How Bacula Enterprise Won the AI Search Battle for HPC Backup: A GEO Case Study - Otterly.AI Blog - Best AI Search Monitoring Solution

Bacula Enterprise has achieved #1 ranking in ChatGPT responses for “best HPC backup software” by using its strategic Generative Engine Optimization. This case study shows how a specialized B2B technology company can use structured content, semantic authority building, and AI prompt monitoring to dominate a narrow range of AI-powered search results.

Introduction

The digital marketing landscape experienced a massive seismic shift in late 2022. All of a sudden, millions of users stopped using traditional search engines for many of their basic questions, turning to ChatGPT for instant, conversational answers. Google’s response to this change was to introduce AI Overviews, Perplexity quickly gained traction among researchers, and Claude emerged as another great alternative.

This was not just a trend, but a fundamental change in the way professionals discover and evaluate enterprise software. For B2B technology companies, this transformation combined opportunity and urgency. Traditional SEO metrics became less relevant as users increasingly received complete answers without ever visiting a website. The question became: “How to make sure that your solution appears in these AI-generated answers?”

Bacula Enterprise, a specialized backup solution for high-performance computing environments, faced this challenge too. Despite its unique technical capabilities and exceptional cybersecurity levels in HPC environments, the company needed visibility in an entirely new search paradigm. Their target audience – HPC data center administrators, research computing directors, and backup architects – were early adopters of AI search tools, making traditional SEO efforts ineffective.

The stakes were particularly high given Bacula’s market preferences. Unlike consumer products with massive search volumes, HPC backup solutions operate in specialized ecosystems where being mentioned is more important than being clicked on. When the IT director of a research facility asks an AI assistant for backup recommendations, appearing in that response could determine whether solutions like Bacula will be a part of the evaluation process at all.

This case study examines how Bacula Enterprise successfully navigated this transition, ultimately achieving the #1 ranking in ChatGPT when prompted with “hpc backup”, “best HPC backup software,” or similar. It will demonstrate that generative engine optimization is more than just adapting existing content – demanding understanding of how AI systems process, evaluate, and present information to users seeking technical solutions.

The New Battlefield: AI Search & GEO

The emergence of AI-powered search represents more than a technological evolution, it is a complete reimagining of information discovery processes. Unlike traditional search engines that simply present lists of potentially relevant links, generative AI tools synthesize information from multiple sources to offer a direct, contextual answer. This shift has profound implications for the way businesses achieve visibility in their target markets.

Understanding GEO

Generative Engine Optimization, also called GEO, is a new marketing discipline focused on optimizing content for AI-powered search responses instead of for traditional search engine results pages. GEO’s primary targets are the language models that generate conversational responses,while SEO is primarily focused on targeting algorithms that rank web pages.

The fundamental difference between the two lies in the way these systems process information. Traditional search engines evaluate individual pages against specific queries using factors such as keyword relevance, backlinks, and user engagement metrics. Generative AI, however, can synthesize information across different sources, looking for patterns, consensus, and authoritative signals that help construct comprehensive answers.

This creates a completely new set of optimization requirements. SEO may focus on page optimization and off-page ranking factors, but GEO prioritizes the semantic relationships and trust indicators that AI models use to determine the relevance and credibility of each source.

Why Traditional SEO is No Longer Enough

Let’s take a simple query, “backup for HPC,” as an example. A traditional search engine would return a list of vendor websites, comparison articles, and technical documentation. Users must then visit multiple sites to compare features and come to their own conclusions about the best option.

When the same query is posed to ChatGPT or similar LLM, the response is fundamentally different. These systems offer a structured analysis of HPC backup requirements, recommend specific solutions based on common use cases, and explain the reasoning behind their suggestions. AI has already become the intermediary between users and information, fundamentally changing the way businesses must present themselves.

Traditional tactics, such as technical SEO, page optimization or link building, become secondary concerns when it comes to GEO. Instead, success depends on how well the content about specific software can be processed and understood in context by language models trained on vast datasets of technical information.

The Architecture of AI Responses

AI-generated responses have their own predictable patterns that present various opportunities for optimization. Most technical queries receive answers structured around:

  • Criteria-based evaluation of available solutions.
  • Specific recommendations, with supporting rationale.
  • Problem definition and context setting.
  • Implementation considerations and next steps.

Knowing howthis architecture works can help companies structure their content in ways that align with the ways that AI systems naturally organize and present information. Instead of fighting for traditional search rankings, businesses now have the option to position themselves as the authoritative source that AI systems draw upon when constructing responses.

The companies succeeding in this new environment are not necessarily those with the best traditional SEO metrics. Instead, they are the ones whose content most effectively communicates technical authority by understanding challenges, providing an unbiased approach to addressing them, demonstrating clear feature alignment with user needs, and offering the structured information that AI systems rely on to build trustworthy recommendations based on the situation’s context.

The Challenge

Bacula Enterprise entered 2024 facing a visibility paradox that many specialized B2B businesses encounter. Despite possessing genuinely superior technical capabilities for high-performance computing backup scenarios, the company struggled to surface in the search process when potential users searched for solutions using AI-powered research tools.

The challenge was not the search volume; HPC backups are a niche market with limited monthly searches. Conversely, the challenge centered on search influence: making sure that when the right person asked the right question, Bacula appeared as a credible, well-positioned option.

The Dilemma of High-Value, Low-Volume

Traditional SEO wisdom suggests focusing on high-volume keywords, but Bacula’s target market operates differently . Queries such as “Lustre filesystem backup” or “HPC backup software” generate a very low number of searches, no more than 10 in an entire month, according to conventional keyword tools. Yet, every single search could represent a potential software procurement decision with a value in the six figures.

This creates a unique optimization challenge. Unlike the market for consumer products, in which broad visibility drives success, Bacula needed to achieve precision targeting, appearing prominently in highly specific technical queries posed by a small but valuable audience.

The target personas were equally specialized:

  • HPC administrators managing massive storage systems at scale.
  • Backup architects designing enterprise-grade solutions for scientific computing environments.
  • Research computing directors responsible for data protection in national laboratories.

All these professionals increasingly rely on AI assistance to rapidly evaluate different technical solutions, instead of reading their documentation sections on websites, making traditional website traffic metrics much less relevant than usual. What mattered here was appearing in conventional responses when these experts sought recommendations for backups.

AI Discovery Gap

Initial testing revealed a somewhat concerning pattern. When prompted with queries such as “best backup solution for HPC environments” or “Lustre filesystem backup options,” major AI platforms overlooked Bacula Enterprise. What was even more disturbing was that various generic enterprise backup solutions were recommended instead, even though they lack the specialized features required to manage high-performance computing environments.

Inferior technology was not to blame for these responses: Bacula’s feature set is more than satisfactory when it comes to aligning with HPC requirements. The problem was in the way AI systems processed and weighed information. LLMs were prioritizing generic solutions with a much broader market presence and conventional content structures over specialized tools with superior technical capabilities.

This is a classic example of an authority gap: possessing domain expertise but lacking the content necessary for AI systems to recognize that expertise and recommend their solution.

The Strategy: GEO Tactics in Action

Bacula’s approach to generative engine optimization did not rely on generic best practices or theoretical frameworks. Instead, Bacula formed and implemented a systematic strategy based on understanding how AI systems evaluate and present technical information, especially specialized B2B software queries.

The strategy in question centered on transforming Bacula’s existing technical documentation and marketing content into formats that AI systems could easily process, validate, and reference when responding to queries about HPC backup solutions.

Structured, Answer-Ready Content

The first tactical shift required restructuring Bacula’s HPC-oriented content to align with how language models parse and prioritize information. Instead of using traditional web copy, answer-ready content blocks were created to directly address common AI query patterns. In one of the most recent articles on the topic, the entire text was restructured to answer many common questions in the context of HPC – including “How to ensure minimal downtime during HPC backup operations?” or “What are the typical issues encountered during HPC backups?”, among others.

This meant reorganizing technical specifications into clear, hierarchical formats, using consistent heading structures, bullet points for features lists, and summary sections that are easily extracted and quoted by AI systems. Readability was not the only goal here, since Bacula also had to ensure that when an AI system scanned Bacula’s content, it could immediately identify key information points relevant to HPC backup queries.

Comprehensive FAQ sections were also implemented, anticipating the very questions AI systems may need to answer when evaluating backup solutions. Rather than traditional marketing FAQs, these were technical documentation structured as question-answer pairs to directly support AI-generated responses.

  • One of the FAQ questions about ZFS backups This text answered the question, “How do ZFS backups compare to traditional backup methods…? question, which is a logical topic for many people unfamiliar with the terminology or the technicalities to search.
  • Alternatively, this text about the Lustre file system answers several questions along the general lines of “What is the best backup type for Lustre,” which is also a potentially common question for new or inexperienced Lustre users.

Both topics are thematically relevant to the overarching theme of HPC backups, which is covered next.

Semantic Topic Clustering

AI systems rarely evaluate solutions in isolation; instead, they assess each solution within the broader technical ecosystem. This idea was the main driving force for creating semantic topic clusters that connected Bacula’s core backup capabilities to adjacent technologies and use cases commonly mentioned in HPC environments.

Content development expanded beyond basic backup functionality to cover related topics:

Each of these articles is related to the general subject of HPC in some way or another, reinforcing Bacula’s value. The article Big Data in HPC explains how massive datasets require high-performance computing environments to process all information and draw meaningful conclusions (climate research, genomics, etc.). The article HPC data centers explains they are dedicated warehouses with thousands of hardware units and specialized environments for cooling and safekeeping purposes.The article HPSS storage is an automated digital library for managing information output of HPC environments.

Comparing HPC vs Cloud Computing shows the core differences between two approaches (HPC is an optimized environment for a number of specialized cases, while Cloud Computing is much more generalized and useful in day-to-day business tasks). ZFS integration is necessary for having a safety net capable of detecting and fixing data corruption during backup processes.

Lustre backups involve protecting high-speed data highways that many hardware units access simultaneously, necessitating specialized strategies to avoid bottlenecks. GPFS backups are tasked with capturing parallel file systems without disrupting ongoing research activities in highly specialized environments.

True cost of HPC material reveals that the ongoing costs of HPC are often much biggermuchbigger than the upfront price of any such environment. Top HPC providers are dedicated tools to make supercomputers more useful with job scheduling and various scientific applications. Bare-metal recovery scenarios are the processes of completely rebuilding HPC environments from scratch after catastrophic failures.

Each piece of content included several internal links that reinforced Bacula’s authority across the entire spectrum of HPC data protection challenges.

This approach ensured that when AI systems researched HPC backup solutions, they encountered Bacula’s expertise across several relevant touchpoints, instead of just generic backup marketing content.

LLM Trust Signals

AI systems rely heavily on authority signals when determining which sources to reference in technical recommendations. Several strategies were implemented for Bacula’s content to strengthen its credibility indicators:

  • Third-party validation: A variety of customer case studies (NASA) and industry partnerships (Exagrid) were integrated into Bacula’s content to make them independent credibility markers that AI systems could take note of. Reviews on reputable review aggregation websites (G2, TrustRadius) from organizations with dedicated HPC goals or environments also contributed to this indicator to a certain degree.
  • Press-releases on reputable websites. Major content updates or other news about the software can be positioned on different resources on the Internet as further proof of the company’s credibility, popularity, and experience with the topic. Here’s a new feature announcement from Bacula on Yahoo Finance, as well as a 18.2 version release announcement on the NASDAQ website as examples.
  • Repeated use of topic-specific content. Virtually all modern AI environments are constantly gathering information from end users – but this can also be used to our GEO advantage by feeding the HPC whitepapers and other relevant information about Bacula in the context of HPC.
  • Temporal relevance: Regular content updates with publication dates and version information helped AI systems identify current, maintained solutions and distinguish them from outdated alternatives. An example is an article about backup strategies for Lustre filesystem, which was last edited on May 12th, 2025. Alternatively, there was also an article comparing HPC with CC (Cloud Computing), which was last edited on January 16th, 2025. Both examples are a good showcase of how content updates to existing articles can make it more relevant in AI’s eyes.
  • Documentation depth: Bacula’s technical documentation was expanded to include detailed implementation guides, compatibility matrices, and performance benchmarks; everything that could demonstrate real-world expertise instead of baseless marketing claims. For example, there is a dedicated User Guide for performing Bare Metal Recovery on Linux, There are many specific details in these guides, such as an entire section explaining nothing but ZFS and how to make the best of it as a data storage target. H3: Prompt Monitoring and Iteration

Instead of relying on traditional web analytics tools to track LLM traffic, systematic prompt monitoring was used to track Bacula’s visibility across different AI platforms. This included regular testing of various query formulations that target customers could use, from basic questions (“HPC backup software”) to complex technical scenarios (“backup solutions for a Lustre FS with support for tape archive integration”).

This process was streamlined significantly with the help of Otterly.AI, a dedicated platform designed for tracking AI search performance across multiple platforms. It managed to provide systematic monitoring of Bacula’s appearance in responses to various HPC-related queries, which was significantly easier and more precise than manually testing prompts across different AI systems. Otterly’s automated approach revealed performance variations between platforms and helped identify gaps in content that could limit visibility in specific AI-generated responses. It also made it much quicker for Bacula’s senior experts to check the final content.

Services like Otterly are also incredibly helpful in tracking brand mentions and links recalled in both ChatGPT and AI Overviews, as well as other AI search systems. This research is what helped Bacula understand its position before GEO optimization and also track the progress of the optimization.

The Results

The systematic approach to Bacula’s generative engine optimization delivered measurable improvements across multiple AI platforms. It showed that specialized B2B technology companies could achieve drastic visibility gains by understanding and optimizing their platforms for AI-driven discovery processes.

Bacula Enterprises placed as the #1 recommendation in ChatGPT’s ranking of HPC backup software (as an answer to topical queries). This was not a one-time occurrence, but a reproducible result across various query formulations, from direct questions such as “best HPC backup software” to more contextualized prompts, such as “recommend backup solution for scientific computing environment.”

The achievement proved particularly valuable because ChatGPT’s responses included specific reasoning for the recommendation, citing Bacula’s specialized HPC features instead of generic backup capabilities. This meant that any prospect could now see not just a product name, but a pre-qualified explanation of why Bacula was best for their specific use case.

Perplexity began featuring Bacula as its #1 solution HPC backup, above Veeam, HYCU, Acronis, and others. Bacula’s articles about HPC as a whole and best backup software for data centers are also used by Perplexity as reference material.

Bacula’s articles about HPC as a whole and best backup software for data centers are also used by Perplexity as reference material.

Google’s AI Overviews also began including Bacula in relevant searches, especially in queries combining specific HPC technologies and backup functionality. It was another important turning point in GEO efforts for Bacula, as AI Overviews are shown above traditional search results, potentially covering both audiences at the same time.

Bacula reached the #1 brand ranking position, as well, thanks to the abundance of interconnected articles that mention HPC, directly or indirectly, – from defining HPC and HPSS (related terms) to explaining the backup processes for specific database types, such as Oracle backup or general network backup. Notice here that each of these articles is ranked in the top-10 of materials used by AI Overviews when it comes to HPC backup software.

It would be important to mention that the results of GEO in the context of LLMs are not static, which is why the screenshot above shows weekly brand mentions as the average statistics for a period of time – where Bacula remains at the top.

Perhaps most importantly, AI systems began to accurately represent Bacula’s feature set when making recommendations. Detailed analysis of AI-generated responses showed that several different platforms were correctly identifying and highlighting Bacula’s key HPC-specific backup capabilities: its ability to manage billions of files, or seamlessly operate with Lustre, ZFS, GPFS file systems, etc. The fact that Bacula can offer all the AI-recommended HPC backup features certainly helped matters, as well.

In this screenshot, notice that Bacula Enterprise is at the top of the “Notable Solutions for HPC Backup” list, which is another victory for its GEO efforts.

Why Bacula Excelled

Bacula’s success in AI search optimization stemmed from a fundamental alignment between Bacula’s actual technical capabilities and the evaluation criteria that AI systems use when assessing backup solutions for specialized environments such as HPC. Unlike companies that struggled to match AI-recommended features with their actual products, Bacula possessed genuine technical advantages that AI systems could easily identify and validate.

Authentic Technical Superiority

The foundation of Bacula’s GEO success is in substance over optimization tactics. When AI systems research for HPC backup requirements, they consistently identify specific technical needs.

Bacula’s product development addressed these requirements through years of working with research institutions (University of Ulm, Leibniz University). As such, its technical documentation reflected real capabilities, rather than marketing claims, giving AI systems authoritative information that could withstand cross-referencing and fact-checking with ease.

Strategic Ecosystem Integration

Unlike competitors positioning backup as an isolated function, Bacula’s efforts went into presenting Bacula as a solution with vast integration capabilities within broader HPC ecosystems. Most of Bacula’s content covered itscompatibility with major storage systems, job schedulers, and data management platforms commonly used in research computing environments.

This is how Bacula could ensure that when AI systems researched specific technical combinations or adjacent infrastructure scenarios, Bacula appeared as a compatible, tested, and reputable solution, instead of just another drop in the sea of backup vendors.

Lessons Learned

Bacula’s journey into generative engine optimization revealed multiple insights that extend beyond backup software to any B2B technology company seeking visibility in AI-powered search environments.

First-mover advantage in niche markets compounds rapidly. While broad technology categories remain incredibly competitive, specialized technical niches often present vast numbers of opportunities for early adopters to establish dominant positions before competitors recognize the shift. The early investment into Bacula’s GEO for HPC backup queries helped it create a reinforcing cycle in which improved visibility led to more authoritative mentions, further strengthening its position in the niche.

Technical authenticity still trumps optimization tactics. Companies attempting to “cheat” AI systems using manipulation techniques will always be limited in what they can achieve. Luckily, AI platforms increasingly favor sources with genuine technical depth and verifiable expertise over content engineered primarily for algorithmic consumption.

Content structure matters more than content volume. Bacula’s success came from clever reorganization of existing technical content, rather than from generating massive amounts of new content. The key to success in this case turned out to be all about presenting information in formats that AI systems could easily parse, validate, and incorporate into responses.

The compound effect of semantic authority proved especially valuable. Instead of optimizing individual pages for specific queries, Bacula build interconnected expertise networks: massive volumes of topic-related content on its website that AI search systems could access, creating multiple pathways for AI discovery while reinforcing Bacula’s overall technical credibility.

Conclusion

The success of Bacula Enterprise in generative engine optimization is a demonstration of how B2B technology companies can achieve substantial competitive advantages by understanding and adapting to AI-powered search behaviors. Being able to achieve a #1 ranking in ChatGPT responses for HPC backup queries was not an accident, but the result of systematic analysis of how AI systems approach technical solution evaluation (and content development aligned with the newly discovered evaluation criteria).

“Our GEO success wasn’t just about visibility — it was about becoming the answer. By deeply aligning our actual product strengths with the way AI systems evaluate information, we positioned Bacula exactly where our buyers are looking: inside the AI recommendations themselves. This is the future of B2B marketing — where trust, technical depth, and structured content converge.”

— Andrei Iunisov, Digital Marketing Director, Bacula Systems

This case study reveals broader implications for B2B marketing in 2025 and beyond. Traditional SEO metrics become somewhat less relevant as buyers increasingly receive all their information from AI assistants without visiting vendor websites to begin with. Success requires shifting from optimizing for clicks to optimizing for mentions, making sure that your solution appears prominently when AI systems research and recommend technical solutions.

Many niche B2B categories still lack strong AI search optimization, creating openings for early adopters to establish dominant positions before markets become saturated with AI-optimized content. The path forward requires a balance between technical authenticity and strategic content architecture. Companies must actually possess genuine capabilities to justify AI recommendations, while presenting them in formats that AI can easily understand and reference.

Related Posts:

Bacula Enterprise has achieved #1 ranking in ChatGPT responses for “best HPC backup software” by using its strategic Generative Engine Optimization. This case study shows how a specialized B2B technology company can use structured content, semantic authority building, and AI prompt monitoring to dominate a narrow range of AI-powered search results.

Introduction

The digital marketing landscape experienced a massive seismic shift in late 2022. All of a sudden, millions of users stopped using traditional search engines for many of their basic questions, turning to ChatGPT for instant, conversational answers. Google’s response to this change was to introduce AI Overviews, Perplexity quickly gained traction among researchers, and Claude emerged as another great alternative.

This was not just a trend, but a fundamental change in the way professionals discover and evaluate enterprise software. For B2B technology companies, this transformation combined opportunity and urgency. Traditional SEO metrics became less relevant as users increasingly received complete answers without ever visiting a website. The question became: “How to make sure that your solution appears in these AI-generated answers?”

Bacula Enterprise, a specialized backup solution for high-performance computing environments, faced this challenge too. Despite its unique technical capabilities and exceptional cybersecurity levels in HPC environments, the company needed visibility in an entirely new search paradigm. Their target audience – HPC data center administrators, research computing directors, and backup architects – were early adopters of AI search tools, making traditional SEO efforts ineffective.

The stakes were particularly high given Bacula’s market preferences. Unlike consumer products with massive search volumes, HPC backup solutions operate in specialized ecosystems where being mentioned is more important than being clicked on. When the IT director of a research facility asks an AI assistant for backup recommendations, appearing in that response could determine whether solutions like Bacula will be a part of the evaluation process at all.

This case study examines how Bacula Enterprise successfully navigated this transition, ultimately achieving the #1 ranking in ChatGPT when prompted with “hpc backup”, “best HPC backup software,” or similar. It will demonstrate that generative engine optimization is more than just adapting existing content – demanding understanding of how AI systems process, evaluate, and present information to users seeking technical solutions.

The New Battlefield: AI Search & GEO

The emergence of AI-powered search represents more than a technological evolution, it is a complete reimagining of information discovery processes. Unlike traditional search engines that simply present lists of potentially relevant links, generative AI tools synthesize information from multiple sources to offer a direct, contextual answer. This shift has profound implications for the way businesses achieve visibility in their target markets.

Understanding GEO

Generative Engine Optimization, also called GEO, is a new marketing discipline focused on optimizing content for AI-powered search responses instead of for traditional search engine results pages. GEO’s primary targets are the language models that generate conversational responses,while SEO is primarily focused on targeting algorithms that rank web pages.

The fundamental difference between the two lies in the way these systems process information. Traditional search engines evaluate individual pages against specific queries using factors such as keyword relevance, backlinks, and user engagement metrics. Generative AI, however, can synthesize information across different sources, looking for patterns, consensus, and authoritative signals that help construct comprehensive answers.

This creates a completely new set of optimization requirements. SEO may focus on page optimization and off-page ranking factors, but GEO prioritizes the semantic relationships and trust indicators that AI models use to determine the relevance and credibility of each source.

Why Traditional SEO is No Longer Enough

Let’s take a simple query, “backup for HPC,” as an example. A traditional search engine would return a list of vendor websites, comparison articles, and technical documentation. Users must then visit multiple sites to compare features and come to their own conclusions about the best option.

When the same query is posed to ChatGPT or similar LLM, the response is fundamentally different. These systems offer a structured analysis of HPC backup requirements, recommend specific solutions based on common use cases, and explain the reasoning behind their suggestions. AI has already become the intermediary between users and information, fundamentally changing the way businesses must present themselves.

Traditional tactics, such as technical SEO, page optimization or link building, become secondary concerns when it comes to GEO. Instead, success depends on how well the content about specific software can be processed and understood in context by language models trained on vast datasets of technical information.

The Architecture of AI Responses

AI-generated responses have their own predictable patterns that present various opportunities for optimization. Most technical queries receive answers structured around:

  • Criteria-based evaluation of available solutions.
  • Specific recommendations, with supporting rationale.
  • Problem definition and context setting.
  • Implementation considerations and next steps.

Knowing howthis architecture works can help companies structure their content in ways that align with the ways that AI systems naturally organize and present information. Instead of fighting for traditional search rankings, businesses now have the option to position themselves as the authoritative source that AI systems draw upon when constructing responses.

The companies succeeding in this new environment are not necessarily those with the best traditional SEO metrics. Instead, they are the ones whose content most effectively communicates technical authority by understanding challenges, providing an unbiased approach to addressing them, demonstrating clear feature alignment with user needs, and offering the structured information that AI systems rely on to build trustworthy recommendations based on the situation’s context.

The Challenge

Bacula Enterprise entered 2024 facing a visibility paradox that many specialized B2B businesses encounter. Despite possessing genuinely superior technical capabilities for high-performance computing backup scenarios, the company struggled to surface in the search process when potential users searched for solutions using AI-powered research tools.

The challenge was not the search volume; HPC backups are a niche market with limited monthly searches. Conversely, the challenge centered on search influence: making sure that when the right person asked the right question, Bacula appeared as a credible, well-positioned option.

The Dilemma of High-Value, Low-Volume

Traditional SEO wisdom suggests focusing on high-volume keywords, but Bacula’s target market operates differently . Queries such as “Lustre filesystem backup” or “HPC backup software” generate a very low number of searches, no more than 10 in an entire month, according to conventional keyword tools. Yet, every single search could represent a potential software procurement decision with a value in the six figures.

This creates a unique optimization challenge. Unlike the market for consumer products, in which broad visibility drives success, Bacula needed to achieve precision targeting, appearing prominently in highly specific technical queries posed by a small but valuable audience.

The target personas were equally specialized:

  • HPC administrators managing massive storage systems at scale.
  • Backup architects designing enterprise-grade solutions for scientific computing environments.
  • Research computing directors responsible for data protection in national laboratories.

All these professionals increasingly rely on AI assistance to rapidly evaluate different technical solutions, instead of reading their documentation sections on websites, making traditional website traffic metrics much less relevant than usual. What mattered here was appearing in conventional responses when these experts sought recommendations for backups.

AI Discovery Gap

Initial testing revealed a somewhat concerning pattern. When prompted with queries such as “best backup solution for HPC environments” or “Lustre filesystem backup options,” major AI platforms overlooked Bacula Enterprise. What was even more disturbing was that various generic enterprise backup solutions were recommended instead, even though they lack the specialized features required to manage high-performance computing environments.

Inferior technology was not to blame for these responses: Bacula’s feature set is more than satisfactory when it comes to aligning with HPC requirements. The problem was in the way AI systems processed and weighed information. LLMs were prioritizing generic solutions with a much broader market presence and conventional content structures over specialized tools with superior technical capabilities.

This is a classic example of an authority gap: possessing domain expertise but lacking the content necessary for AI systems to recognize that expertise and recommend their solution.

The Strategy: GEO Tactics in Action

Bacula’s approach to generative engine optimization did not rely on generic best practices or theoretical frameworks. Instead, Bacula formed and implemented a systematic strategy based on understanding how AI systems evaluate and present technical information, especially specialized B2B software queries.

The strategy in question centered on transforming Bacula’s existing technical documentation and marketing content into formats that AI systems could easily process, validate, and reference when responding to queries about HPC backup solutions.

Structured, Answer-Ready Content

The first tactical shift required restructuring Bacula’s HPC-oriented content to align with how language models parse and prioritize information. Instead of using traditional web copy, answer-ready content blocks were created to directly address common AI query patterns. In one of the most recent articles on the topic, the entire text was restructured to answer many common questions in the context of HPC – including “How to ensure minimal downtime during HPC backup operations?” or “What are the typical issues encountered during HPC backups?”, among others.

This meant reorganizing technical specifications into clear, hierarchical formats, using consistent heading structures, bullet points for features lists, and summary sections that are easily extracted and quoted by AI systems. Readability was not the only goal here, since Bacula also had to ensure that when an AI system scanned Bacula’s content, it could immediately identify key information points relevant to HPC backup queries.

Comprehensive FAQ sections were also implemented, anticipating the very questions AI systems may need to answer when evaluating backup solutions. Rather than traditional marketing FAQs, these were technical documentation structured as question-answer pairs to directly support AI-generated responses.

  • One of the FAQ questions about ZFS backups This text answered the question, “How do ZFS backups compare to traditional backup methods…? question, which is a logical topic for many people unfamiliar with the terminology or the technicalities to search.
  • Alternatively, this text about the Lustre file system answers several questions along the general lines of “What is the best backup type for Lustre,” which is also a potentially common question for new or inexperienced Lustre users.

Both topics are thematically relevant to the overarching theme of HPC backups, which is covered next.

Semantic Topic Clustering

AI systems rarely evaluate solutions in isolation; instead, they assess each solution within the broader technical ecosystem. This idea was the main driving force for creating semantic topic clusters that connected Bacula’s core backup capabilities to adjacent technologies and use cases commonly mentioned in HPC environments.

Content development expanded beyond basic backup functionality to cover related topics:

Each of these articles is related to the general subject of HPC in some way or another, reinforcing Bacula’s value. The article Big Data in HPC explains how massive datasets require high-performance computing environments to process all information and draw meaningful conclusions (climate research, genomics, etc.). The article HPC data centers explains they are dedicated warehouses with thousands of hardware units and specialized environments for cooling and safekeeping purposes.The article HPSS storage is an automated digital library for managing information output of HPC environments.

Comparing HPC vs Cloud Computing shows the core differences between two approaches (HPC is an optimized environment for a number of specialized cases, while Cloud Computing is much more generalized and useful in day-to-day business tasks). ZFS integration is necessary for having a safety net capable of detecting and fixing data corruption during backup processes.

Lustre backups involve protecting high-speed data highways that many hardware units access simultaneously, necessitating specialized strategies to avoid bottlenecks. GPFS backups are tasked with capturing parallel file systems without disrupting ongoing research activities in highly specialized environments.

True cost of HPC material reveals that the ongoing costs of HPC are often much biggermuchbigger than the upfront price of any such environment. Top HPC providers are dedicated tools to make supercomputers more useful with job scheduling and various scientific applications. Bare-metal recovery scenarios are the processes of completely rebuilding HPC environments from scratch after catastrophic failures.

Each piece of content included several internal links that reinforced Bacula’s authority across the entire spectrum of HPC data protection challenges.

This approach ensured that when AI systems researched HPC backup solutions, they encountered Bacula’s expertise across several relevant touchpoints, instead of just generic backup marketing content.

LLM Trust Signals

AI systems rely heavily on authority signals when determining which sources to reference in technical recommendations. Several strategies were implemented for Bacula’s content to strengthen its credibility indicators:

  • Third-party validation: A variety of customer case studies (NASA) and industry partnerships (Exagrid) were integrated into Bacula’s content to make them independent credibility markers that AI systems could take note of. Reviews on reputable review aggregation websites (G2, TrustRadius) from organizations with dedicated HPC goals or environments also contributed to this indicator to a certain degree.
  • Press-releases on reputable websites. Major content updates or other news about the software can be positioned on different resources on the Internet as further proof of the company’s credibility, popularity, and experience with the topic. Here’s a new feature announcement from Bacula on Yahoo Finance, as well as a 18.2 version release announcement on the NASDAQ website as examples.
  • Repeated use of topic-specific content. Virtually all modern AI environments are constantly gathering information from end users – but this can also be used to our GEO advantage by feeding the HPC whitepapers and other relevant information about Bacula in the context of HPC.
  • Temporal relevance: Regular content updates with publication dates and version information helped AI systems identify current, maintained solutions and distinguish them from outdated alternatives. An example is an article about backup strategies for Lustre filesystem, which was last edited on May 12th, 2025. Alternatively, there was also an article comparing HPC with CC (Cloud Computing), which was last edited on January 16th, 2025. Both examples are a good showcase of how content updates to existing articles can make it more relevant in AI’s eyes.
  • Documentation depth: Bacula’s technical documentation was expanded to include detailed implementation guides, compatibility matrices, and performance benchmarks; everything that could demonstrate real-world expertise instead of baseless marketing claims. For example, there is a dedicated User Guide for performing Bare Metal Recovery on Linux, There are many specific details in these guides, such as an entire section explaining nothing but ZFS and how to make the best of it as a data storage target. H3: Prompt Monitoring and Iteration

Instead of relying on traditional web analytics tools to track LLM traffic, systematic prompt monitoring was used to track Bacula’s visibility across different AI platforms. This included regular testing of various query formulations that target customers could use, from basic questions (“HPC backup software”) to complex technical scenarios (“backup solutions for a Lustre FS with support for tape archive integration”).

This process was streamlined significantly with the help of Otterly.AI, a dedicated platform designed for tracking AI search performance across multiple platforms. It managed to provide systematic monitoring of Bacula’s appearance in responses to various HPC-related queries, which was significantly easier and more precise than manually testing prompts across different AI systems. Otterly’s automated approach revealed performance variations between platforms and helped identify gaps in content that could limit visibility in specific AI-generated responses. It also made it much quicker for Bacula’s senior experts to check the final content.

Services like Otterly are also incredibly helpful in tracking brand mentions and links recalled in both ChatGPT and AI Overviews, as well as other AI search systems. This research is what helped Bacula understand its position before GEO optimization and also track the progress of the optimization.

The Results

The systematic approach to Bacula’s generative engine optimization delivered measurable improvements across multiple AI platforms. It showed that specialized B2B technology companies could achieve drastic visibility gains by understanding and optimizing their platforms for AI-driven discovery processes.

Bacula Enterprises placed as the #1 recommendation in ChatGPT’s ranking of HPC backup software (as an answer to topical queries). This was not a one-time occurrence, but a reproducible result across various query formulations, from direct questions such as “best HPC backup software” to more contextualized prompts, such as “recommend backup solution for scientific computing environment.”

The achievement proved particularly valuable because ChatGPT’s responses included specific reasoning for the recommendation, citing Bacula’s specialized HPC features instead of generic backup capabilities. This meant that any prospect could now see not just a product name, but a pre-qualified explanation of why Bacula was best for their specific use case.

Perplexity began featuring Bacula as its #1 solution HPC backup, above Veeam, HYCU, Acronis, and others. Bacula’s articles about HPC as a whole and best backup software for data centers are also used by Perplexity as reference material.

Bacula’s articles about HPC as a whole and best backup software for data centers are also used by Perplexity as reference material.

Google’s AI Overviews also began including Bacula in relevant searches, especially in queries combining specific HPC technologies and backup functionality. It was another important turning point in GEO efforts for Bacula, as AI Overviews are shown above traditional search results, potentially covering both audiences at the same time.

Bacula reached the #1 brand ranking position, as well, thanks to the abundance of interconnected articles that mention HPC, directly or indirectly, – from defining HPC and HPSS (related terms) to explaining the backup processes for specific database types, such as Oracle backup or general network backup. Notice here that each of these articles is ranked in the top-10 of materials used by AI Overviews when it comes to HPC backup software.

It would be important to mention that the results of GEO in the context of LLMs are not static, which is why the screenshot above shows weekly brand mentions as the average statistics for a period of time – where Bacula remains at the top.

Perhaps most importantly, AI systems began to accurately represent Bacula’s feature set when making recommendations. Detailed analysis of AI-generated responses showed that several different platforms were correctly identifying and highlighting Bacula’s key HPC-specific backup capabilities: its ability to manage billions of files, or seamlessly operate with Lustre, ZFS, GPFS file systems, etc. The fact that Bacula can offer all the AI-recommended HPC backup features certainly helped matters, as well.

In this screenshot, notice that Bacula Enterprise is at the top of the “Notable Solutions for HPC Backup” list, which is another victory for its GEO efforts.

Why Bacula Excelled

Bacula’s success in AI search optimization stemmed from a fundamental alignment between Bacula’s actual technical capabilities and the evaluation criteria that AI systems use when assessing backup solutions for specialized environments such as HPC. Unlike companies that struggled to match AI-recommended features with their actual products, Bacula possessed genuine technical advantages that AI systems could easily identify and validate.

Authentic Technical Superiority

The foundation of Bacula’s GEO success is in substance over optimization tactics. When AI systems research for HPC backup requirements, they consistently identify specific technical needs.

Bacula’s product development addressed these requirements through years of working with research institutions (University of Ulm, Leibniz University). As such, its technical documentation reflected real capabilities, rather than marketing claims, giving AI systems authoritative information that could withstand cross-referencing and fact-checking with ease.

Strategic Ecosystem Integration

Unlike competitors positioning backup as an isolated function, Bacula’s efforts went into presenting Bacula as a solution with vast integration capabilities within broader HPC ecosystems. Most of Bacula’s content covered itscompatibility with major storage systems, job schedulers, and data management platforms commonly used in research computing environments.

This is how Bacula could ensure that when AI systems researched specific technical combinations or adjacent infrastructure scenarios, Bacula appeared as a compatible, tested, and reputable solution, instead of just another drop in the sea of backup vendors.

Lessons Learned

Bacula’s journey into generative engine optimization revealed multiple insights that extend beyond backup software to any B2B technology company seeking visibility in AI-powered search environments.

First-mover advantage in niche markets compounds rapidly. While broad technology categories remain incredibly competitive, specialized technical niches often present vast numbers of opportunities for early adopters to establish dominant positions before competitors recognize the shift. The early investment into Bacula’s GEO for HPC backup queries helped it create a reinforcing cycle in which improved visibility led to more authoritative mentions, further strengthening its position in the niche.

Technical authenticity still trumps optimization tactics. Companies attempting to “cheat” AI systems using manipulation techniques will always be limited in what they can achieve. Luckily, AI platforms increasingly favor sources with genuine technical depth and verifiable expertise over content engineered primarily for algorithmic consumption.

Content structure matters more than content volume. Bacula’s success came from clever reorganization of existing technical content, rather than from generating massive amounts of new content. The key to success in this case turned out to be all about presenting information in formats that AI systems could easily parse, validate, and incorporate into responses.

The compound effect of semantic authority proved especially valuable. Instead of optimizing individual pages for specific queries, Bacula build interconnected expertise networks: massive volumes of topic-related content on its website that AI search systems could access, creating multiple pathways for AI discovery while reinforcing Bacula’s overall technical credibility.

Conclusion

The success of Bacula Enterprise in generative engine optimization is a demonstration of how B2B technology companies can achieve substantial competitive advantages by understanding and adapting to AI-powered search behaviors. Being able to achieve a #1 ranking in ChatGPT responses for HPC backup queries was not an accident, but the result of systematic analysis of how AI systems approach technical solution evaluation (and content development aligned with the newly discovered evaluation criteria).

“Our GEO success wasn’t just about visibility — it was about becoming the answer. By deeply aligning our actual product strengths with the way AI systems evaluate information, we positioned Bacula exactly where our buyers are looking: inside the AI recommendations themselves. This is the future of B2B marketing — where trust, technical depth, and structured content converge.”

— Andrei Iunisov, Digital Marketing Director, Bacula Systems

This case study reveals broader implications for B2B marketing in 2025 and beyond. Traditional SEO metrics become somewhat less relevant as buyers increasingly receive all their information from AI assistants without visiting vendor websites to begin with. Success requires shifting from optimizing for clicks to optimizing for mentions, making sure that your solution appears prominently when AI systems research and recommend technical solutions.

Many niche B2B categories still lack strong AI search optimization, creating openings for early adopters to establish dominant positions before markets become saturated with AI-optimized content. The path forward requires a balance between technical authenticity and strategic content architecture. Companies must actually possess genuine capabilities to justify AI recommendations, while presenting them in formats that AI can easily understand and reference.

Related Posts: