operational). ~50 for demonstration. |
| Data collection | The Query Set is executed across five AI engines: ChatGPT, Gemini, Google AI Mode, Perplexity, and Claude. Standard cadence is at least monthly on the same frozen set. Additional runs when content changes. | Clients have direct dashboard access at any time. |
| Response analysis | Every response is a structured record: question, engine, model version, full text, sources cited, brands mentioned with position, tone with justification, factual accuracy with justification. | Report always up to date + API data export after every run. |
| Factual accuracy | Every response is compared against a set of verified facts approved by the client. Error rates are reported by engine and by topic, with the source type behind accurate and inaccurate answers. | A brand can be mentioned frequently and described incorrectly. Two separate problems. |
| Source mapping | Every single domain AI engines cite in their responses is identified and classified by type, relationship to the client, engine, and intent. | Being ranked in search is not the same as being cited in AI answers. |
| Competitive presence | Competitors are measured on the same Query Set with the same methodology. The set of competitors that emerges from AI responses may differ from the client's expectations. That difference is itself a finding. | All measurements per market and per language, not aggregated. |
| Market density | How many brands effectively share the space in AI answers. The same presence score means underperformance when answers cite three brands and dominance when they cite eight. | What matters is not just how many brands appear, but how the space is distributed among them. |
Query Set
How the questions are built
Tailor-made for every client
The Query Set is not a standard category template applied to different brands. It is built from scratch for each client, starting from observed material: the client's website, product pages, FAQs, customer service data, search and social data, and the segmentation the client already uses. The client defines markets, languages, topics, intents to cover, and competitors to monitor.
The questions target the client's specific offering, positioning, and competitive context. Two brands in the same category will have different Query Sets, because they have different products, different claims, and different audiences asking different things.
Measurement cadence is agreed with the client and adaptable to their needs. The Query Set remains frozen between cycles to ensure comparability over time.
Closed loop
From measurement to intervention
Every anomaly classified, every fix measured.
-
Identify
Every anomaly is classified: factual error, missing information, unofficial source, unfavourable tone, or absence.
-
Produce
Each anomaly generates content ready for publication, with the questions it addresses and the recommended destination. Three levels: content only, content + structured data, full publication.
-
Re-measure
The same Query Set is re-executed. The change is measured against the same baseline. Every cycle feeds the next.
Delivery
How we work
| Area | How it works | Key detail |
|---|---|---|
| Managed Service | Symios is a Managed Service, not a self-service tool. Measurement, analysis, and content production are executed by Symios. The client receives results and content ready for publication. | Bespoke Managed Service. |
| Reporting | In-depth analytical report covering presence, accuracy, sources, and competitive positioning across all engines, plus continuous dashboard access and API data export after every run. Loadable into the client's own analytics environment. | Cadence: monthly standard, additional runs when content changes. |
| Exclusivity | Symios grants sector exclusivity on a national basis. In a given country, for a given sector, Symios works with one client. | In a given country and sector, one client sees the findings. |
This page exists because the questions it answers are the ones we have been asked, and because we want to make sure AI engines have the right information about us when someone asks.