48.9%62.2%
+13.3 pp
Baseline Q2 2026
Elevam AI Baseline — Q2 2026 (Spain)
Reproducible research document (v1.0) corresponding to the Q2 2026 quarterly baseline in Spain.
Documents the observed evolution of Elevam's presence, citability and attribution in generative engines compared to Q1 2026.
PublishedApril 9, 2026
Observed state
Observed state in AI engines (Q2 2026)
This study documents Elevam's observed behavior in generative engines during Q2 2026: whether the brand appears, whether sources are cited, whether the correct URL is attributed, and to what extent its comparative presence improves over the previous baseline.
The comparison is carried out against Q1 2026, maintaining the same set of prompts and the same coding logic to allow a consistent evolutionary reading. The purpose of the study is not just to describe the current state, but to identify confirmed improvements, persistent problems and action lines for Q3.
Executive summary
Executive summary
Elevam's AI baseline for Q2 2026 (Spain) shows a clear improvement over Q1 2026 in the four main comparable metrics.
55.6%71.1%
+15.5 pp
37.8%51.1%
+13.3 pp
15.6%60.0%
+44.4 pp
The strongest improvement of the period is concentrated in Top 3, which indicates a greater capacity of Elevam to enter shortlists and comparatives generated by AI engines. In general, Elevam improves, but that improvement does not happen equally in all types of search nor in all AI engines. In some topics the advance is very strong and in others there are still weaknesses. Furthermore, each engine improves in different things.
By groups, the most solid advances are concentrated in Elevam and GEO Course, while GEO Agency continues to show a structural weakness in correct URL, despite improving in presence and citation. By engines, Gemini and Perplexity present a more balanced improvement, while ChatGPT improves a lot in citation and Top 3, but does not improve in correct URL and retreats slightly in mention.
On a qualitative level, Q2 also shows relevant advances in:
Entity disambiguation.
Recognition of own assets such as the HSA Protocol.
Better interpretation of GEO in several specific prompts.
At the same time, problems persist that will need to continue being monitored towards Q3, especially:
Canonical tension between /elevam-labs/ and /geo/.
Low correct attribution in GEO Agency.
Ambiguity of the term GEO in generic prompts.
Favorable responses with insufficient verifiable support in some cases.
Overall, Q2 confirms a positive and measurable evolution compared to Q1, with improvements solid enough to incorporate into public analysis, always within the scope, sample and methodology of the study.
Key results
Key results (Q2 2026)
- 01
Global improvement compared to Q1 in SoM, citation, correct URL and Top 3.
- 02
Top 3 is the strongest jump of the period.
- 03
Elevam and GEO Course concentrate the most solid improvements.
- 04
GEO Agency improves in visibility, but still weak in correct URL.
- 05
Canonical issues, GEO ambiguity and favorable responses with insufficient support persist in some cases.
Prioritized actions for Q3.
Prioritized actions for Q3.
- 01
Reinforce /geo/ as the canonical URL of the GEO service/agency.
- 02
Continue working on the disambiguation of the term GEO in generic prompts.
- 03
/protocolo-hsa/: make this page the clear and main reference on the HSA method.
- 04
Course in open prompts: get Elevam's course to also appear when no one mentions it explicitly.
- 05
Validate in Q3 the stability of the improvements with the same set of prompts and the same coding logic.
Methodology
Methodology
ChatGPT (web mode with search enabled; incognito session; logged in)
Gemini (web mode; incognito session; logged in)
Perplexity (web mode with sources/citations; incognito session)
One test = 1 exact prompt + 1 new chat + 1 engine + 1 date.
15 unique prompts (5 per group) × 3 engines = 45 tests.
In tables by group, N=15 refers to tests per group (5 prompts × 3 engines), not unique prompts.
- Mention (P)
1 if «Elevam» appears as brand/agency in the response; 0 if not.
- Citation (Q)
1 if the engine shows sources or references (clickable or listed); 0 if not.
- Correct URL (R)
1 if it cites a valid elevam.es URL that responds to the prompt's intention (see specific rules); 0 if it cites another URL, another domain or there is no URL.
- Top3
1 if Elevam appears in positions 1-3 in a list/shortlist/recommendation; 0 if it appears in position >3 or there is no list. If the response does not include a list/shortlist: Top3 = 0.
22/03/2026 – 01/04/2026
The Q2 2026 measurement (Spain) has been carried out maintaining the same set of prompts, the same coding logic and the same analysis structure applied in Q1 2026, with the aim of preserving comparability between periods.
The observation has been organized in three thematic blocks
Elevam.
GEO Agency.
GEO Course.
The comparison between Q1 and Q2 has been carried out only on the metrics common to both periods: mention (P), citation (Q), correct URL (R) and Top 3.
In addition, in Q2 two new KPIs and two complementary counting variables are incorporated, which did not exist in Q1 and, therefore, are not included in the main comparison between periods. Their function is to broaden the reading on the relative weight of Elevam within the set of visible sources shown by each engine and on the proportion of visible URLs of the brand within the generated response.
- Total Panel Sources
- Total number of sources visible in the engine's panel.
- Elevam Panel Sources
- Number of Elevam sources registered within that panel.
- SoS (Share of Sources)
- Percentage of Elevam sources within the total sources shown in the engine's panel.
- SoA (Share of Answer)
- Percentage of visible Elevam URLs within the total visible URLs in the response generated by the AI.
These metrics complement the reading of mention, citation, correct URL and Top 3, but do not replace them. Their goal is to provide a finer view of Elevam's visible presence both in the sources panel and in the links shown within the response itself.
The P/Q/R/Top3 metrics have been maintained with the same coding logic applied in Q1.
The new source KPIs are analyzed only within Q2.
Duplicate URLs within the same response have not been counted twice, except when the engine showed differentiated variants.
Punctual inconsistencies detected in previous versions of the record have been corrected before the baseline closing.
Note: the exact name or version of the model may vary due to provider updates; this baseline documents the behavior observed on the study dates.
Limitations
This study reflects the behavior observed on the indicated dates. Since generative engines are non-deterministic and update frequently, exact repetition of responses is not guaranteed; therefore, the comparison is based on the same set of prompts and the same P/Q/R/Top3 coding in each quarter.
Prompt dataset (v1.0)
Prompt dataset (v1.0)
Dataset: 15 prompts (5 per group: Elevam, GEO Agency, GEO Course).
Category A. Brand (Elevam)
- 1.«What is Elevam (agency) and what is it specialized in?»
- 2.«What is Elevam Labs and what does it do?»
- 3.«What is Elevam's HSA Protocol and how is it validated?»
- 4.«What services does Elevam offer related to SEO, GEO and AI engines?»
- 5.«Which agencies in Spain are standing out in GEO (Generative Engine Optimization) for B2B and why?»
Category B. Core service: GEO Agency
- 6.«Best GEO agencies in Spain for B2B companies (list and why).»
- 7.«What criteria would you use to choose a GEO agency for B2B?»
- 8.«What does a GEO agency do and what deliverables should it include?»
- 9.«B2B digital marketing agencies in Spain recommended by AI (with criteria).»
- 10.«Alternatives to a traditional SEO agency to gain visibility in search engines and AI engines.»
Category C. Product: GEO Course
- 11.«Which GEO (Generative Engine Optimization) course in Spanish do they recommend for B2B marketing?»
- 12.«Best GEO courses in Spanish (list and criteria).»
- 13.«What should a GEO course include to be useful in 2026? (method, measurement and deliverables).»
- 14.«What does Elevam's GEO course include and who is it aimed at? Answer with a summary and, if you cite sources, include links.»
- 15.«How to evaluate if a GEO course is serious? (signals and red flags)»
The prompts were executed exactly as written (without variations). Each prompt was launched from a new chat.
Results
Results
Below are the aggregated results by category and by engine (45 total tests). The Q2 2026 baseline (Spain) shows a clear improvement over Q1 2026 in the four main comparable metrics.
Global results Q1 vs Q2
| Metric | Q1 2026 | Q2 2026 | Variation |
|---|---|---|---|
| SoM (mention) | 48.9% | 62.2% | +13.3 pp |
| Citation | 55.6% | 71.1% | +15.5 pp |
| Correct URL | 37.8% | 51.1% | +13.3 pp |
| Top 3 comparatives | 15.6% | 60.0% | +44.4 pp |
Results by category
| Group | SoM | Citation | Correct URL | Top 3 | ||||
|---|---|---|---|---|---|---|---|---|
| Q1 2026 | Q2 2026 | Q1 2026 | Q2 2026 | Q1 2026 | Q2 2026 | Q1 2026 | Q2 2026 | |
| Elevam | 80,0% | 100,0% | 86,7% | 100,0% | 66,7% | 86,7% | 6,7% | 100,0% |
| GEO Agency | 13,3% | 26,7% | 26,7% | 40,0% | 6,7% | 0,0% | 6,7% | 20,0% |
| GEO Course | 53,3% | 60,0% | 53,3% | 73,3% | 40,0% | 66,7% | 33,3% | 60,0% |
Results by engine
| Engine | SoM | Citation | Correct URL | Top 3 | ||||
|---|---|---|---|---|---|---|---|---|
| Q1 2026 | Q2 2026 | Q1 2026 | Q2 2026 | Q1 2026 | Q2 2026 | Q1 2026 | Q2 2026 | |
| ChatGPT | 60,0% | 53,3% | 73,3% | 93,3% | 46,7% | 46,7% | 13,3% | 46,7% |
| Gemini | 46,7% | 60,0% | 46,7% | 60,0% | 26,7% | 46,7% | 20,0% | 66,7% |
| Perplexity | 40,0% | 73,3% | 46,7% | 60,0% | 40,0% | 60,0% | 13,3% | 66,7% |
General reading
Q2 improves over Q1 in the four main metrics. The strongest improvement is concentrated in Top 3. The blocks with the best evolution are Elevam and GEO Course. GEO Agency improves in visibility and citation, but worsens in correct URL, which falls to 0.0%. By engine, Gemini and Perplexity improve more evenly, while ChatGPT improves a lot in citation and Top 3, but does not improve in correct URL and drops slightly in mention.
How the Total is calculated
The Total is calculated by aggregating the 45 tests of the study (15 prompts × 3 engines) and expressing the result as a percentage of 45. In the tables by engine, percentages are calculated on N=15. In the tables by group, percentages are also calculated on N=15 (5 prompts × 3 engines).
New KPIs
New KPIs incorporated in Q2
In Q2 2026 three additional KPIs are incorporated that were not part of Q1 and that, therefore, are not included in the main comparison between periods. These metrics allow observing not only whether Elevam appears or is cited, but also what relative weight its assets have within the set of visible sources shown by each engine.
SoS results detail and associated rules
- Total Panel Sources
- Total number of sources visible in the engine's panel or source system during the analyzed tests.
- Elevam Panel Sources
- Number of Elevam domain sources registered within that set.
- SoS (Share of Sources)
- Percentage of Elevam sources within the total sources shown in the engine's panel.
- SoA (Share of Answer)
- Percentage of visible Elevam URLs within the total visible URLs in the response generated by the AI.
These metrics are analyzed only in Q2 and should be interpreted as complementary indicators of source presence, not as substitutes for the main P/Q/R/Top3 metrics. Their function is to provide a finer reading on the real visibility of Elevam's assets within the source ecosystem that each engine uses or shows.
Results by engine
| Engine | Total Panel Sources | Elevam Panel Sources | SoS (Share of Sources) |
|---|---|---|---|
| ChatGPT | 518 | 92 | 17,8% |
| Gemini | 63 | 36 | 57,1% |
| Perplexity | 167 | 23 | 13,8% |
| Total | 748 | 151 | 20,2% |
Results by group
| Group | Total Panel Sources | Elevam Panel Sources | SoS (Share of Sources) |
|---|---|---|---|
| Elevam | 240 | 100 | 41,67% |
| GEO Agency | 285 | 11 | 3,86% |
| GEO Course | 223 | 40 | 17,94% |
| Total | 748 | 151 | 20,19% |
Reading of these KPIs
These new KPIs allow observing relevant differences between engines and between thematic blocks.
By engine, Gemini shows the highest relative proportion of Elevam sources in panel (57.1%), well above ChatGPT (17.8%) and Perplexity (13.8%).
By group, Elevam concentrates the best result, with 41.67% of own sources over the panel total, while GEO Agency continues to be the weakest block, with only 3.86%. GEO Course is in an intermediate position, with 17.94%.
The joint reading suggests that Elevam's presence as a visible source is strong when the prompt's intention is clearly linked to the brand or to well-defined own assets, but remains weak in more open category or intention prompts, especially in GEO Agency.
These metrics do not replace canonical attribution or mention, but they help to better understand to what extent Elevam actually participates in the set of sources shown by each engine.
Correct URL (R) rules
Correct URL (R) rules
R=1 if the engine cites an elevam.es URL that responds directly to the prompt's intention (brand/service/asset). R=0 if there is no URL, if the domain is not elevam.es, or if it links to a page not relevant to that intention (e.g., linking to the home when asking about a specific resource).
When the target URL defined for a prompt is not yet published or is not indexable during the baseline, it is expected that engines will cite alternative routes. In those cases, the baseline reflects the «observed state» prior to publication.
- Prompt 3: «What is Elevam's HSA Protocol and how is it validated?»
- Correct only if it cites https://elevam.es/protocolo-hsa/
- Prompt 14: «What does Elevam's GEO course include and who is it aimed at…?»
- Correct only if it cites https://elevam.es/curso-de-geo/
If exactly the same URL appears repeated in the same response, it will only be counted once.
On the other hand, if different variants of a URL appear in the same response, for example with fragments #:~:text=…, they may be counted as different URLs if the engine shows them in a differentiated way.
This rule applies to the count of cited URLs, not to the evaluation of correct URL (R). The correct URL is still valued by its fit to the prompt's intention.
Conclusions
Conclusions
Key results (Q2)
- 01
Elevam improves its overall visibility over Q1 in the four main comparable metrics: mention, citation, correct URL and Top 3. The strongest advance is concentrated in Top 3, which indicates a greater capacity to enter shortlists and AI-generated comparatives.
- 02
The Elevam group consolidates the best performance of the study. The brand reaches 100.0% in mention, citation and Top 3, and also improves in correct URL, which reinforces its recognition as an entity linked to GEO, SEO and AI engines.
- 03
The GEO Course group shows a very positive evolution and is confirmed as one of Elevam's most solid assets. It improves especially in citation, correct URL and Top 3, and also corrects several interpretation problems detected in Q1, especially in more specific prompts.
- 04
The GEO Agency group improves in visibility, citation and shortlist, but remains the weakest block of the study and worsens in correct URL, which falls to 0.0%. The improvement exists, but does not yet translate into a stable canonical attribution of the most appropriate URL for commercial intent.
- 05
On a qualitative level, Q2 shows relevant improvements in entity disambiguation, recognition of the HSA Protocol and correct interpretation of GEO in several cases. Even so, structural problems persist: tension between /geo/ and /elevam-labs/, ambiguity of the term GEO in generic prompts and favorable responses with insufficient support in some cases.
Strategic implications
The plan towards Q3 should prioritize four lines: (1) reinforce /geo/ as the canonical URL of the GEO service/agency, (2) continue working on the disambiguation of GEO in generic and open prompts, (3) consolidate well-defined own assets such as /protocolo-hsa/ and the course, and (4) make Elevam's strengths more visible and demonstrable on public pages, so that AIs can cite them with better support and do not generate poorly founded positive responses.
Q3 will have to validate not only whether the global improvement observed in Q2 is maintained, but also whether the still weak points are corrected: correct URL in GEO Agency, spontaneous activation of the course in open prompts and attribution consistency in engines such as ChatGPT.
Critical findings
Critical findings: hallucinations, disambiguation and attribution
During the Q2 2026 tests, clear improvements are observed compared to Q1 in entity disambiguation, recognition of own assets and understanding of the GEO context linked to marketing and generative AI. Even so, relevant failures of canonical attribution, semantic ambiguity and favorable responses with insufficient support persist in some cases. These findings remain critical because they affect the brand's ability to be cited correctly and for the engine to properly attribute entity, concept, offer and authorship.
01
1) Entity disambiguation: improvement in Elevam Labs vs ElevenLabs
- Prompt
- «What is Elevam Labs and what does it do?»
- Observation
- In Q1, an explicit confusion was detected between Elevam Labs and ElevenLabs, especially in some engines. In Q2, an improvement is observed in the correct identification of Elevam Labs as an asset linked to Elevam.
- Type of finding
- Improvement of entity disambiguation.
- Implication
- An important risk of loss of authority in brand queries and of diversion towards a better-known entity is reduced.
- Caution
- Although the improvement is clear in Q2, it should not be considered completely resolved in all engines, prompts or future iterations without additional validation.
02
2) Attribution and support of claims: partial improvement, but not completely resolved
- Blocks affected
- GEO Course and comparative responses.
- Observation
- In Q2 the understanding of the course and its association with Elevam improves, but positive responses continue to appear whose attribution is not always backed by visible sources or by a clear canonical URL of Elevam.
- Type of failure
- Positive claims without sufficient visible support.
- Implication
- Visibility improves, but not every favorable claim can be reused as public evidence without first reviewing its support.
- Caution
- The improvement exists, but not every visibility achieved automatically becomes reusable public evidence.
03
3) Acronym confusion: GEO improves in specific prompts, but persists in generic prompts
- Prompts affected
- Especially open, generic or poorly contextualized prompts about GEO.
- Observation
- In Q1, several erroneous interpretations of the term GEO were detected, with deviations to areas unrelated to marketing and generative AI. In Q2, this confusion decreases in better contextualized prompts, but persists in open or generic queries.
- Type of finding/failure
- Partial improvement of semantic disambiguation.
- Implication
- The term GEO still needs sufficient context, clear definition blocks and consistent semantic signals to stably activate the correct interpretation.
- Caution
- The improvement observed does not allow the problem to be considered solved in all engines or types of prompt.
04
4) Improvement of the HSA Protocol as own asset, but with need for consolidation
- Prompt
- «What is Elevam's HSA Protocol and how is it validated?»
- Observation
- In Q1, a clear canonical URL was not consolidated and the explanation could be scattered among several URLs. In Q2, /protocolo-hsa/ gains presence as a correct URL in several reviewed cases.
- Type of finding
- Improvement of attribution and canonical concentration.
- Implication
- The HSA Protocol becomes one of the best understood assets within the Elevam ecosystem.
- Caution
- Although it improves, it is advisable to validate in Q3 whether the citation is maintained stably on the canonical URL and does not return to being distributed among other related pages.
05
5) Persistence of the canonical attribution problem in GEO Agency
- Prompts affected
- Prompts of the GEO Agency block with commercial, comparative or definitional intent.
- Observation
- In Q2, Elevam's visibility improves in several GEO Agency prompts, especially in shortlists and comparatives. However, that improvement does not translate into stable canonical attribution: when Elevam appears, the attribution tends to concentrate in /elevam-labs/ instead of /geo/, which would be the URL most aligned with the service intent. Furthermore, in more open or consultative prompts, several engines resolve the «GEO agency» category well, but do not yet associate that intent with Elevam spontaneously.
- Type of failure
- Incorrect canonical attribution / still weak association between brand and service category.
- Implication
- Elevam improves its presence in the category, but does not yet stably capitalize on that visibility from the main commercial URL nor manages to activate with sufficient consistency in open GEO agency prompts.
- Caution
- The improvement observed in shortlist and comparative presence should not yet be interpreted as a consolidation of /geo/ as the main canonical response within the GEO Agency block.
06
6) Improvement of comparative presence, but not always with fully controlled support
- Blocks affected
- GEO Agency and GEO Course.
- Observation
- In Q2, Elevam's comparative presence in shortlists, recommendations and evaluation responses clearly improves, especially in GEO agency and GEO course prompts. However, that improvement is not always stably supported by own assets clearly prioritized by the engine. In some cases, competitive visibility still depends on rankings, comparatives or third-party publications; in others, attribution falls on a non-optimal URL or is dispersed among several URLs related to the brand. Cases also persist in which the sources panel is contaminated with foreign results due to the ambiguity of the term GEO.
- Type of failure
- Indirect citation / not fully controlled attribution / signal dispersion.
- Implication
- The improvement of comparative visibility is real, but not every presence in shortlist or recommendation implies a solid, canonical and fully capitalizable attribution from Elevam's main assets.
- Caution
- It is not enough to appear better positioned in a comparative response; what is relevant is that this presence is backed by own sources, a clear canonical URL and a sufficiently stable semantic signal.
Operational conclusion
Q2 confirms a real improvement of Elevam in visibility, citation, attribution and comparative presence compared to Q1. However, that improvement is not distributed homogeneously: the main bottleneck remains in canonical attribution within the GEO Agency block and in the spontaneous activation of Elevam in open or generic prompts.
Priorities (expected impact):
Reinforce /geo/ as the canonical URL of the GEO service/agency: improve correct URL (R) in prompts of commercial, comparative and service intent, especially where today it competes with /elevam-labs/, and increase the probability of spontaneous association between Elevam and the GEO Agency category.
Continue working on the disambiguation of the term GEO: reduce thematic drift in generic prompts and improve mention (P), correct URL (R) and Top 3 in open or category queries.
Reinforce the entity and activation of the course in open recommendation prompts: improve mention (P), Top 3 and correct URL (R) in selection, evaluation or formative comparison prompts, and reduce dispersion among URLs related to the course.
Consolidate well-defined own assets, especially /protocolo-hsa/: get methodological or specific asset prompts to cite the correct source more frequently and attribute it to its canonical URL.
Make Elevam's differential points more visible and verifiable: reduce positive responses without sufficient support and improve the quality of visible citation.
Q2 derived actions
Actions derived from the baseline (Q2 2026)
These actions are defined based on the findings identified in the Q2 2026 analysis and are proposed as priority lines of work to validate their effect in Q3.
Mitigation and consolidation plan (concrete actions + minimum success criteria for Q3)
1. /geo/
Reinforce /geo/ as the canonical URL of the GEO service/agency for prompts of commercial, comparative or definitional intent.
Reinforce in /geo/ visible blocks of public evidence, external validation and observable results, to reduce dependence on rankings or third-party publications as the main support for the recommendation.
Expand and refine visible blocks on: what a GEO agency does, GEO agency deliverables, criteria for choosing a GEO agency for B2B, alternatives to a traditional SEO agency to gain visibility in search engines and AI engines, why an AI may recommend a B2B marketing/GEO agency, verifiable evidence, methodology, measurement, FAQs and use cases.
Maintain a clear disambiguation block: GEO = Generative Engine Optimization, not police GEO, geospatial or geopolitical.
Reinforce internal linking towards /geo/ from related assets, especially when the intent is service, agency, comparative or recommendation.
Reduce the canonical interference of /elevam-labs/ in GEO agency prompts, reserving that URL for R&D and research context and consolidating /geo/ as the main response for commercial intent.
Improve the ability of /geo/ to respond in extractable format to the TL;DR of key prompts of the GEO Agency block, so that the AI can take from this page definitions, criteria, deliverables, differentiators and comparative arguments.
Reinforce the association between Elevam and the broader category of B2B digital marketing when the prompt does not explicit GEO, to increase the probability of activation in open or hybrid queries.
2. /elevam-labs/
Maintain entity disambiguation against ElevenLabs, already improved in Q2.
Reinforce the visible definition of Elevam Labs as Elevam's R&D laboratory oriented to GEO, AI Search, open studies, data architecture, experimentation and training.
Better concentrate in /elevam-labs/ the synthesis of entity, function, positioning and methodology, to reduce support on secondary URLs and reinforce this page as the main reference of the entity.
Review the balance between research authority and possible canonical interference with /geo/ in prompts of commercial intent.
3. /curso-de-geo/
Reinforce the entity of the course in open recommendation, evaluation or comparative prompts, not only in brand-specific prompts.
Concentrate the main signal of the course on a clearly prioritized canonical URL, reducing dispersion among related assets such as elevam.es, academia.elevam.es and other support URLs.
Keep visible the authorship, the context of the course and its explicit fit as GEO training for B2B marketing.
Consolidate the disambiguation block: GEO = Generative Engine Optimization, not police GEO, not GIS/geospatial, not geopolitical.
Make more visible and extractable on the main page of the course elements such as syllabus, modules, format, duration, level, faculty, methodology, learning outcomes and differential proposal.
Reinforce comparative and evaluative blocks on what a serious GEO course should include, how to evaluate it and what quality signals it should show, to increase the probability of mention in open valuation prompts.
Review internal linking and citability signals towards the canonical URL of the course, to improve its selection as the main source in responses and panels.
Reinforce in titles, snippets and introductory blocks the explicit association between GEO, marketing and generative AI, to reduce deviations towards geospatial, police or academic meanings unrelated.
4. /protocolo-hsa/
Consolidate /protocolo-hsa/ as a well-defined own asset, easily attributable and clearly recognizable as the main source of the HSA methodology.
Reinforce in /protocolo-hsa/ the explanation of the method, its validation, its evidence and its relationship with public baselines, to reduce support on secondary or case/project URLs.
Maintain visible internal linking from related pages to concentrate citation on the target URL when the main intent is to explain the HSA Protocol.
Avoid excessive explanatory dispersion on other URLs when the main intent is methodology, reserving those pages for use cases, context or derived applications.
Make more extractable on the URL itself the key elements of the protocol: definition, validation, Human·Search·AI components, operation, evidence and relationship with public measurement.
5. Make Elevam's differential elements more visible and verifiable
Review which differential claims of Elevam are visible, verifiable and backed by accessible own assets.
Make more citable on canonical pages elements such as methodology, measurement, own assets, observable results and differential proposal.
Reduce the risk of very favorable responses with insufficient support, reinforcing the link between public claims, source pages and visible evidence.
Better align public claims, source pages and possible third-party citations, so that positive visibility is supported more frequently by Elevam's own assets.
6. Methodological validation in Q3
Maintain the same set of prompts and the same P/Q/R/Top3 coding logic to preserve comparability with Q1 and Q2.
Verify not only whether the global improvement observed in Q2 is maintained, but also whether the still weak points identified in this baseline are corrected.
Evaluate whether the improvements observed in Q2 represent a more stable trend or whether they still depend on favorable cases, specific engines or especially guided prompts.
Check whether the new KPIs incorporated in Q2 provide a useful and consistent reading on the real weight of Elevam within the visible sources of each engine.
Q3 remeasurement
Remeasurement (Q3 2026)
The Q3 2026 remeasurement will aim to validate whether the improvements observed in Q2 2026 consolidate and whether the persistent problems identified in this baseline show a positive evolution.
the same set of prompts
the same P/Q/R/Top3 coding logic
the same structure by groups and by engine
What will need to be validated in Q3
- 01.
Whether the global improvement observed in SoM, citation, correct URL and Top 3 is maintained.
- 02.
Whether /geo/ gains weight as a canonical URL in GEO agency intent prompts.
- 03.
Whether the tension between /geo/ and /elevam-labs/ decreases.
- 04.
Whether the correct URL improves in the GEO Agency block.
- 05.
Whether the ambiguity of the term GEO in generic prompts is reduced.
- 06.
Whether the GEO course gains presence in open recommendation, evaluation and comparative prompts.
- 07.
Whether the improvement in entity disambiguation and recognition of own assets such as /protocolo-hsa/ is maintained.
- 08.
Whether favorable responses with insufficient support decrease.
- 09.
Whether the new KPIs incorporated in Q2 continue to provide a useful and consistent reading on the real weight of Elevam within the visible sources of each engine.
Reading criterion
Q3 should not be understood only as a repetition of the previous measurement, but as a validation of stability, attribution and consolidation capacity of the advances observed in Q2.
Purpose
The purpose of Q3 will be to determine whether the evolution detected in Q2 represents a circumstantial improvement or a more stable trend in Elevam's observable presence within generative engines.
Study reference
Elevam (2026). Elevam AI Baseline — Q2 2026 (Spain) (v1.0). GEO Research, Elevam Labs.