Is AI traffic a red herring? + the 2026 GTM channel scorecard

Your weekly summary of tactics and trends. Trusted voices only.

Week of Nov 2, 2026 · 5-minute read

Top of mind

We read 664 posts from 102 SaaS marketing experts over the last two weeks. Here's what got folks talking:

1. Does AI visibility pay off? About 1 in 7 posts were about AI search, and the question has shifted from "how do I get cited?" to "what is getting cited worth?" Three of this week's tactics are about it (2, 5 and 6). Kevin Indig and Rand Fishkin disagree on whether chatbot recommendations pay off (the Debate).
2. AI agents in go-to-market. AI SDRs and agentic workflows were everywhere, but they aren't paying off yet. AI SDRs landed in the bottom tier of Kyle Poyar's channel survey (Tactic 1), and only 7% of marketing job posts mention working with AI agents (Emily Kramer, in the Spotlight).
3. Is content marketing over? Kieran Flanagan says blog-and-form content is done, and creators, founder voice and original data are taking its place. B2B brands pay creators $1–15k per collab (Quick hits).

New tactics and ideas

Cool ideas we ran into this week, and how we'd try them

1. The 2026 GTM channel scorecard

Kyle Poyar · Oct 23
300 B2B companies ranked their channels by ROI. Top tier: network and intros, exec dinners, and AI search (the fastest grower). Bottom: PR and AI SDRs.
→ Read Kyle's post · Full report
What we'd try: check next year's channel plan against the top and bottom tiers before budgets lock.

2. Five credibility statements = 62% more AI visibility

Ross Hudgens (Siege Media) · Oct 26
Across 520 tracked companies, mid-authority brands with 5+ provable claims on their homepage (review counts, case-study results, analyst placements) had 62% more AI visibility than those with none.
→ Read Ross's post
What we'd try: count the provable claims on your homepage. Aim for five, written as text rather than logo images.

3. Your pricing now goes stale in 6 months

Kyle Poyar · Oct 20
From advising 20+ Series A–D founders: pricing used to last about 18 months and now lasts about 6. Your freshest pricing data is sitting in your sales call recordings.
→ Read Kyle's post
What we'd try: run a monthly AI pass over call transcripts to pull out pricing objections, and track what changes month to month.

4. Links still kill social reach, at least for Rand

Rand Fishkin (SparkToro) · Oct 19
Two new studies say links barely hurt reach. Fishkin's own last ~100 posts say otherwise: posts with links got about 1/12th the impressions.
→ Read Rand's post
What we'd try: compare your own posts with and without links before you change how you post.

5. ChatGPT is warning users about self-ranking comparison pages

Lily Ray · Oct 19
In ~700 ChatGPT answers to B2B software prompts, 19 warned users about who wrote the rankings, including vendors who rank themselves on their own comparison pages.
→ Read Lily's post
What we'd try: if your "best X tools" page puts you at #1, back it with outside proof like reviews and analyst reports.

6. AEO runs on two clocks

Ethan Smith (Graphite) · Oct 21
Cited answers can change in days. Answers that come from a model's training data take at least 2–5 months, until the next model version ships.
→ Read Ethan's post
What we'd try: when a new model ships, re-run your key prompts with web search turned off and compare against the previous version.

Debate: Do chatbot recommendations pay off?

Exposure is what counts

• Kevin Indig: AI referral traffic grew 5.6x in the US over two years, but it's still about 0.5% of traffic (across 10k domains). He calls AI traffic a red herring. What matters much more, he says, is how often you're recommended.

Outside Google, it may not move the business

• Rand Fishkin: after reading three new studies, he's "less confident than ever" that more visibility in ChatGPT, Claude or Perplexity "would have much business impact for most companies." Google's AI Overviews, AI Mode and Gemini are another matter, which he thinks probably matter a lot.

Where they agree: both are asking what showing up in AI answers is worth. On Google, Cyrus Shepard's case study points to an answer: when a site opted out of Google's AI features, its regular organic clicks dropped too. For ChatGPT and the other chatbots, it's still an open question.

Quick hits

• 3 of 100: Andy Crestodina's team tested 100 free "AI visibility audit" tools, and only 3 asked about your buyer.
• $1–15k: what B2B brands pay a creator per collab, depending on audience size, from Kyle Poyar with beehiiv's ad data.

Expert spotlight: Emily Kramer

Co-founder of MKT1 and writer of the MKT1 Newsletter. This week she analyzed 1,100+ marketing job posts from 2026. Only 47% ask for AI skills, and only 7% mention working with AI agents.
→ Read Emily's post
One interesting opinion: job descriptions are years behind how marketing teams work today.
Follow her for: data-backed B2B marketing team building, hiring and planning.