David Ly Khim

I'm a Cambodian-American entrepreneur and growth practitioner. Currently building Omniscient Digital, an organic growth agency helping B2B software companies turn SEO, LLMs, and content into revenue growth channels.

Previously, I was Head of Growth at People.ai where I led the implementation of a product-led growth motion for the new freemium products, PeopleGlass and ClosePlan, overseeing product strategy, go-to-market, and customer acquisition initiatives.

Before that, I spent 6 years working on HubSpot's freemium business through multiple cross-functional projects including: driving the implementation of referral loops across the free product, bottom-of-funnel content strategy, and business development.

One key project I worked on was developing and executing a strategy to acquire users through marketplaces and integrations, which required collaboration across marketing, business development, and multiple product teams. This resulted in spinning up a new Growth Acquisition Team in Dublin. We revived and grew adoption of (a previously sunset) HubSpot WordPress integration from 8k to 110k+ installs (adopted by 22% of the customer base), drove 1000s of product signups a month, and influenced $3.6M+ in ARR. 

Skills: product management, growth experimentation, growth marketing, data analysis, user research, user acquisition, project management, SEO, content strategy, partnerships

Personally, born and raised in Long Beach, CA, now based in Boston. Into weight lifting, running, Muay Thai, reading, and whiskey. Not all at the same time.

Co-Founder, Head of Strategy & Operations

Co-founder @ Omniscient Digital | Growth Advisor

I'm the co-founder of an organic growth agency - Omniscient - that drives attributable growth outcomes through content & SEO for ambitious B2B brands. 

Before that, I ran the experimentation team and program at Workato. We went from zero true controlled A/B tests to 20+ per quarter and revamped our digital analytics infrastructure while I was there. 

Before that, I worked on freemium growth at HubSpot (also having a short stint doing product management for a new social launch). We invented a new organic strategy called the "surround sound strategy." Later on, Semrush built a tool based on our new approach. My team drove thousands of net new freemium signups across our various product suites. 

Before that, I ran growth and content at CXL and helped launch CXL Institute. I did a little bit of everything there, from managing our Google Adwords and Facebook Ads to running webinars, creating content, and running A/B tests. 

Now, I write mainly at alexbirkett.com and live in Brooklyn, NY with my dog Biscuit

Alex Birkett

How Convert Increased LLM Visibility 81% and AI Citations 140% in 60 Days

In early 2025, Convert, a leading A/B testing and experimentation platform, recognized a fundamental shift in how buyers conduct research. Prospects were increasingly turning to AI tools like ChatGPT and Perplexity instead of traditional search, yet Convert’s presence in AI engines lagged behind competitors. 

Convert partnered with Omniscient to build a strategy tailored for this new reality. 

Rather than relying solely on conventional SEO, the team developed a Generative Engine Optimization (GEO) framework that measured and drove visibility specifically inside LLMs. This included tracking prompt-level visibility, competitive citation share, and brand mentions, metrics designed for AI search environments to be coupled with traditional search metrics. 

The approach combined rigorous baseline analysis, targeted citation-building with high-value third-party domains, and strategically structured owned content designed to be easily cited by AI models. Omniscient also aligned the work with Convert’s business goals and legacy KPIs, ensuring the effort was both measurable and sustainable. 

LLM visibility rose 81%, and AI citation share increased by 140% across tracked queries. Convert began appearing more consistently in prompts such as “top A/B testing tools,” “best CRO tools,” and “privacy-first experimentation platforms.” 

Tactically, nothing incredibly fancy happened here. 

We uncovered questions and queries that mapped to problem aware and decision aware customer journey stages. We created very, very good content that answered these questions directly. We optimized historical posts to better answer these questions, with cleaner and clearer structure. And then we conducted Surround Sound SEO to build brand mentions and ubiquity in off page citation sources influencing these prompts. Simple, not easy.

Co-founder @ Omniscient Digital | Ex-Growth @ HubSpot, Workato, CXL

Open article: Convert Grows LLM Visibility 81% & AI Citation Share 140% by beomniscient.com

Convert Grows LLM Visibility 81% & AI Citation Share 140%

We ran a study to test how often LLMs recommend brands when users aren’t explicitly shopping.

Across 180 prompts and more than 5,000 responses, Cate D. looked at queries across the customer journey. Brand mentions increase as queries move closer to purchase intent, but they don’t begin there. Even for broad, informational questions, brands appeared in roughly one out of five responses.

This suggests LLMs act less like neutral reference tools and more like category interpreters. When explaining problems or approaches, they often ground abstractions in familiar, widely recognized examples. Brands become shorthand for what’s standard or proven, whether or not the user asked for recommendations.

The effect varies by industry and model. B2B SaaS shows higher rates of brand surfacing than professional services, and different LLMs introduce brands at different points. There is no single “AI search” behavior.

The implication isn’t that brands should chase LLM mentions directly. It’s that brand salience now emerges earlier in the journey, embedded in how problems and categories are explained, well before explicit buying intent.

In complex or fast-changing industries, having someone internal to guide content can be a meaningful advantage.

Not just someone who can write, but someone who’s embedded in the business—close to product, go-to-market, and the conversations that shape strategy.

This kind of role often becomes a useful throughline as the company evolves. They’re the person who remembers why decisions were made, which messages have already been tested, and how the narrative has shifted over time.

That context can be hard to replicate externally, especially in companies where the product itself changes frequently.

External teams still play an important role, whether it’s scaling production, bringing in specialized skills, evolving distribution for AI search or emergent channels, or pushing execution forward. But it’s difficult to rely entirely on outsiders if the foundation isn’t clear and well-maintained.

In our experience, the most effective content programs tend to be built on a combination of:
-Internal ownership of messaging and priorities
-External support for strategy, scaling, and specialized execution
-A feedback loop that helps both sides stay aligned

This doesn’t mean every company needs a full in-house team from day one. But especially in technical, fast-moving spaces, having at least one person internally who carries the narrative forward tends to make external collaboration much more successful.

The more complex the subject matter, the more important it becomes to preserve continuity. Allie Konchar (Decker)

New research: what content formats LLMs cite when answering branded questions

When large language models talk about your brand, they rely heavily on third-party validation, not only your thought leadership or brand pages.

Cate D. analyzed 23,000+ citations across branded queries in ChatGPT, Perplexity, and Gemini to understand which content types shape AI-generated brand narratives. The pattern is interesting: reviews, forums, listicles, and other forms of what we bucketed as "social proof" account for the majority of citations. In aggregate, these sources function as the primary trust layer for LLMs, closely mirroring how human buyers evaluate credibility.

The second most cited category is third-party reference and directory content - product profiles, documentation summaries, and aggregators. These sources appear to serve as structured grounding for models, especially when answering factual or comparative questions about a company.

By contrast, brand-owned thought leadership and educational content plays a surprisingly small direct role in branded responses. Blog posts, research, and POV pieces are cited relatively infrequently, despite their importance for shaping narratives elsewhere in the ecosystem. Foundational brand pages (about, homepage, FAQs, contact pages) and formats like video and news are cited even less often.

The implication is not that thought leadership is unimportant, but that its influence is mostly indirect, likely influencing broader category narratives and decision points earlier in the customer journey. LLMs appear to privilege content that reflects external consensus and decision-stage evaluation over brand-authored explanations.

In AI-mediated discovery, brands are increasingly defined not only by what they publish, but by what others say about them, repeatedly, in structured and comparable ways.

Full research here: https://lnkd.in/eutmhQ8e

A growing number of companies are seeing traffic declines and pointing to AI summaries or ChatGPT as the culprit.

And in some cases, they’re right.

In many cases, this is an acceleration of a trend that had been emerging before Gen AI. It's the removal of friction to accessing low variance and high consensus queries, and often the recommendation of solutions within a single prompt or set of prompts.

In many cases, those queries look like this: 

“What is a CRM?”
“Email open rate benchmarks”
“Top features of marketing automation software”

These are consensus queries, and were so often the bedrock of SaaS SEO programs. The information isn't necessarily scarce, though the search volume was high. So traffic charts went up and to the right for those that could arbitrage these answers. But now, users can get it summarized instantly, often without clicking through.

What’s being lost here isn’t just traffic. It’s measurable publisher value.

We’re seeing clients shift focus in three key ways:
-Prioritizing information that’s hard to fake (original research, deep comparisons, proprietary insights)
Re-underwriting the value of different traffic segments(was that lost traffic ever converting? Or was it propping up vanity metrics?)
Leaning harder into brand gravity(so when people do click, they know who you are and why you’re worth listening to)

AI may flatten some of the surface-level SERPs, but it also raises the bar on what’s actually worth creating.

Some traffic is disappearing. That doesn’t necessarily mean it was valuable.

Success in AI search depends on alignment across comms, product marketing, paid, SEO, PR, content, and brand. Who’s making that happen?

We’ve seen many SEO teams absorb the AEO remit by default. But the scope is already broader than traditional SEO. And it’s only expanding.

Consider what success looks like in AI search:
-Influence on AI-generated recommendations
-Visibility across third-party aggregators and comparison pages
-Clarity on brand messaging and attribution paths
-Cross-channel alignment on buyer intent

You can’t do that in a silo.

Which is why T-shaped marketers (broadly fluent, with depth in a core area or two) are often best positioned to take ownership here. They know how to interface with different teams. They can translate between disciplines. And they can connect the dots between influence, visibility, and business outcomes.

Someone has to step up, make a case, and show what good looks like. It might be an SEO lead. It might be someone on growth. In many orgs, no one has claimed it yet.

That gives you a chance to define what it becomes.

I think one of the bigger paradigm shifts with AI search is simply one of measurement. 

Strategically, there are factors at play that reduce the apparent value of certain types of content - namely, generic top-of-the-funnel information. 

Now, was generic top-of-the-funnel information ever valuable from a business standpoint, or were just able to see traffic and click charts go up as a result of chasing high volume keywords (relevance be damned)? 

That's the "traffic trap" we've talked about for many years. 

One could also there's still value here in topical authority, or potentially even (gasp) directly for readers in order to clarify and educate about topics in your space and industry. Different discussion. 

A bigger strategic shift is the emergent impact and influence of off-page sources, aka public relations, comms, brand, and customer marketing, that now inform a large portion of your brand perception in AI summaries and answers. 

Tactically, you have some marginal differences in how content should be structured. This is evolving, and by and large, not incredibly different than in the past. Query fan outs represent another research methodology for ideating and prioritizing content and pages. 

But not being able to track a significant portion of the user journey has caused a whole lot of bedlam, some excitement, and a lot of product marketing for new AI search startups. 

So far we've tried to apply similar models to a new paradigm in attempt to bring certainty and rigor to planning. Prompts representing keywords, prompt volume representing keyword volume, prompt visibility representing rankings, etc. But the models are not isomorphic. And these open up more questions than answers. 

I expect we'll converge on performance tracking standards, but for now it's a bit of the wild west on the measurement side of things.

The tradeoff isn’t quality vs. quantity. It’s structure vs. sprawl.

As companies scale content, campaigns, or experiments, the conversation often turns to how to maintain quality at higher velocity.

But the real question isn’t whether you should prioritize quality or quantity. It’s whether your system is designed to handle scale in the first place.

On one end of the spectrum, you get total decentralization: anyone can launch anything. There’s no governance, no oversight, and very little friction. It’s fast, but risky. Mistakes compound quickly. Yours truly often gets brought in after the fact to do plumbing work from sporadic and messy SEO programs. 

On the other end, everything runs through one person or team (or a daisy chain of stakeholders, each of whom have their own scrutiny and standards bearing down on the unit of production). Quality is high, but throughput is bottlenecked. You spend more time waiting than shipping.

Neither extreme works at scale.

What actually matters is establishing the right level of governance for the complexity of your operations, then setting clear quality standards that allow for distributed execution without introducing unnecessary risk.

AI adds another layer here. Done well, it can encode certain quality thresholds into the process itself. Not to remove humans, but to reduce friction where judgment isn’t needed, and to create guardrails where it is.

In practice, scaling is a systems problem.

The more nodes you add (people, tools, channels, experiments) the more important it becomes to clarify:

Who decides what gets launched?
What does “good enough” look like?
Where do we tolerate risk, and where do we lock it down?

Velocity doesn’t require lower standards. It requires better systems.

We are hiring! Many open roles, including organic growth strategy, technical SEO, editorial, digital PR, etc.

Director of Organic Growth Strategy at Omniscient Digital

Things are still moving in 2026. Which means it’s time for us to hire more exceptional people at Omniscient Digital.

We're looking to recruit two more Organic Growth Strategists. People who are already operating at a high level and want more room to think, build, and lead meaningful organic strategy.

We’re also hiring our first Senior Technical SEO Specialist (a role very close to my heart). This is a senior, hands-on role for someone who deeply understands technical SEO and wants to help shape how technical excellence shows up across our work and the org.

Finally, we’re hiring an Organic Growth Analyst.

This role is for someone with strong fundamentals, sharp instincts, and a real interest in organic growth, who wants to develop quickly in an environment where the work matters, the expectations are clear, and the people around you are very good at what they do. You’ll learn by doing, work closely with experienced strategists, and grow into bigger responsibility over time.

If any of these roles sound like a good next step, we'd love to hear from you! Job specs attached in the comments.

taplio

The Battle

David Ly Khim

Alex Birkett

Followers

Posts

Engagements

Likes

Performance ScoreAlex Birkett's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

David Ly Khim

Alex Birkett

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Alex Birkett's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAlex Birkett's LinkedIn Performance