AI in search: Why CMOs must shift down from top-of-funnel content

There’s this common misconception in content marketing and SEO that we have to start educating our target audience from scratch as if they’re entirely unfamiliar with the subject.

Understandably, 64% of businesses say content marketing is most effective at the top of the funnel (TOFU), compared to 54% in the middle and 53% at the bottom, according to this research. On top of that, 65% of buyers say B2B content works better when it’s not blatantly trying to sell to them, which is especially true in developer marketing.

But the problem is, there’s little chance that prospective customers will make a buying decision based on a TOFU article. What makes this even worse is that AI Overviews(AIO) now take over 96% of searches with purely informational intent, according to research by Mark Traphagen, who analyzed over a million keywords in the U.S. This leads to a noticeable drop in traffic—sometimes by as much as 10%—due to zero-click searches, and there’s really no escaping it.

For SaaS startups or businesses operating on tight budgets, investing in TOFU content often turns out to be a waste of resources, as it rarely delivers the conversions needed to justify the investment.

At Hackmamba, we want our customers to be successful by creating technical content that converts, and prioritizing top-of-funnel marketing is a massive conversion mistake. This article is a wake-up call for CMOs and technical content managers to rethink their strategies. We'll show you how the middle-of-funnel (MOFU)/Bottom-funnel (BOFU) content strategy can better help you position your marketing content for conversion while withstanding the challenges AI poses.

Why TOFU content is vulnerable to AI

TOFU is the awareness stage of the customer journey, where technical marketing content is designed to educate and build brand awareness among potential customers who are not yet ready to explore specific products. Typically, blog posts at this stage focus on broad, high-level concepts aimed at attracting search engine traffic by covering general topics.

But where it all falls apart is that TOFU content is incredibly vulnerable to AI, and there are reasons why.

Data aggregation is a key factor.

AI models like GPT or BERT are trained on massive datasets, including publicly available content, documentation, and FAQs. Since top-of-the-funnel content often covers basic, well-known topics (e.g., "What is cloud computing?"), AI can quickly generate accurate answers to these queries. Users don’t even need to click through to the original web page because AI pulls from the common knowledge that TOFU content typically rests on.

Then there’s the power of natural language processing (NLP).

Modern AI systems can accurately understand and respond to natural language queries. So when someone searches for general or introductory information, NLP models can deliver the right answer directly in the search results, thanks to AI-powered snippets.

This significantly reduces the likelihood of users visiting your website to consume your TOFU content.

Semantic search and summarization capabilities also play a major role.

AI models don’t just rely on keyword matching. They also understand the context behind a query, allowing them to provide more accurate answers. This means AI can easily summarize content from multiple sources, distilling the key points of TOFU material into concise responses without needing to direct traffic to any one source. It’s faster, and for users, it satisfies their search intent in one go. But that’s not the worst part.

TOFU content is often redundant across multiple sites.

Since it addresses frequently asked questions and general topics, not much makes it unique. AI models thrive on this redundancy—they can aggregate the information, synthesize it, and churn out answers without breaking a sweat. And when everyone produces the same type of content, it becomes even easier for AI to take over. Read more in my article on how to create differentiated SEO content.

From the research, only 6% of AI Overview results contained the actual search query, meaning AI creates responses based on context, not necessarily on pulling directly from sources.

What’s more concerning is that most AIO results don’t cite sources. No citation means no click-through—devastating for anyone relying on informational content to bring in traffic.

Why AI will struggle to replicate MOFU and BOFU technical content

There’s been growing concern that AI will eventually evolve to generate complex, lower-funnel content, just as it has mastered handling top-of-funnel informational queries. That’s because if AI search engines can provide decent responses to basic queries, it’s only a matter of time before they start generating deeper, more complex content that lives at the middle and bottom of the sales funnel.

However, based on data, this fear might be a bit premature. Research from Mark shows that AIO was triggered for just 1.2% of transactional queries and a mere 0.4% for navigational ones. While AI’s capabilities are undoubtedly improving, I believe that lower-funnel content—especially in the SaaS and developer tools space—will remain a challenge even for the most advanced AI models. Here’s why.

The complexities of MOFU/BOFU content are beyond AI’s current capabilities.

MOFU/BOFU content digs deep into our product features, configurations, and real-world use cases. For example, when developing MOFU content for a CI/CD tool like CircleCI or GitLab, we need to cover complex technical workflows, optimal configuration setups, or performance benchmarks under various conditions.

AI search engines offer high-level insights based on public data but lack the technical depth and access to proprietary tools or internal optimization strategies we use to produce detailed articles like "Optimizing CircleCI Pipelines for Kubernetes" or "Advanced Caching Techniques in GitLab CI/CD Pipelines."

Product-specific use cases are a challenge for AI.

In a SaaS comparison between Datadog and Prometheus, our BOFU content might explain how Datadog’s out-of-the-box integrations allow faster multi-cloud monitoring setups, while Prometheus may require more manual effort.

While AI can summarize basic differences, it can't create the product-specific stories rooted in real user experiences, feedback, or competitive research that make our BOFU content effective.

MOFU/BOFU content is tailored to buyer personas and context

When targeting a senior DevOps engineer evaluating Kubernetes management tools, our content might emphasize how Lens IDE offers more comprehensive cluster monitoring compared to other open-source solutions. This is how MOFU/BOFU content is designed to address distinct buyer personas' specific pain points and objectives.

AI struggles to generate content that speaks directly to the needs of these personas and their decision-making criteria, which is key to effective lower-funnel content.

AI lacks contextual insights for complex decisions

When comparing AWS Lambda and Google Cloud Functions for serverless architectures, we discuss cost structures, performance data, and scaling considerations for particular enterprise use cases. This way, our MOFU/BOFU content supports final purchasing decisions, which are highly contextual.

AI lacks the real-world insights from performance tests, customer testimonials, or product usage that our buyers need at this stage.

Building trust through human experience is irreplaceable

When you're at the BOFU stage, trust plays a massive role in the buyer's journey. Our audience wants to hear from their peers, read case studies, and see real-world examples before they commit. SaaS and developer tools thrive on trust earned through experience.

A developer might read, "We reduced deployment time by 40% after switching to Vercel." AI can’t replicate that level of trust-building content because it lacks the authority and firsthand experience. It can generate broad, generic statements or pull in aggregated reviews, but it can’t create content rooted in a real user’s journey.

AI attempts to handle topics as complex as MOFU/BOFU content will lead to hallucinations, frustrating users, and creating trust issues. In the worst case, this could result in lawsuits.

The closest AI search engines can get to answering transactional and commercial queries is crawling and aggregating content from legitimate sites—requiring citations. This ensures better accountability and is a win for users and content creators.

The way forward for CMOs and Technical Content Managers

Google's use of AI to deliver quick, summarized information to users is understandable. It cuts down the need to sift through countless blogs, saving time. However, as content managers, we need to defend the dollars invested in our content. If your content doesn't generate leads, signups, or demos, it's not doing its job.

Far too often, informational, technical blog posts barely mention a company's product or service, if at all. This is largely due to the developer marketing mindset, as mentioned earlier, where being too "salesy" is frowned upon.

The hard truth is that you're missing out on conversions if you're not discussing your product or service. And if you’re not talking about it, someone else will.

Therefore:

• While it’s true that customers need to be educated before they make a purchase, your company doesn't always have to be the one doing it. Many potential buyers already know their problem and are searching for a solution, except if your product is so unique that it requires customers to adopt an entirely new behavior. But let's be honest, that’s extremely rare.

• It’s better to reverse-engineer the sales funnel and prioritize BOFU content to reap the rewards first. While top-of-funnel (TOFU) content has its place, BOFU content should always come first.

• Understand your audience's pain points and how they align with your product's use cases. Find keyword ideas that closely match those pain points and use cases from there. This approach will significantly increase your chances of attracting and converting valuable traffic.

With this strategy, you’re building a protective moat around your content, shielding it from AI disruptions and positioning it for conversions.

Talk to us for help with technical content strategy or creation.