Method in brief
- Prompts: 24
- Prompt types: category, comparison, bottom of funnel and persona
- Engines: ChatGPT, Claude, Gemini, Perplexity, Grok, Mistral and DeepSeek
- Mode: augmented
- Web search: on
- Runs per prompt: 3
- Responses per measurement: 504
24 prompts × 7 engines × 3 runs = 504 responses.
Context
The company sold invoicing and management software to small and medium businesses. The site had good functional coverage and established organic presence. The brand, however, appeared irregularly in comparison questions, software-choice questions and recommendations by company profile.
In some answers it was presented as a relevant option. In others, competitors with similar propositions took most of the recommendations and citations.
The main problem was not absence. It was the lack of consistency between engines and prompt types.
Baseline measurement, T0
- Date
- 8 January 2026
- Citation rate
- 7.3%
- Mention rate
- 21.8%
- Recommendation rate
- 10.5%
- Share of voice
- 5.8%
- Engines recommending
- 1 of 7
- Average position
- 5.6
Reading the baseline
The brand appeared in questions directly about invoicing, but was rarely backed by a qualifying citation.
Recognition was concentrated in a single engine and depended too much on prompts containing wording close to the site's own.
More open questions, comparisons or those aimed at specific segments favoured other brands.
Work carried out
Structure
January 2026
- Clarified the primary entity and the relationship between brand, product and features.
- Reviewed structured data.
- Reorganised product, feature, integration and segment pages.
- Reduced duplication between commercial pages.
- Improved internal linking architecture.
Information
January and February 2026
- Built comparison pages by use case.
- Rewrote feature pages with verifiable information.
- Added selection criteria.
- Explained integrations more clearly.
- Created frequently asked questions.
- Published bottom-of-funnel content.
- Standardised the terminology used to describe the product.
Notability
February and March 2026
- Reviewed mentions in directories and sector sources.
- Corrected inconsistencies in the company description.
- Reinforced relevant external sources.
- Consolidated the association between brand, category and market.
Authority
February 2026
- Named authors clearly.
- Added sources and references.
- Explained the methodology in comparative content.
- Strengthened proof of experience.
Links
January to March 2026
- Linked category, feature and segment pages.
- Linked informational content to commercial pages.
- Built coherent internal paths for each search intent.
Final measurement, T1
- Date
- 8 May 2026
- Citation rate
- 29.6%
- Mention rate
- 48.2%
- Recommendation rate
- 33.9%
- Share of voice
- 16.9%
- Engines recommending
- 5 of 7
- Average position
- 2.8
Observed change
- Date
- 8 January 2026
- Citation rate
- 7.3%
- Mention rate
- 21.8%
- Recommendation rate
- 10.5%
- Share of voice
- 5.8%
- Engines recommending
- 1 of 7
- Average position
- 5.6
- Date
- 8 May 2026
- Citation rate
- 29.6%
- Mention rate
- 48.2%
- Recommendation rate
- 33.9%
- Share of voice
- 16.9%
- Engines recommending
- 5 of 7
- Average position
- 2.8
| Metric | T0 | T1 | Change |
|---|---|---|---|
| Citation rate | 7.3% | 29.6% | 22.3 pp |
| Share of voice | 5.8% | 16.9% | 11.1 pp |
| Recommendation rate | 10.5% | 33.9% | 23.4 pp |
| Engines recommending | 1 of 7 | 5 of 7 | up 4 |
| Average position | 5.6 | 2.8 | closer to the start |
Result
Over the period measured:
- Citation rate went from 7.3% to 29.6%.
- The change was 22.3 percentage points.
- Share of voice went from 5.8% to 16.9%.
- The brand went from recommended by 1 engines to 5.
- Average position went from 5.6 to 2.8, meaning the brand began to appear closer to the start of the answers.
Conclusion
The change did not come from a single page or an isolated edit.
The work set out to improve how the brand, the product, the features, the segments and the external sources were understood as one coherent whole.
The results correspond only to the method, prompts, engines and period measured.
Methodology and definitions
T0 and T1 use exactly the same prompts, engines, mode, web-search setting and number of runs, and the same definitions and formulas. Without that correspondence the change would not be computed.
- Unit of observation. Case × measurement × prompt × engine × run.
- Citation rate. Responses with a qualifying citation ÷ total valid responses × 100.
- Mention rate. Responses with a mention ÷ total valid responses × 100.
- Recommendation rate. Responses with a recommendation ÷ total valid responses × 100.
- Share of voice. Recommendations of the brand ÷ total recommendations of every brand observed × 100.
- Engine recommending. An engine with a recommendation rate of 25% or higher across the prompts in the case.
- Average position. Simple mean of the brand's ordinal position, counted only in responses where it appears. A lower figure is better.
Limitations
- Results apply only to the prompts, engines, mode and period measured. They do not extrapolate to other questions or dates.
- Results in AI engines vary by version, mode, location and moment of collection.
- A citation is not an endorsement, and a recommendation does not measure service quality.
- No effect on sales, revenue, enquiries or pipeline is claimed, because none was measured.
Nature of the evidence
Baseline and final exports from the Visibility Tracker, with prompt, engine, run, mention, recommendation, citation, position and retrieved source.
Back to all cases · The State of AI Visibility for Portuguese B2B SaaS, 2026