Is the Internet Too Noisy for B2B? What the Data Says Before You Move Your Budget to Events

How we built this answer

We ran four studies on 14 October 2026. The first gathered measurements of how noise has changed each online channel: organic social reach, cold outreach, search after AI answers and the price of paid media. The second gathered what events and online channels cost, how long each takes to produce a result and how B2B buyers now choose a supplier. The third read what 14 marketing leaders whose posts our buyers follow write and do about events and noise: 332 of their posts published between April and October 2026, plus their newsletters, event pages and interviews. The fourth read 2,287 comments written by buyers under those leaders' posts: 1,431 on LinkedIn, collected on 8 October 2026, and 856 replies on X, collected on 9 October.

Many of the largest datasets on this subject come from companies that sell the channel they measure: an SEO agency measuring search, an events platform measuring events, an outreach tool measuring outreach. We use the direction of such data and name the seller's interest each time. Where only the seller has published a number and the method is hidden, we leave the number out.

1. The noise is measurable, and it is concentrated where B2B publishes

In April to July 2026, Pangram Labs scanned 1,002,627 posts on LinkedIn, X, Reddit, Medium and Substack, collected through a browser extension from people who opted in. More than 40% of long LinkedIn posts were written entirely by AI. LinkedIn made up about a third of everything scanned and 62% of all the AI-written content found. On the open web, Graphite checked 43,000 random articles published from January 2020 to May 2025: machine-written articles overtook human ones in November 2024 and reached 51.7% in May 2025, then stopped growing.

Search lost clicks when AI summaries arrived. Pew Research Center tracked the browsing of 900 US adults through March 2025: when Google showed an AI summary, people clicked an ordinary result in 8% of visits, against 15% without one, and clicked a link inside the summary in 1%. SparkToro puts US Google searches ending without a click on the open web at 68.01% in early 2026, against 60.45% in 2024; the two years come from different clickstream panels, so part of the gap may be the panels.

Organic reach fell, unevenly. Across 16,645 LinkedIn company pages, Socialinsider measured video views down 36% from 2024 to 2025. On X, Metricool's 1.1 million posts show impressions per post down only 5%, while clicks through to the author's profile fell 31%: people react inside the feed and leave it less often. The most quoted figure, "LinkedIn reach down 60% in two years", comes from a paid report on active creators whose method is not published, so we do not rely on it.

Outreach answers less to the template. Expandi analyzed 13,218,869 LinkedIn connection requests sent from May 2025 to April 2026: replies to the request itself fell from 3.5% to 2.2%. In cold email, Belkins measured 7.5 million emails by one method on one base and saw replies fall from 0.50% in the first half of 2025 to 0.40% in the second.

Paid media did not get dearer everywhere. In Meta's filing for 2025, the average price per ad rose 9% while ad impressions grew 12%. Tinuiti, tracking advertisers who kept their strategy unchanged, saw Google search cost per click falling slightly at the end of 2025. B2B advertising on LinkedIn is the exception: in Dreamdata's B2B data, the cost of a thousand impressions rose from €26.62 to €34.33 (+29%) and the cost per click by 11.8% in a year.

2. Access to people held. Replies to the template fell

The same Expandi data that shows replies to connection requests falling also shows what did not change: people accepted 28–30% of requests all year, and replies to actual messages held at 10–11%. The door opens as often as it did. What fell is the response to a request that looks like every other request.

Smaller and more precise wins in every channel where it has been measured. Facebook pages with 1,000 to 5,000 followers reach 4.35% of their audience, pages with 100,000 to a million reach 0.50% (Socialinsider, 872,075 posts). Cold email campaigns to fewer than 50 people get 5.8% replies, campaigns to more than 1,000 get 2.1% (Woodpecker, 20+ million emails). Campaigns to fewer than 200 prospects get nearly twice the reply rate of campaigns to 500+ (Saleshandy, 30.31 million emails, July–October 2026).

The mass of machine text is also not where buyers find answers. Graphite, which measured that more than half of new articles are machine-written, writes that "these articles largely do not appear in Google and ChatGPT". In Google, brands cited inside the AI summary get 2.07% clicks per impression against 0.94% for brands that are not cited (Seer Interactive, 53 brands, 2.43 billion impressions). And visits from ChatGPT to B2B company websites grew from 645,000 to 2.6 million a month within a year (Demandbase, 11 billion visits, June 2025 to July 2026).

Buyers describe the same thing in their own words. Under a post about AI letting everyone publish ten times more, an agency owner wrote: "Volume without judgement just trains the market to ignore you." Under a post about bot comments, a company owner wrote that AI personalization lines in inboxes "all read identical now, kinda makes real outreach stand out more" (first post, second post).

Why we conclude this.

Four companies measuring different channels on different customer bases show the same sign: the smaller and more targeted the effort, the higher the response. In the cleanest outreach dataset, the measures of access (acceptance, replies to messages) held while the response to the request itself fell. Expandi itself writes that its dataset does not isolate the cause, and names look-alike connection notes, changes in how LinkedIn shows them and audience habituation as possible ones. Our reading is that noise grows in volume and the response goes to whatever stands out from it. The conclusion does not say that online is easy: the falls in organic reach, in clicks from search and in replies to templates are real.

3. The exhibition hall carries the same noise

Under a post addressed to event teams, a marketing lead wrote: "At RSA this year, every booth sounded like the same company. And they were all experts in AI." A business owner under the same post: "At this point, 'AI-powered' on a booth sign tells me about as much as 'electricity-powered.' Great, but what do you actually do for me?"

Peep Laja, who founded Wynter and runs its annual conference, explains what an event still offers: "You never have to travel for content anymore (every piece of content ends up online), so the only point to attend an event is the other people there." Emily Kramer of MKT1, in her newsletter "Dinners are the new trade shows", calls conference booths "dollars in trashcans" and quotes Sheena Badani of RevAura on aiming for "~12 people who are actually in your pipeline" instead of "500 badge scans from people who will never respond."

An event does not take a company out of the noise. What it adds is a conversation with particular people, and that conversation is worth having with people who came for a reason.

4. In First Page Sage's table, trade shows are the most expensive in-person channel per customer

The only open table that compares the cost of a customer across online and offline channels comes from First Page Sage, an agency that sells SEO. It covers campaigns its agency ran for about 120 client firms between December 2021 and November 2024, which it describes as high-performing campaigns, averaged over three years. Its interest lies with search, and its table should be read with that in mind. The table below shows 11 of its 20 channels; the others, from content marketing ($1,254) and podcasts ($1,472) to PR ($1,720) and TV advertising ($2,028), do not change the order of the in-person channels. The table measures what a customer cost to win; what that customer paid back is not in it.

The same agency puts a single trade show at $10,000 to $250,000, and in its conversion data 54% of leads from executive events become qualified, against 28% from conferences and 24% from trade shows. Several online routes cost more than a trade show: account-based marketing, an outbound sales team, basic SEO. Among the ways of meeting buyers in person, the stand is the most expensive, while networking and public speaking come out among the cheapest routes in the whole table.

The price of exhibiting keeps rising. The Exhibitor Advocate read 224 published exhibitor manuals and rate forms in 23 US cities: material handling now costs $2.28 a pound, 21.3% more than in 2022 and 9.5% more in 2025 alone against 2.7% inflation; electrical overtime labor rose 41.2% since 2022. Exhibiting in New York costs about 75% more than in Atlanta.

The largest dataset on what events bring comes from HockeyStack, an attribution platform that read the CRMs of 198 B2B software companies: 2,642,337 deals from January 2024 to September 2025, with equal credit to every marketing touch. In-person and virtual events together account for about 6.5% of closed deals. Opportunities from in-person events reach qualification in 5.50% of cases, against 4.82% for other channels, and close in 12.1% against 11.1%. HockeyStack's own conclusion is "Events drive quality, not quantity", and its earlier analyses found that opportunities from events carry larger contracts and shorter sales cycles. The share of deals says nothing about payback by itself, because the share of budget those companies spent on them is not published. For mid-market companies, which HockeyStack defines as 100 to 999 employees, 2–3% of event opportunities reach qualification, the weakest result by company size.

Those who pay for events judge them coolly. In Forrester's survey of more than 400 event decision-makers (published June 2026), only 38% of CMOs agree that events perform well, nearly 70% of organizations are cutting the number of events they run, and only 18% plan more large hosted events. In a survey by Vendelux, a platform that sells event tooling, of 120+ marketing and event leaders, 86% of teams cannot attribute event ROI.

Two leaders have published what their own events cost. Peep Laja's 2017 conference sold out with 450 attendees, a $520,000 budget committed to the hotel a full year ahead; then the card processor froze ticket revenue for six months, and he lent his personal savings to the company to make payroll. Tycho Luijten of Dapper spent €15,000 and hundreds of hours on an event for 70 invited people, and judges it by how many people remember it.

Why we conclude this.

An agency's cost per customer across its clients and an audit of exhibition price lists, which has no interest in either side, point the same way: exhibiting at a trade show is the expensive way to meet buyers in person, and its price keeps rising. The agency's interest favours search, and even in its table the small formats where people meet come out among the cheapest. The attribution data add that events bring fewer deals and better ones. The conclusion concerns exhibiting at a trade show; meeting buyers in person comes out well in the same data.

5. Neither route is fast. Online work has a fast part and a slow part

A trade show takes 7 to 9 months to produce a result in First Page Sage's ranking of channels. Search is slower to start than most expect: Ahrefs followed millions of new pages and found that only 1.74% reach Google's top ten within a year; in a second sample of 2 million non-empty English pages, which Ahrefs considers the fairer figure, the share was 6.11%. Of the pages that do reach the top ten, 40.82% get there in the first month. A search campaign breaks even between 5 months (construction) and 14 months (legal services), 7 to 10 months for most industries, with the peak in the second or third year.

Precise outreach brings the first meetings fastest. In Saleshandy's 30.31 million emails, the best campaigns book 2 to 3 meetings per 100 emails sent, and 57% of positive replies come from follow-ups, which is why the result depends on a sequence being run to the end. None of these sources supports a breakthrough online within a quarter. What they support is outreach for the first conversations while content and search build up.

6. Both routes leave something behind, and both need the work to continue

Pages that rank in Google tend to be old: 72.9% of pages in Google's top ten are more than three years old, and the average number-one page is five years old. This measures the age of the page, not how long it has held its position. Content without upkeep fades fast: Fractl followed 719 articles for 28 months and found that a typical piece peaks four months after publishing and loses about 76% of its peak traffic within three months of that peak; how-to pieces keep about 30% of the peak a year later. Visibility in AI answers fades faster still: Matt Heinz cites Profound's finding that a typical page cited in an AI answer loses half its visibility in 11 days.

The broadest evidence on what accumulates comes from Les Binet and Peter Field's analysis of 996 campaigns in the IPA Databank: short-term activation gives a quick lift that fades, while brand building accumulates, with long-term effects overtaking after roughly six months. For B2B, their work with the LinkedIn B2B Institute puts the balance at 46% brand to 54% activation.

Events leave something behind as well. In a field study with industrial distributors, Smith, Gopalakrishna and Smith (2004) found that follow-up selling was more productive when the customer had seen the product at a trade show. After Tycho Luijten's event for 70 people, more than 40 guests posted about it. What nobody has published is the size of that effect and how long it lasts: the one academic study designed to measure it, Sridhar, Voorhees and Gopalakrishna (2015) on 2,299 attendees of ten trade shows, keeps its effect sizes behind a paywall.

How much an event leaves behind depends on what happens around it. In Forrester's survey of more than 400 event decision-makers, 62% of organizations whose event platform is connected to their sales and marketing systems are satisfied with the results, against 37% of those without that connection.

Why we conclude this.

On the online side, durability and decay are both measured on large samples: pages that rank are mostly years old, and content without upkeep fades within months. On the event side, the effect is there, and its size and duration are unmeasured. Both point to the same practical conclusion. Content needs upkeep, and an acquaintance made at an event needs further conversations. The result builds up where that continuing work happens, and steady work online is what keeps it going between one meeting and the next.

7. The shortlist is made before the first conversation

In 6sense's survey of more than 4,000 B2B buyers (November 2025), 94% of buying groups ranked their preferred vendors before first contact, and 77% bought from that early favorite. The typical buying cycle is about ten months, with around 16 interactions per buyer with the vendor that wins. In Gartner's survey of 645 B2B buyers (August–September 2025), 67% prefer to buy without a salesperson and use seven sources of information on average; 69% turn to a salesperson to check what AI told them. Forrester counts 13 people inside the buying organization and 9 outside influencers in a typical decision.

G2 asked 1,076 B2B software buyers in March 2026 what shapes their shortlist: AI chatbots (54%), software review sites (43%), market research firms (36%), the vendor's own site (36%), peers and colleagues (33%), independent forums (30%), thought leadership (28%). Events are not among the options. In TrustRadius's survey of 1,862 buyers (January 2026), 74% rely on reviews and conferences are not mentioned. Both surveys cover buyers of software, and the absence of events from the G2 list is partly a choice made by whoever wrote the questionnaire.

A first acquaintance can start anywhere. A buyer under Rand Fishkin's post on "how did you hear about us" described one common path: "Ask an accountancy client and they say Google, because Google is where they typed the firm name after someone at a networking lunch told them about it." The name came from a lunch, and the check happened in search. In G2's survey, peers and colleagues shape a third of shortlists, and that word can travel in person as well as online.

Why we conclude this.

Large buyer surveys from different vendors agree that buyers rank the candidates before they speak to any vendor, using sources they read on their own. The first mention of a company can come from a lunch, a colleague, a conference or an answer from an AI assistant. After that the buyer checks the company alone, and steady presence helps the name survive the check: the website, reviews, mentions on other sites, answers from AI assistants and the company's own content let the buyer verify it, remember it and keep it on the list.

8. None of the 14 leaders argues for moving offline

We read all 332 posts of the 12 leaders in our collection and searched the open web for all 14, including the two whose posts we had not collected. Not one of them argues for replacing online work with events. Each of them runs an online engine (a newsletter, LinkedIn, a podcast, original research, free online sessions), and six of them added an in-person format once they already had an audience:

• Dave Gerhardt (Exit Five) runs ticketed meetups of about 100 marketers and a 400-person event in Vermont, on top of a newsletter, a podcast and LinkedIn;
• Matt Heinz has hosted in-person CMO breakfasts for ten years, alongside a weekly online meeting of the same community;
• Emily Kramer runs small monthly dinners next to a newsletter-first business;
• Peep Laja's Spryng is capped at about 200 people, with no booths;
• Ross Simmonds is launching a conference for 200 marketers, "small on purpose", while three of his nine open roles are for Reddit and three for AI search;
• Tycho Luijten's event for 70 people exists, in his words, to be remembered, and its payoff was the 40+ people who posted about it.

Rand Fishkin, whose company measures where audiences pay attention, paused his own conference for 2026 and speaks at other people's. Leon Abboud states the long-term logic most directly: "Paid ads is renting attention, and rent goes up. An owned audience is building an asset that grows in value over time."

Only one leader puts the stage of a company into numbers. In Kyle Poyar's survey of 300 B2B software and AI companies (October 2026), network and introductions bring the most growth up to $10M in annual recurring revenue; intimate events enter the playbook at $1–10M and become the main play at $10–50M, where AI search also enters the top four, ahead of classic SEO. Matt Heinz reports a marketing leader, Michele Hayes, moving budget from corporate programs into smaller regional events, because without that human connection "we're kind of noise." Both describe small rooms, and neither describes leaving online.

9. Buyers ask how to be heard. None of them asks about going offline

In all 2,287 buyer comments, the word "offline" does not appear once, and "trade show" does not appear either. Buyers talk about the bar having moved up. The mechanism they describe most often is this one, written by a company owner under Rand Fishkin's post about measuring LinkedIn:

"A founder posts consistently for three months, LinkedIn referral traffic looks flat, and they conclude it isn't working. Then you check branded search and direct enquiries and both have climbed. The buyer saw the posts, remembered the name, and came in through Google or WhatsApp a few weeks later."

One buyer asked the question this article answers, under a post about B2B software marketing: "I'm curious how many B2B teams would see different results if they shifted some ad spend into creating spaces people actually want to be part of." Nobody replied with a number. Another wrote under a post on investing in brand: "yes let's judge our marketing after 3 weeks when our buyers take 9 months on average... makes total sense".

The buyers were identified by a model reading each commenter's profile headline: owners, chief executives and heads of marketing, sales, revenue and operations. A headline does not show company size, so this is the voice of people in the right roles at companies of unverified size. Most of the X replies came from solo founders, and we weight them low.

Where the answer changes

Company size and stage

In Kyle Poyar's survey of software companies, small events gain importance between $1M and $50M in annual recurring revenue, the range where most companies with 20 to 200 people sit. In HockeyStack's data, events work best for small companies of 1 to 99 employees (11% of opportunities from in-person events reach qualification) and worst for mid-market companies of 100 to 999 (2–3%). A company with 20 to 200 people falls across both groups.

Industry

Dave Gerhardt puts it plainly: "When your buyers are in wholesale distribution, manufacturing, and logistics, they aren't scrolling LinkedIn all day. They're at the trade show and the association dinner." Forrester found that nearly half of respondents in manufacturing and in professional, business and technology services give events more than 30% of their program budgets. In such industries a trade show stays part of the plan, and work online decides how many visitors come to the stand already knowing the company.

Sales cycle and the number of people deciding

With a ten-month cycle, 16 interactions with the winning vendor, and 13 people inside the buyer plus 9 outside, a single event cannot carry the relationship. The longer the cycle and the more people involved, the more the result depends on presence that runs for months.

Geography

The events our leaders run take place in hub cities: New York, Austin, Toronto, Boston. Exhibiting in New York costs about 75% more than in Atlanta, and for a team outside those cities every in-person meeting adds travel.

What nobody has measured

An online system has never been measured against an events program head to head for a company with 20 to 200 people; any claim that one overtakes the other in a given number of months is an estimate. Nobody has published how long the effect of an event lasts. The figures that circulate for small dinners and round tables (cost per guest, meetings per dinner) come only from companies that sell such events, without samples or periods. The time from zero to a first inquiry has not been measured for a newsletter, a community or a founder's personal content. Most large datasets cover software companies or corporations with billions in revenue. We built the answer from what has been measured and drew no conclusion from these gaps.

Our answer for a company with 20 to 200 people

Where the next budget should go depends on where your buyers make their choice and on what they find when they check you.

1. If a buyer who hears your name would find a weak website, few reviews and no mentions elsewhere, start there. Every route in this article, a lunch, a show or an AI answer, ends with the buyer checking the company alone. Gilles Argivier puts the risk plainly: "More demand does not repair a weak growth system. It amplifies whatever is already there." Rand Fishkin adds that the website "has not become less important. If anything, the information there, and the experience from landing to conversion, is MORE important in the Zero Click Era."
2. If your buyers find and check suppliers on their own, put the main investment into steady, precise work online. Real lists of the people who buy from you, with outreach for the first conversations. Content built on your own data, cases and judgment, which AI cannot produce for anyone else. Presence where shortlists form: answers from AI assistants, review sites, mentions on other people's sites and your own website. All of it kept up for months, and measured as buyers behave: branded search, direct inquiries and the quality of first conversations alongside clicks, since last-click attribution credits the last touch for work that happened months earlier.
3. Once you have an audience, add your own small event. A dinner or round table for 10 to 30 people invited from those who already follow your work, prepared online weeks in advance and followed up online afterwards. In Vendelux's survey, the strongest event programs start outreach 8 to 12 weeks ahead, while 55% of teams start less than four weeks before.
4. If your buyers do business at trade shows, keep the show and build the work around it. Dave Gerhardt's description fits many industries: buyers in wholesale distribution, manufacturing and logistics "aren't scrolling LinkedIn all day. They're at the trade show and the association dinner." Forrester found that nearly half of companies in manufacturing and in professional, business and technology services give events more than 30% of their program budgets. For such a company the decision is how to get more from each show: meetings arranged with the right people weeks before it, a website and content that visitors can check afterwards, and follow-up that keeps the conversations going. Forrester's 62% satisfaction among organizations that connect events to their sales and marketing systems, against 37% without, is the measured case for that.

Where Ayant-X fits

This kind of online work runs on a long stretch of routine work: research, publishing, search pages, campaigns, checking what moved and keeping everything current. Inside a busy team of 20 to 200 people, this is the work that slips first. Ayant-X builds and runs this system for B2B companies across the website, search, social media, advertising and visibility in AI answers. Our agents carry the routine on a schedule, and our team configures the system for each client and checks the quality of its work. Events stay with the client; the website, search and content that buyers check before and after a show are part of the system we run.

Sources

Noise by channel

• Pangram Labs, AI in Your Feed: 1,002,627 posts, from 24 April 2026, published 9 July 2026; PPC Land on the LinkedIn share
• Graphite: 43,000 Common Crawl articles, January 2020 – May 2025, published 14 October 2025
• Pew Research Center: 900 US adults, 68,879 searches, March 2025
• SparkToro: US Google searches, early 2026 against 2024
• Socialinsider, LinkedIn Benchmarks 2026: 1.3M posts, 16,645 pages, 2024–2025; Social Media Reach Statistics 2026: 872,075 posts
• Metricool, X statistics: 1,123,528 posts, 15,116 accounts, 2024 against 2025
• Expandi, LinkedIn Outreach Benchmarks 2026: 13,218,869 requests, May 2025 – April 2026
• Belkins: 7.5M cold emails, 2025; Woodpecker: 20M+ emails; Saleshandy: 30.31M emails, July–October 2026
• Seer Interactive: 53 brands, 5.47M queries, January 2025 – February 2026
• Demandbase: 11 billion visits, June 2025 – July 2026
• Meta, results for Q4 and full year 2025 (SEC filing); Tinuiti, Digital Ads Benchmark Q4 2025; Dreamdata, LinkedIn Ads Benchmarks 2026

Cost, time and what lasts

• First Page Sage, CAC by channel: ~120 client firms, 2022–2024, updated 18 June 2025; lead generation channels ranked; lead-to-MQL rates; SEO ROI statistics
• The Exhibitor Advocate, 2025 Annual Survey of Exhibition Rates: 224 rate documents, 23 US cities, published 12 May 2026
• HockeyStack, The State of Event Marketing in 2025: 198 companies, 2,642,337 deals, January 2024 – September 2025
• Forrester, Q1 2026 State of B2B Events Survey: 400+ event decision-makers
• Vendelux, 2026 B2B Events Survey: 120+ marketing and event leaders
• Ahrefs, How Long Does It Take to Rank: 1M and 2M URL samples, published 15 May 2025
• Fractl, Content Decay: 719 articles, March 2024 – June 2026
• Binet and Field, The Long and the Short of It (IPA); LinkedIn B2B Institute, The 5 Principles of Growth in B2B Marketing
• Sridhar, Voorhees and Gopalakrishna (2015), Customer Needs and Solutions 2(3)

How buyers choose

• 6sense, 2025 B2B Buyer Experience Report: 4,000+ buyers, published 12 November 2025
• Gartner buyer survey and its June 2026 release: 645 buyers, August–September 2025
• Forrester, The State of Business Buying 2026
• G2, 2026 AI Search Insight Report: 1,076 buyers, March 2026
• TrustRadius, 2026 B2B Buying Disconnect: 1,862 buyers, January 2026
• Smith, Gopalakrishna and Smith (2004), International Journal of Research in Marketing 21

Leaders

• Peep Laja: on conferences, on the 2017 conference
• Emily Kramer: Dinners are the new trade shows
• Matt Heinz: B2B Marketing Meet-Up, Palo Alto, on AI-answer visibility decay
• Kyle Poyar: channels by company stage
• Dave Gerhardt: NYC meetup, on buyers in distribution and manufacturing
• Ross Simmonds: Citation Conference
• Tycho Luijten: on measuring events by memorability
• Rand Fishkin: SparkTogether
• Leon Abboud: on owned audience