We launched a company from scratch: no reputation, a brand-new website, no paid media, and no sales outreach. Here’s the experiment we ran the experiment on ourselves…

Before starting No Fluff, I spent years in corporate marketing for a company with multiple services, multiple buyer groups, and multiple senior leaders with strong opinions about what the next campaign should be. Some ideas came from the service leads, some from sales feedback, and some were mandated directly from the C-suite. (Of course channel and ultimate messaging decisions were based on what had been tried and tested in the past.)

However, because each line of business had its own revenue goals (and just a small full-time marketing team to support them), it was easy for marketing to become a constant race to execution.

( P.S. That is not a criticism, it’s how a lot of B2B marketing gets done when a company has many services, many audiences, a lean team, and a lot of pressure to produce results. )

But when I started No Fluff, I knew I couldn’t build that way.

We did not have the time, budget, or brand authority to test scattered campaigns across different audiences and services. We could not afford to target multiple audiences. So to protect my time as a founder, discipline was essential.

So before we published a single page, launched the website, or wrote our first LinkedIn post, we mapped 150 buyer questions. Actual questions a founder, CEO, VP of Marketing, or revenue leader might ask ChatGPT, Gemini, or Perplexity when trying to understand AI search visibility, compare vendors, or decide whether this was a problem worth solving.

Those 150 questions became the foundation of everything we did at No Fluff from the day we launched on Dec. 11th through the next 90 days. And honestly, they changed how I think about marketing strategy entirely.

Why? Our customer-centric view vs. launch campaigns

The best marketing answers real buyer questions. That sounds obvious, but most B2B teams still build content around campaigns, internal priorities, channel plans, or keywords. Those inputs matter, but they don’t always show you how buyers actually think when they are researching a problem.

AI search made this gap visible in a way Google never did. When a buyer asks ChatGPT, “Who can help us with AI visibility for our SaaS company?”, the answer is not a list of some sponsored, some earned links. There is no ranking page to optimize for and no position 1 through 10 in AI answers (possibly 5). It’s simply, whether or not you were the best fit for the answer.

According to Forrester, 89% of B2B buyers have already adopted generative AI for self-guided vendor research. When your buyers are asking AI those questions today, the only honest way to know whether you show up is to ask the same questions yourself, systematically, and read the answers. If you do not know which questions buyers are asking, you cannot know whether AI includes you at all. That is why we started with the questions.

The six question categories

We did not pick 150 questions at random. We built a structured set of buyer questions across six clusters, each designed to test a different dimension of how AI perceives a brand.

Identity: Does AI know who you are?

These test whether AI can accurately describe your company when a buyer asks about you by name (“What does No Fluff do?”), and the goal is 100% recall. If AI cannot get this right, nothing else matters. (Technical SEO/AEO people call these Branded and Semi-branded Questions)

Category fit: Does AI connect you to the right services?

These are the hardest to win because they do not mention any brand by name (“Who are the best GEO partners for B2B SaaS?”). AI defaults to whichever brands have the strongest, clearest signals, and we targeted 30-70% visibility here (we did NOT get there, but learned more about the brands that do get picked… more on that below)

Problem fit: Does AI understand what you solve?

Problem and pain-point questions test whether AI understands what problems you solve, not just what category you belong to. For answers we want AI to come to our website, pull from our content when someone asks, “Why is my company not ranking in AI-generated answers?”

Competitive fit: Does AI compare you correctly?

Comparison prompts help determine whether AI sees you as a legitimate alternative (“No Fluff vs iPullRank”).

Authority/depth: Does AI see you as credible enough to cite?

More advanced prompts test whether AI views your brand as a thought leader, not just a participant (“Which companies are doing original research on AI visibility?”).

Taken together, these six question categories gave us a map of the dimensions AI uses to decide whether to name a brand: identity, category fit, problem-solving, and depth of authority. We ran all 150 questions across ChatGPT, Gemini, and Perplexity every week for 90 days. Same questions, same engines, same methodology. That gave us a controlled dataset we could actually learn from, week over week.

What the questions taught us

The prompt set was not just a measurement tool. It became the strategy itself.

When we saw which questions AI could not answer about us, we knew exactly what to build:

• If AI could not explain what made us different from competitors, we wrote content that made the distinction explicit.
• If AI confused us with a similarly named brand (and it did constantly in the early weeks) we strengthened entity clarity across every channel.

In other words, we did everything in our power to answer those questions clearly.

The results

The point was not that we “hacked AI visibility in 90 days,” it was that starting with buyer questions made every marketing decision sharper.

Over 90 days, we went from zero AI visibility to appearing in 39 of 150 tracked prompts (about 26%) from a company that didn’t even exist 3 months before. We earned 74 AI mentions (where our company was explicitly named) with 42 citations (where our content was pulled from directly). The majority of the work was getting AI to understand the RIGHT No Fluff and showing up correctly in the questions where people asked about us.

Here is what I didn’t expect… such focus lead to measurable gains in traditional search and founder reach:

• 158 Google search queries where No Fluff now appears.
• 20.6 total impressions, growing from zero to roughly 2,500 monthly impressions.
• Number-one Google ranking for several AI-related long-form queries.
• Founder impressions on LinkedIn up 49.1%, and engagement up 61%.
• 42 citations earned
• Inbound inquiries with little to no sales outreach

Even got a few calls from people that mistook us for an agency of a similar name!

Those are not vanity numbers; they are essential awareness numbers NEEDED for a new business in a highly competitive space.

They are the side effects of building our website, brand, content, marketing, socials around one strategy.

AI visibility is not instant, and our progress was volatile, with mention rates climbing one week and dropping the next. However, building based on clear buyer questions sharpened our messaging, improved traditional search visibility, and created momentum across channels while AI pickup built over time.

Why this matters for your team

Most marketing strategies are still built from the inside out. They start with company goals, current channels, campaign calendars, sales priorities, and historical performance. Those inputs matter, but they often miss the actual questions buyers are asking before they ever reach your website.

Mapping 150 buyer questions gave us a clearer view of the moments that actually shape buyer decision making. It showed us where our brand was understood, where we were missing the mark, where our competitors were being named, and where AI did not yet have enough confidence to include us.

For leadership teams, this is not a “content exercise.” It is about looking outward at the customer, understanding how buyers now research and decide, and aligning your strategy from there. The strongest question set might come from data already within the business or easily in reach; sales calls, objections, RFPs, demo questions, lost-deal notes, customer interviews, support themes, reviews, search data, competitor comparisons, analyst language, and internal proof.

You do not need to run a 90-day experiment to start: Ask ten questions your best buyer would ask when researching your category. Look at which companies are mentioned. Look at which sources are cited (hint: it likely wont cite your website). Look at whether your brand appears, and whether it is described accurately.

If you are absent, the answer is probably not “publish more blog posts.” The real issue is that AI does not yet have enough clear, consistent, trusted signals to recommend to you with confidence. (and of course, we can help you with that!)

What do you think of our approach to AI visibility?