Predictable B2B Inquiries from Google and AI Search: A Plan for Companies That Have Lived on Cold Calling and Referrals
Cold calling and existing customers carry less every year, and buyers now research with Google and ChatGPT before they talk to anyone. Here is a realistic plan to turn search into steady inquiries, with timelines and what to do while it grows.
Kenneth Kather · Founder & CEO, KNK Outbound
Tools in this post
Key takeaways
- Search can deliver steady B2B inquiries, but it takes six to twelve months to carry real weight. Plan the bridge: outbound and existing customers keep the pipeline full while search grows.
- Start with the questions buyers actually ask, word for word: from sales calls, emails, Search Console and AI prompts. Every important question gets a page that answers it in the first two sentences.
- AI assistants now preselect vendors. They cite pages with concrete answers, prices, processes and evidence, and they trust what third parties say about you, from trade media to directories.
- Measure inquiries, not traffic. Impressions and mentions grow long before clicks do, and in B2B a handful of the right inquiries a month changes the business.
The short answer
Yes, a B2B company that has lived on cold calling and existing customers can build steady inquiries through Google and AI search. It works if you answer the questions your buyers really ask, one page per question, with concrete content that a competitor cannot copy in an afternoon. It takes six to twelve months to carry real weight, so the existing channels have to keep running while it grows. Done this way, search does not replace sales, it hands sales warmer conversations.
Why the old channels carry less
Three things have changed at the same time. Decision makers are harder to reach by phone, because fewer of them sit at a desk with a direct line. Buying committees have grown, so one good contact no longer decides alone. And buyers do their research before they talk to anyone: first Google, increasingly ChatGPT, Perplexity or Copilot. When the assistant answers "which providers are there for X in Germany?", it names three to five companies. Whoever is not among them does not get the call.
That is the opportunity. Most established B2B companies in the German-speaking market have websites that describe themselves but answer no questions. The bar is lower than it looks.
Step 1: Collect the real questions
Do not start with keywords, start with questions. Gather them from four sources:
- Sales: What do prospects ask in the first call? What do they ask just before signing?
- Emails and quotes: Which questions come up again and again in writing?
- Search Console: Which queries already show your site, even on page three?
- AI prompts: Long conversational queries in Search Console often come from AI assistants. They show the exact wording buyers use with ChatGPT.
You will end up with 30 to 80 questions. They fall into a few groups: what you offer, what it costs, how it works, who it is for, how you compare, and what can go wrong.
Step 2: Build four kinds of pages
| Page type | Answers | Example |
|---|---|---|
| Service page per offer | What exactly do you do, for whom, at what price level | "Contract manufacturing of stainless steel parts" instead of one page "Services" |
| Page per industry or region | Only where the offer is genuinely different there | Different standards, buyers or rules per industry |
| Guides | Costs, process, comparison, decision help | "What does X cost and what drives the price?" |
| Evidence | Case studies, numbers, own data, references | A customer project with figures, a small study from your own data |
- Answers
- What exactly do you do, for whom, at what price level
- Example
- "Contract manufacturing of stainless steel parts" instead of one page "Services"
- Answers
- Only where the offer is genuinely different there
- Example
- Different standards, buyers or rules per industry
- Answers
- Costs, process, comparison, decision help
- Example
- "What does X cost and what drives the price?"
- Answers
- Case studies, numbers, own data, references
- Example
- A customer project with figures, a small study from your own data
The rule for every page: the answer comes first. The first two sentences answer the question completely. Then come the details, a table if things are compared, and at the end the questions that usually follow, as an FAQ. This is how people read, and it is how AI assistants extract answers.
The most underrated page type is evidence. A competitor can copy your service text. They cannot copy your project numbers, your process photos or the evaluation of your own data. That is also what trade media link to.
Step 3: Make it easy for AI assistants to find and cite you
AI search does not work completely differently from Google, but it rewards a few things more strongly:
- Clear facts on the page: prices or price ranges, timelines, minimum order sizes, regions served. Assistants cannot recommend you for something your site does not state.
- Third-party mentions: trade portals, industry directories, comparison lists, interviews. Assistants check whether others confirm what you say about yourself.
- Technical basics: indexable pages, fast loading, structured data, and registration in both Google Search Console and Bing Webmaster Tools. ChatGPT search also draws on the Bing index.
How to check whether it is working is described in measuring AI visibility, traffic and leads.
Step 4: Plan the timeline honestly
| Phase | What happens | What you see |
|---|---|---|
| Months 1 to 2 | Questions collected, first 10 to 15 pages live, technical basics done | Little, apart from indexing |
| Months 3 to 6 | More pages, first evidence pages, first mentions by third parties | Impressions rise, first AI mentions, isolated inquiries |
| Months 6 to 12 | Pages move up, assistants cite you more often | Regular inquiries on the topics you cover |
- What happens
- Questions collected, first 10 to 15 pages live, technical basics done
- What you see
- Little, apart from indexing
- What happens
- More pages, first evidence pages, first mentions by third parties
- What you see
- Impressions rise, first AI mentions, isolated inquiries
- What happens
- Pages move up, assistants cite you more often
- What you see
- Regular inquiries on the topics you cover
Our own numbers show the pattern. Our site went from a few hundred to more than 7,500 search impressions a week within a few months, and AI assistants cite us for many buyer questions. Clicks grew much more slowly, because assistants answer a lot themselves. The first meeting from AI search came after about six months of consistent work. Impressions are not inquiries, but they come first.
Step 5: Bridge the gap with outbound
The six to twelve months are the dangerous part, because nothing visible happens for a long time. That is why we recommend keeping an active channel running in parallel: targeted outbound by email and LinkedIn to the accounts that fit, with research and a real reason for contact. The two strengthen each other. Prospects who get a good email look you up and find answers. Buyers who found you through search recognize the name when a message arrives. How that works as a service is described on our page for B2B lead generation, and how content and ads amplify proven messages afterwards is on content and ads.
The most common mistakes
- Writing about yourself instead of answering questions. "We are a leading provider" answers nothing.
- Hiding prices completely. A range is enough, and without it you will not be recommended for budget questions.
- Twenty almost identical city pages. Search engines and assistants recognize copies. Only build a page where the content is genuinely different.
- Measuring traffic instead of inquiries. In B2B, five fitting inquiries a month can matter more than 5,000 visitors.
- Stopping after three months. That is exactly when the curve usually starts.
Frequently asked questions
Our sales live on cold calling and existing customers, but it carries less and less. How can we generate predictable inquiries through organic search?
Collect the questions your buyers really ask, build one page per important question that answers it in the first two sentences, add evidence such as case studies and your own data, and get mentioned by third parties. Plan six to twelve months and keep outbound running as a bridge until search carries weight.
How long does it take for SEO to bring B2B inquiries?
Usually three to six months for the first isolated inquiries and six to twelve months for a steady flow, depending on the competition and how much useful content you publish. Impressions and AI mentions rise first, clicks and inquiries follow later.
How do we get recommended by ChatGPT as a B2B company?
Answer buyer questions concretely on your own site, state facts such as price ranges, regions and processes, and make sure trade media, directories and comparison lists mention you. Assistants recommend companies whose claims are confirmed by other sources.
Can search replace cold calling completely?
Rarely in the first year. Search brings buyers who are already looking, outbound reaches the right accounts before they look. Most B2B companies do best with both: outbound fills the pipeline now, search lowers the cost per inquiry over time.
What content brings B2B inquiries from Google and AI search?
Service pages per offer, guides on costs, process and comparisons, and evidence pages with case studies or your own data. Generic company descriptions bring little. Pages that answer one concrete buyer question best bring the most.