This post was sponsored by Expert Callers. The opinions expressed in this article are the sponsor’s own.
Why did my inbound traffic drop suddenly?
Are B2B leads dropping everywhere?
Is AI Overviews hurting B2B leads?
For two decades, the logic was simple: more traffic meant more leads, more leads meant more revenue. In 2026, that logic is breaking down.
In this article, we’ll dig into how this change might be a fix in disguise.
In This Article
Why Inbound Traffic Volume Dropped (But Your Deals Got Bigger)
Generative AI has effectively absorbed the early research phase of the buying journey. It’s still happening, just not on your site.
Where Your Organic Top-Of-Funnel Traffic Actually Went
AI Overviews and other LLM-based answer engines now synthesize information from across the web to answer top-of-funnel (TOFU) questions right on the search engine results page (SERP).
When a procurement manager searches for ‘best CX outsourcing vendors for mid-market SaaS,’ they increasingly encounter AI-generated summaries, recommendations, and curated results before reviewing traditional search listings. They get a synthesized shortlist, with vendor summaries drawn from across the web: case studies, reviews, analyst mentions, editorial coverage. The buyer forms a view, often a near-final one, before ever visiting a vendor’s website.
Why Some B2B Brands Are Still Getting Clicks
Seer Interactive’s 2026 AIO study, spanning 5.47M queries and 2.43 billion organic impressions across 53 brands, found that brands appearing on AIO-present SERPs but not cited within the AI Overview saw organic CTR fall 67% over 2025 (Seer Interactive, 2026). Brands that were cited in the AIO earned +120% more organic clicks per impression than their uncited competitors on the same SERP. The gap between cited and non-cited brands, not a universal traffic collapse, is the operative dynamic.
Seer’s 2026 update also found early signs of CTR stabilization in Q1 2026 after 18 months of decline. The recovery is accruing to cited brands. The structural pressure on non-cited brands hasn’t reversed; it’s simply stopped getting worse at the same rate.
Most Of B2B Research Now Happens Before You See A Lead
Generative AI has become buyers’ primary research method, returning a vendor shortlist before they visit any website (Forrester, 18,000 buyers, 2026). 80% of the B2B buying journey now happens without vendor involvement. By the time contact is made, the shortlist is largely settled.
Why A Smaller Pipeline Is Probably A Better One
AI models synthesize vendor credibility signals such as case studies, third-party citations, verified reviews, editorial coverage.
Through this process, AI models surface the vendors with the strongest corroborated presence. Low-credibility vendors don’t rank lower in this environment. They get bypassed at the research stage entirely, before a buyer ever forms an intent to click.
The funnel hasn’t disappeared. The top of it has. What remains is a filtered pipeline: buyers who arrive having already concluded their vendor research, carrying a procurement decision rather than a discovery question.
How To Get Cited By AI So The Right Buyers Find You
AI decides which vendors show up in a buyer’s search before that buyer ever clicks. These five steps make sure you’re one of them.
Step 1: Audit Where You Show Up (And Where You Don’t) (Weeks 1–2)
Pull your landing page data.
Pull your top 50 organic landing pages from Google Search Console over the trailing 90 days. Record the query cluster, query type (informational/navigational/transactional), and CTR for each. High impressions with low CTR on transactional queries is a credibility problem, not a visibility problem, and it requires a different fix.
Build a third-party mention inventory.
Use both Ahrefs and SEMrush; they return different datasets, so you need both. Export, deduplicate, and classify each external mention: editorial, directory, review, analyst citation, forum, social. Calculate your ratio of earned mentions (editorial, analyst, verified review) to unearned. For most B2B services companies, this ratio is significantly worse than expected.
Benchmark against 3 competitors.
Build a gap table: publications that cite them but not you, review platforms where they’re established and you’re absent, analyst reports that name them. That gap list is your outreach target list for Steps 3 and 4.
Audit your AI surface manually.
Open ChatGPT, Claude, and Perplexity. Run six to eight queries as a buyer would: “best [service category] for [client type],” “compare [service] vendors.” Screenshot every response. Note whether you appear, how you’re described, which competitors surface consistently, and which sources appear to be shaping the response. Repeat this audit every quarter.
Step 2: Fix Your Case Studies So AI Can Actually Read Them (Weeks 2–8)
A credibility-grade case study requires:
- A named or specifically described client
- A quantified baseline (“average handle time was 8:42 and CSAT was 61%,” not “they were struggling”)
- A specific description of work performed including key decisions made
- A defined timeline
- Outcomes in absolute terms, not percentages alone
- A client quote on the specific outcome, not a generic endorsement
- A named author with a linked professional profile
Most companies fail on components 1, 2, 3, and 7. Anonymous case studies with vague outcomes carry minimal weight with search algorithms or AI models.
Production process.
Identify your five to ten strongest outcomes from the past 24 months. Schedule 45-minute structured interviews with both the client contact and your internal delivery lead. Use a fixed template that forces numbers: metrics before, metrics after, what changed and when. Assign a named senior author to write each one, a real person with an existing professional presence, not a generic company byline. Get written client sign-off on public metric citation. Publish with Schema.org markup and submit each URL for indexing via Google Search Console immediately. Done properly: three to four weeks per case study from interview to publication.
Step 3: Get Bylines In Publications AI Trusts (60–90 Days)
Build your target list from real bylines.
Build a media target list from actual bylines published in the last 90 days, not a PR database. For CX outsourcing, that means Customer Think, ICMI, Contact Center Pipeline, CX Today. Build a tracking spreadsheet with name, publication, recent topics covered, and a note on the angle most relevant to their specific beat.
Write a three-paragraph pitch.
Each pitch is three paragraphs: why this story fits this editor’s beat right now; what the story is in one sentence; what you’re offering (data, interview, exclusivity). Send individually. Follow up once at seven days. Expect 10–15% positive response. For five placements in a quarter, plan 35–50 individual outreach contacts.
Link every placement back to your site.
After every placement, add it to a press page on your site and link back to the original. The cross-referencing strengthens the credibility signal in both directions.
Step 4: Get Reviews On The Platforms AI Cites (Weeks 4–6)
Prioritize the platforms AI actually cites.
Prioritize review platforms that appeared in ChatGPT, AI Mode, Claude, or Perplexity’s citations for your category during your Step 1 audit. Assign review outreach to account managers, not marketing; the request carries more weight from the relationship owner. Send a personalized email with a direct link to the submission form, not a homepage. No incentives: platform policies prohibit them and flagged reviews are removed.
Build review requests into your delivery process.
Expect 30–40% conversion on warm personal outreach. Build the request into your delivery process at 90 days post-engagement and at project completion.
Respond to every review within 72 hours.
Including critical ones. A specific, considered response to a negative review is itself a credibility signal; it demonstrates that a real person is accountable for outcomes.
Step 5: Make Your Authors Verifiable Across The Web (Weeks 1–4)
Set up each author’s identity trail.
For every team member who will author content: update their LinkedIn profile with specific expertise domains and a verifiable career history; create an author bio page on your website that links to their LinkedIn and describes their specialization in concrete terms; ensure all content they produce links back to that bio page; and when external placements land, include a link to their company author page in the byline.
Why this matters for AI and search.
This creates a verifiable identity trail across multiple web properties. A search engine or AI model encountering content by a named individual can cross-reference that identity across your website, LinkedIn, and external publications, and interpret the consistent pattern as genuine subject-matter expertise. Without this infrastructure, even strong content returns a fraction of its potential credibility signal.
How To Coordinate This AI Search Strategy
Running these workstreams in parallel requires at minimum: a content strategist capable of structured interviews and external publication drafting; an account management resource for review outreach; a senior subject-matter expert available for media interviews and author attribution; and a project coordinator managing client approvals across multiple case studies simultaneously. At a well-resourced company with existing content and PR capability: four to six months to measurable movement. Building from scratch: six to nine.
SEO used to reward visibility. It now rewards credibility. And only one of those compounds.
In the AI-search era, the companies with the strongest digital credibility may attract fewer visitors, but they will increasingly attract the right ones.
Image Credits
Featured Image: Image by Shutterstock. Used with permission.