In June 2026 I attended a LinkedIn webinar on how B2B marketing measurement is changing with AI search, and the session gave a clear, data-backed view of how B2B buying and measurement are changing. The session was moderated by Reuben Webb, B2B Editor at The Drum, and brought together Inna Meklin, who leads web growth and AI search at LinkedIn, Eren Özgen, data insights and operations leader for Microsoft EMEA, and Adam Holmgren, CEO and co-founder of the B2B attribution company Fibbler.
What stayed with me most was that LinkedIn chose to put its own marketing numbers on the screen and talk openly about what AI search has done to them. Traffic is down and leads are down, and yet the business is getting better quality enquiries than it was a year ago, which is a far more useful story for B2B leaders than a simple traffic chart.
The session was valuable because it was a marketing team being candid about its own data, with Microsoft and Fibbler adding the view from people who measure the same shift at scale, and I think every leadership team in a growing B2B company would benefit from hearing it.
Why does AI search matter so much to B2B buying decisions?
AI search matters to B2B buying decisions because buyers increasingly settle on a preferred supplier before they contact anyone, and AI search tools are now where a large share of that early research happens. Forrester's research into B2B buyer behaviour, cited by The Drum ahead of the webinar, ranks AI conversational search tools first among discovery sources, ahead of vendor websites, sales reps and Google.
- 68% of B2B buyers already have a preferred vendor in mind before the formal purchase process begins, according to Forrester. - That preferred vendor wins around 80% of the time. - LinkedIn is the second most cited source across major AI search engines, including ChatGPT, Microsoft Copilot and Google AI Mode, according to Meltwater research on 9 million citations.
Forrester describes the modern B2B buying journey as a process of confirmation, and I think that phrase captures the shift well, because it means the shortlist is usually formed long before a company knows a buyer is looking, and AI answers are increasingly the place where that shortlist takes shape.
What did LinkedIn find about AI search in its own data?
LinkedIn's marketing team, presenting its own data at the June 2026 webinar, found that AI search reduced website traffic and lead volume while improving the quality of the leads that did arrive.
- AI Overviews are now triggered for 47% of the keywords LinkedIn tracks, up from 7% a year earlier. - Organic search traffic to its marketing websites fell year on year, and so did marketing qualified leads. - LLM traffic grew 2.3 times, and the visitors who did arrive were more qualified, converting at 1.7 times the rate.
If a company with LinkedIn's resources is seeing its traffic fall, then a growing Irish company noticing the same pattern has good reason to look at how buyers are researching before assuming anything is wrong with its marketing.
Should B2B companies worry when website traffic falls?
B2B companies should look closely at a fall in website traffic, and read it alongside qualified sales conversations, because Forrester's research shows buyers now form a preferred shortlist early, increasingly with the help of AI answers, before they ever visit a website.
I have sat in plenty of leadership meetings where the first slide is a traffic graph heading the wrong way, and the reaction is usually the same, with a quiet room, a question about whether the website needs rebuilding and a request for the agency to explain itself.
My view is that the graph on its own asks the wrong question. Buyers now ask ChatGPT, Perplexity or Google who the main suppliers in their category are, and they read the summary before they click anywhere, so by the time they reach your website they have already done a good part of the shortlisting. Fewer people arrive as a result, but the people who do are much closer to a decision. The question I put to leadership teams is whether qualified sales conversations are holding up or growing, because if they are, the falling graph is mostly telling you how buyers research today.
This is also why I would never advise a company to stop investing in SEO. AI tools lean heavily on well structured, crawlable websites, so sound SEO is the foundation that AI visibility is built on, and good technical foundations give your expertise the best chance of being found and quoted.
What is the difference between brand and non-brand AI search topics?
Brand topics are AI searches that name a company directly, while non-brand topics are the general questions buyers ask before they know which companies exist, and LinkedIn tracks both. For me this was the most useful idea in the whole session, and it also explains why measurement was the theme, since a team that only measures brand searches and website visits will miss the stage where buyers form their shortlist.
- Brand topics. Searches about LinkedIn by name, such as LinkedIn Ads or Sales Navigator, where the goal is accuracy and making sure AI answers describe the products correctly. - Non-brand topics. Broader buyer searches such as B2B marketing, sales prospecting and B2B lead generation, which LinkedIn sees as the main lever for growth and competitive positioning.
When companies first look at AI search, they tend to type in their own name, see a reasonable answer and feel reassured. That check is worth doing, but it is the easy part, because nobody who has never heard of you will search for you by name. The pipeline sits in the general questions a buyer asks first, such as who can help us enter the Irish market or how we should build a marketing function for the first time, and a company that does not appear in those answers is unlikely to make the shortlist, however good its website looks.
Where do AI tools get information about B2B companies?
AI tools such as ChatGPT, Perplexity and Google AI Overviews build their picture of a B2B company mostly from sources outside its own website, and LinkedIn listed six types of source that shape how these tools describe and recommend companies.
- Publications cited in external media. - Expert opinions. - Content that answers customer questions. - Mentions in forums and reviews. - Customer communities and creators. - Customer stories of first-hand experience.
What strikes me about this list is how much of it sits outside a company's own website, which means a strong presence in AI search depends on the wider reputation a business builds across LinkedIn, the press, reviews and the communities its buyers belong to.
Which AI search visibility levers work fastest for B2B companies?
LinkedIn Articles, paid activity and website content can be activated within weeks, while PR, partners and creator work usually take eight weeks or more, according to the impact and effort matrix LinkedIn shared, which grouped each visibility lever by how quickly it can be activated.
- Immediate, within a week. LinkedIn paid activity, Reddit and Google paid, LinkedIn Articles and ads on AI platforms. - Quick, one to four weeks. B2B web microsites, blogs and help centre content. - Medium, four to eight weeks. Forums and community discussion, and product reviews. - Long, eight weeks or more. Partners, PR and communications, and influencer and creator work.
LinkedIn was careful to say that results depend heavily on industry, and I agree. The right mix for any business depends on its industry, its buyers and where its visibility stands today, and working that out is where a clear plan makes the biggest difference.
What should B2B leadership teams ask about AI search?
B2B leadership teams can judge how ready they are for AI search by asking three questions about their own business.
- Do we know what ChatGPT, Perplexity and Google currently say about our company? - Do we show up in AI answers to the questions our buyers ask before they know our name? - Are we measuring qualified conversations and pipeline alongside website traffic?
If any of those questions are hard to answer, and you would like to see how your company currently shows up in ChatGPT, Perplexity and Google, and where the biggest opportunities are for you, I would be glad to walk through it with you on a short call.