Every hiring conversation I've had in the last six months follows the same arc. The strategy questions — positioning, segmentation, ICP — get nodded through in ten minutes. Then comes the real interrogation: "LinkedIn feels dead. Cold email feels dead. So what's actually working right now?"
It's a fair question. It's also a dangerous one.
Here's the thing about the "what's working" question: the moment a channel earns that label, the clock starts running out on it. That's not cynicism. That's the Power Law at work — and if you don't understand why it applies to GTM, you'll be chasing tactics forever.
Let me explain what I mean, then get into what I actually think is working in mid-2026 and where the next real leverage is coming from.
Why Channel-First Thinking Will Get You Fired
A 195-company survey published late 2025 by Kyle Poyar and Maja Voje identified six dominant GTM motions and found, perhaps surprisingly, no correlation between GTM motion and growth rate. Every motion can work. No motion is automatically superior.
That's the buried lede that most teams miss entirely.
The same report found that the average B2B software company runs 5 core GTM channels plus 5.5 experiments on top of them. That's 10+ channels in flight simultaneously — mostly at small teams without the headcount or data infrastructure to know which one is actually driving closed revenue.
This is the channel trap: the solution to channel saturation is not more channels. It's better signal discipline.
The Three Laws You Need to Internalize First
Before any channel conversation, every revenue leader should be able to articulate three dynamics that govern channel performance over time.
1. The Power Law Doesn't Lie
In any given company, at any given stage, one or two channels will produce a disproportionate share of qualified pipeline. Not 80/20. In modern B2B, it's closer to 95/5. The problem is that most companies don't know which channel is their 95 because their attribution models can't see dark social, can't handle multi-touch accurately, and can't account for the two podcast episodes a champion listened to before ever raising their hand.
Chasing 10 channels isn't optionality. It's a refusal to do the measurement work required to find your power channel and go all-in on it.
2. Channel Decay Is Accelerating
Every channel follows a predictable arc: Discovery → Early Adopter Advantage → Mainstream → Saturation → Decay. What's changed in the last three years is the speed of that arc. AI-powered tooling means the moment a tactic works, it's productized, templated, packaged into a Clay workflow, and distributed to 50,000 SDRs within a quarter.
Cold email didn't die because inboxes got stricter (though they did). It died because the median quality of cold email collapsed when AI made volume free. The same thing happened to LinkedIn connection-request outreach. The same thing is happening right now to AI-generated SEO content. The channel didn't fail — the signal-to-noise ratio in it did.
Channel decay is not an argument against channels. It's an argument for getting there earlier and building moats before the flood.
3. Tactic Fatigue Is a Buyer Problem, Not a Marketer Problem
Most GTM teams treat tactic fatigue as a creative problem. "Our sequences aren't personalized enough." "Our subject lines need work." This misses the root cause entirely.
Tactic fatigue is a buyer trust deficit. Decision-makers have been on the receiving end of enough AI-generated "personalization" — referencing their LinkedIn posts, mentioning their recent funding, congratulating them on their new role — that they've developed a sophisticated filter against it. The filter isn't conscious. It's reflexive. The moment a message pattern-matches against their prior experience of being sold to, it's gone.
The solution isn't a better subject line. It's a fundamentally different relationship to trust and signal.
The Contrarian Take: LinkedIn and Cold Email Aren't Dead
They're dead as broadcast channels. They're very much alive as precision instruments.
The companies still generating outbound pipeline from email and LinkedIn in 2026 share three characteristics:
- —They are operating with real intent signal stacks — not just one trigger but layered signals: job posting changes + G2 category page traffic + LinkedIn content engagement + funding event + technology install data. They're reaching out to the 3% of the market that is actively in-market, not the 97% who aren't.
- —They have human judgment in the loop — not AI SDRs. Human-identified, AI-drafted, human-reviewed and sent. The AI SDR failure rate the 2025 State of B2B GTM survey documented ("We tried an AI SDR for six months and were unable to generate a single opportunity") was predictable. Buyers can tell. The value isn't in the send; it's in the judgment of who to send to and why now.
- —They are leading with insight, not pitch — the outreach itself is the value. A genuine, specific, non-generic observation about the prospect's business that demonstrates you've done the work. Not "I noticed you're using Salesforce..." but something that reveals real situational knowledge.
What Is Actually Working Mid-2026
With that framing established, here's my honest mid-2026 channel and tactic assessment for growth-stage B2B SaaS.
High-Signal, Working Now
Intent Signal Stacking
The companies winning in outbound aren't using Clay to spray 10,000 contacts with "personalized" emails. They're using it to identify the 50 accounts per month where three or more signals align — and then doing genuine, human-quality outreach to those 50. Volume is the enemy. Signal precision is the advantage.
Warm Outbound Through Network Nodes
Not founder brand in the "post every day on LinkedIn" sense. Specifically: identifying the 20-30 people in your extended network whose word carries disproportionate weight with your ICP and engineering a reason for those people to genuinely recommend you. This is slower than any other tactic. The conversion rate is 5-10x higher. It doesn't scale with software — it scales with trust capital.
Intimate Events Over Large Conferences
The 2025 data confirms what practitioners have known for a while: the ROI on the $80K conference booth is a fiction maintained by vanity metrics. A dinner with 12 CFOs or a half-day workshop with 15 engineering leaders at the right companies outperforms it in pipeline per dollar by an order of magnitude. The constraint is curation, not cost.
Ecosystem and Integration-Led Distribution
Being native in your buyer's existing workflow is a moat that compounds over time. A Salesforce AppExchange listing, a native Slack integration, a Snowflake data share partnership — these are distribution channels that are hard to replicate quickly and generate warm, high-intent leads by definition. Underinvested by most mid-market companies.
Emerging, Worth Real Experimentation
AEO (Answer Engine Optimization)
This is not SEO with a new name. It's a fundamentally different discipline. Google SEO optimizes for crawlers and backlink graphs. AEO optimizes for LLM consumption — structured content, clear factual claims, authoritative sourcing, schema markup designed for machine readability. Your buyers are increasingly starting research in Perplexity, Claude, and ChatGPT. If your brand, product, and positioning don't surface in those answers, you don't exist in their consideration set before they've even Googled you.
The companies getting ahead of this are publishing practitioner-grade content that LLMs cite as authoritative — detailed technical comparisons, credible point-of-view pieces, third-party validation that makes it into training and retrieval contexts.
Dark Social Infrastructure
The attribution problem gets worse every year. A significant and growing percentage of B2B pipeline originates in conversations you can't track: Slack communities, private Discord servers, WhatsApp buyer groups, LinkedIn DMs, newsletter forwards. The right response is not to solve the attribution problem (you can't) — it's to earn presence in those spaces. Participating in the communities where your buyers talk to each other. Being the practitioner that gets cited in the private conversation you'll never see.
Community Distribution (Not Community Building)
Building a community takes 18-24 months to generate meaningful return. Most companies don't have that runway. The faster play is identifying the three to five communities where your ICP is already active and concentrated, and earning trusted distribution rights inside those communities — through genuine expertise, useful content, and consistent presence. You're not building the audience; you're borrowing trust from communities that already have it.
The Frontier: Agentic Discoverability and Buying
This is where I'd place the highest-upside, longest-horizon bet for the next 18-36 months.
The buying journey is being reconstructed. Buyers — particularly technical buyers and operators — are increasingly delegating early-stage research to AI agents. These agents browse, compare, synthesize, and shortlist. By the time a human is involved, the consideration set is already narrowed.
This creates a new category of GTM problem that almost nobody is solving for yet: being visible to, and trusted by, the agents your buyers are using.
Practically, this means:
- —Structured, machine-readable web presence — your website, product pages, and case studies need to be optimized for agent consumption, not just human reading. JSON-LD schema, clear factual claims, structured comparison data.
- —G2, Capterra, and third-party review profile health — agents pull from these sources. Stale, sparse, or poorly structured profiles are invisible in AI-assisted research.
- —API accessibility and developer documentation — technical buyers using agents often start with "can this tool be integrated" before "should I talk to sales." Your developer docs are now a top-of-funnel asset.
- —LLM-cited authority content — long-form, specific, well-sourced content that earns citations inside AI responses. Not for Google traffic. For the model that's doing your buyer's research for them.
The companies that build this layer now will have a compounding advantage in 18 months that new entrants cannot easily replicate, because LLM citations and agent trust are themselves subject to the Power Law — early authority compounds, late entrants fight for table scraps.
What This Means Practically for 2026 GTM Planning
A few principles I'd carry into any GTM leadership role based on all of this:
Audit your actual power channel first.
Before adding anything new, do the attribution forensics to identify where closed-won deals actually originated — including dark social proxies (survey customers at close, look for content engagement patterns). Find your 95% channel and protect it.
Set a channel decay monitoring cadence.
Assign someone to track leading indicators of decay in each active channel quarterly: response rates, meeting-set rates, CPL trends, organic share of voice. Treat declining indicators as a forcing function for experimentation before you're in crisis.
Separate experiments from motions.
Run 2-3 genuine channel experiments at small budget with clear success thresholds. Don't let experiments live indefinitely in a gray zone — they either graduate to motions or they get cut.
Invest in the AEO/agentic layer now, before it's a survival tactic.
The window to build early authority in AI answer surfaces is closing. This is the AEO equivalent of building domain authority in 2012 — the practitioners who started early had structural advantages that persisted for years.
Lead with signal discipline over volume.
In every channel, the question isn't "how do we reach more people" — it's "how do we reach the right 50 people with the right message at the right moment." That's a data and judgment problem, not a scale problem.
The hiring managers asking "what's working" deserve an honest answer. The honest answer is: the channels that are working are working because the teams using them have better signal quality, better judgment about when to deploy them, and a healthier relationship with buyer trust than the teams that are just running volume plays with better AI tooling.
The channel is rarely the problem. The discipline around it almost always is.