The state of GTM and AI tools now

Contents

Below we study:

• Categories of GTM tools and what’s popular now
• A end-to-end concrete example through a B2B SaaS
• Different distribution models and their limitations (and a PLG and sales led case study)
• Product case studies
• Current valuations and future winner predictions

Popular tools and their categories

Like other industries, GTM landscape has split into two camps: AI-native and legacy platforms with AI bolted on as an afterthought. By categories we have the following most popular tools.

Data augmentation & outbound prospecting

Clay: 100M ARR. AI native for outbound. Replaced 4 separate tools (ZoomInfo, Clearbit, PhantomBuster, and custom scrapers) in one spreadsheet-like interface that connects to 100+ data providers. GTM Engineer Club

Apollo: 150M ARR. popular mid-market/startup option. Apollo's free tier for outbound combined with Clay's starter plan for enrichment can give you enterprise-grade prospecting capabilities for under $200/month

ZoomInfo: 1B ARR. Still the enterprise standard for contact/company data, though increasingly challenged.

Prediction - zoominfo will die due to legacy, and Apollo and Clay will win

ABM

The fundamental insight is: in B2B, you don't sell to individuals, you sell to organizations. A deal at a large company involves 6-10 decision-makers on average. ABM acknowledges this reality and coordinates everything around the account, not the individual lead.

Company: 6sense.

Conversation intelligence & revenue forecasting

Gong

Attribution & Analytics

HockeyStack is gaining serious traction. HockeyStack is one of the few attribution tools that actually explains the "why" behind numbers. It caught a gap between the competitive existing tools.

Startup vs. Enterprise Stack

The Full Workflow, Step by Step

We now walk through a concrete, realistic example using a fictional enterprise software company.

The Company: "DataFlow" (B2B SaaS, $50M ARR, sells data pipeline software to mid-market and enterprise)
Target customer: Head of Data Engineering or CTO at companies with 200–2,000 employees
Deal size: $50K–$300K ACV
Sales cycle: 3–6 months
CRM: Salesforce
Marketing automation: Marketo
Key tools: 6sense, Clay, ZoomInfo, Outreach, Gong, HockeyStack, Default

STAGE 1: Identify Who to Target

Owner: Marketing + RevOps | Tools: ZoomInfo, 6sense, Clay

The RevOps team builds an Ideal Customer Profile (ICP) in Salesforce — companies in fintech or healthcare, 500–2,000 employees, using Snowflake or Databricks, growing headcount in data engineering.

• ZoomInfo surfaces a list of 4,000 companies that match this profile
• 6sense overlays intent data — it detects that 180 of those companies are actively researching "data pipeline" and "ETL tools" based on web behavior across thousands of sites
• Clay enriches those 180 accounts further — pulling in technographic data (what tools they use), recent funding rounds, hiring signals (are they posting data engineer jobs?), and finds the right contact at each company (Head of Data Engineering, VP of Engineering)
• All of this syncs into Salesforce as new Account and Contact records, tagged as "High Priority ICP"

Output: A clean, prioritized list of 180 high-intent target accounts in Salesforce, enriched with context.

STAGE 2: Marketing Warms Them Up

Owner: Marketing | Tools: 6sense, Marketo, Jasper/Copy.ai, LinkedIn Ads

The marketing team doesn't just hand the list to sales cold. They run a coordinated warming campaign first.

• 6sense runs targeted display ads specifically to those 180 companies — the Head of Data Engineering at a fintech in Austin starts seeing DataFlow ads on industry sites, LinkedIn, even generic news sites. They've never heard of DataFlow but now they've seen the brand 12 times.
• Marketo runs a nurture email sequence to any contacts at those accounts already in the database — a 4-email sequence over 6 weeks sharing a benchmark report, a case study, and a product webinar invite
• Jasper was used to generate the content assets (case studies, ad copy, email sequences) at scale
• 6sense tracks account engagement — it scores each of the 180 accounts and surfaces when an account crosses a threshold (e.g., 3+ people from the same company visited the pricing page)

Output: 40 of the 180 accounts are now "hot" — multiple people engaged, intent score spiking.

STAGE 3: The MQL → SQL Handoff

Owner: Marketing hands to Sales | Tools: Default, Marketo, Salesforce

A Head of Data Engineering at a fintech company (let's call her Sarah, at "FinCo") fills out a form on DataFlow's website requesting a demo.

• Default fires the moment the form is submitted — it enriches Sarah's record in real time (confirms she's the right persona, FinCo has 800 employees, Snowflake user, matches ICP perfectly), scores her as a high-priority lead, and routes her instantly to the enterprise AE (Account Executive) responsible for fintech accounts
• The AE gets a Slack notification within 60 seconds with full context: Sarah's title, FinCo's tech stack, the 6sense data showing 4 people from FinCo visited the pricing page last week, and the fact that FinCo has been seeing DataFlow ads for 3 weeks
• Marketo updates Sarah's lifecycle stage from MQL to SQL and logs the handoff in Salesforce
• A Salesforce Opportunity is automatically created for FinCo

Output: Sarah gets a reply from the AE within 10 minutes. The old world: she might have waited 2 days and gotten a generic response.

STAGE 4: Sales Outreach & Discovery

Owner: Sales (AE + SDR) | Tools: Outreach, Gong, Salesforce, Clay

The AE and their SDR partner pick up the account.

• Clay runs a quick research workflow on FinCo — pulls recent news (they just announced a Series C), LinkedIn activity from the data team, job postings (they're hiring 3 data engineers), and generates a personalized outreach angle: "Congrats on the Series C — scaling data infrastructure at that inflection point is exactly what our customers deal with"
• Outreach manages the multi-touch sequence: the AE sends a personalized email (AI-drafted by Clay/Outreach), the SDR follows up on LinkedIn, a call is scheduled
• Sarah replies and books a discovery call
• The discovery call happens over Zoom — Gong records and transcribes it automatically
• After the call, Gong's AI summarizes the key pain points Sarah mentioned (latency issues with their current pipeline, engineer time wasted on maintenance), flags that Sarah mentioned a competitor ("we've also been looking at Fivetran"), and automatically logs call notes to the Salesforce Opportunity

Output: Discovery complete. AE has clean notes in Salesforce, knows the pain, knows the competition, knows the timeline (Sarah said "we need something in place before Q3").

STAGE 5: Multi-Stakeholder Deal Management

Owner: Sales | Tools: Gong, Salesforce, 6sense, Outreach

Enterprise deals involve multiple stakeholders. Sarah's boss (the CTO) and the procurement team get pulled in.

• 6sense detects that the CTO's email domain has now visited DataFlow's security/compliance page — the AE gets an alert and proactively sends the CTO a security overview without being asked
• Gong analyzes subsequent calls and flags a risk: the word "budget" has come up 4 times and the CTO went quiet when pricing was discussed. The AE's manager is alerted to coach the rep before the next call.
• Salesforce tracks the opportunity stage, the multi-threaded contacts, all activity, and rolls into the CRO's pipeline forecast
• Outreach manages follow-up sequences to the procurement contact who went cold

Output: Deal is multi-threaded, risks are visible early, forecast is accurate.

STAGE 6: Closed-Won → Attribution & Learning

Owner: RevOps + Marketing | Tools: HockeyStack, Salesforce

FinCo signs a $120K ACV deal. Now the question is: what actually worked?

• HockeyStack traces the full journey: 6sense ads (12 impressions) → Marketo nurture email (opened the case study) → pricing page visit → demo form → closed in 94 days
• Marketing learns that the benchmark report email had the highest engagement and influenced 8 of the last 12 deals — they produce more of that content
• The 6sense ABM campaign is confirmed to have influenced deal velocity — accounts that saw ads closed 22 days faster on average
• All of this feeds back into Salesforce and refines the ICP and scoring model for the next 180 accounts

The full picture (end-to-end)

Every stage has a handoff — and every handoff is a place where deals used to die. A form that took 2 days to respond to. A rep who didn't know the prospect had been looking at the pricing page. A manager who didn't know a deal was at risk until it was too late. A marketer who had no idea which campaigns actually mattered.

The entire GTM tool stack exists to eliminate those gaps — making the handoffs instant, data-rich, and visible to everyone in Salesforce. The AI layer is what makes it possible to do this at scale without hiring an army of SDRs and analysts.

The famous Peter Thiel line captures it well: "A product without distribution is not a business."

Distribution Models and Their GTM Implications

1. Direct Sales-Led Distribution

How it works: You hire salespeople who go directly to customers. GTM tools involved: The full stack — ZoomInfo, Clay, Outreach, Gong, Salesforce When it makes sense: High ACV ($50K+), complex products, enterprise buyers DataFlow example: The entire workflow we walked through is direct sales-led distribution

2. Product-Led Distribution (PLG)

How it works: The product itself is the distribution mechanism — free tier, viral loops, self-serve. Users find, adopt, and expand without touching a salesperson. GTM tools involved: Product analytics (Mixpanel, Amplitude), in-app messaging (Intercom), usage-based alerts feeding into Salesforce for expansion plays When it makes sense: Developer tools, collaboration software, anything with a natural viral loop (Slack, Figma, Notion, Dropbox) Key insight: The "product" does the work that salespeople do in sales-led — it qualifies, nurtures, and converts on its own

3. Partner/Channel Distribution

How it works: Other companies (system integrators, resellers, consultants) sell your product on your behalf or bundle it with their own GTM tools involved: Partner portals, PRM software (like Crossbeam, PartnerStack), CRM deal registration When it makes sense: When you can't hire enough salespeople to cover the market, or when your product is better sold alongside a services engagement (think: Salesforce implementation partners) Key insight: This is how companies like Microsoft and Salesforce reached global scale — their partner ecosystems dwarf their own direct sales

4. Content/Inbound Distribution

How it works: SEO, YouTube, podcasts, newsletters, LinkedIn — you create content that attracts buyers to you organically GTM tools involved: HubSpot, Jasper/Copy.ai (creation), HockeyStack (attribution), Marketo (nurture once they arrive) When it makes sense: When your buyers are actively searching for solutions (high search intent), or when you can build an audience around a problem you solve Key insight: This is the lowest CAC (Customer Acquisition Cost) distribution at scale, but it takes 12–24 months to build and doesn't work for every category

5. Platform/Marketplace Distribution

How it works: You list on AWS Marketplace, Salesforce AppExchange, Google Workspace Marketplace, or similar — buyers find you through a trusted platform they already use GTM tools involved: Minimal — the platform does the distribution; you need good reviews, listing optimization, and a way to handle inbound When it makes sense: When your buyers already live in a platform ecosystem and procurement is easier through it (large enterprises can use existing AWS spend commitments) Key insight: Increasingly important — AWS Marketplace alone does tens of billions in software transactions annually

The "Distribution + GTM Tools" Connection

Here's where it gets interesting: your distribution model determines which GTM tools you need.

A common mistake companies make is building a sales-led GTM tool stack when they actually have a product-led distribution model — they hire SDRs and buy Outreach when what they need is better in-product onboarding and a usage-based expansion trigger.

Most successful companies in 2026 actually combine multiple distribution models simultaneously:

DataFlow's full distribution picture:

• Content — SEO blog posts, benchmark reports
• Paid distribution — 6sense ABM ads, LinkedIn
• Partner distribution — AWS Marketplace listing, Snowflake partner ecosystem
• Direct sales — AEs working the 180 target accounts
• Product-led — free trial that converts to paid

Each distribution channel feeds different parts of the GTM funnel and requires different tools. The RevOps/GTM engineering function exists largely to stitch all of these together into a single view in Salesforce, so leadership can see which distribution channels are actually producing revenue.

The GTM tools — Clay, Gong, Outreach, 6sense — are largely available to everyone. Your competitor can buy the same stack tomorrow. What they can't easily replicate is:

• A trusted brand built through 5 years of content
• A partner network of 200 system integrators who recommend you
• A viral product loop baked into the product experience
• A marketplace presence with 500 five-star reviews
• A community of 20,000 practitioners who evangelize you

This is why the best GTM strategists think about distribution first — before deciding on tools, headcount, or messaging. The tools execute your distribution strategy; they don't replace it.

The short version: distribution is the channel infrastructure, marketing is the message you put through it, sales is the conversion at the end of it, and the GTM tool stack is what makes the whole machine measurable and scalable.

PLG vs. Sales-Led: The Fundamental Difference

How PLG Relates to the GTM Stack
PLG doesn't eliminate GTM tools, it changes which ones matter and how they're used.
PLG is not "no sales," it's "sales at a different moment.". Most successful PLG companies eventually layer sales on top, creating what's called "Product-Led Sales" (PLS). The model looks like this:
Free/trial users → product usage signals → sales identifies high-value accounts → sales reaches out with context → enterprise deal closes

Tools That Become Central in PLG

• Product Analytics (Amplitude, Mixpanel, Pendo) replace Gong and Outreach as the primary signal source. Instead of listening to sales calls to understand deal health, you're watching in-product behavior to understand: Who activated (hit the "aha moment")? Who is stuck in onboarding? Who is using the product so heavily they're a natural expansion candidate?This behavioral data is the PLG equivalent of Gong's conversation intelligence. It goes into the CRM just like call notes would.
• In-App Messaging (Intercom, Pendo, Appcues) Instead of Outreach sending email sequences, PLG companies trigger contextual messages inside the product — "You've hit your free tier limit, here's what you unlock with Pro" or "3 of your teammates are on Slack — invite them to collaborate."
• Usage-Based CRM Triggers The CRM (Salesforce or HubSpot) gets fed product usage data instead of just sales activity. A Salesforce opportunity might automatically be created when a user hits a usage threshold — say, a team of 10 people actively using the free tier for 30 days. This is the PLG equivalent of a demo request.
• Self-Serve Infrastructure Stripe (billing), Auth0 (authentication), in-app upgrade flows — the "checkout" experience replaces the sales proposal. The product needs to be able to close deals without a human.

Slack, Figma, Notion, Atlassian all have significant enterprise sales teams. The difference is that sales doesn't initiate the relationship. The product does. Sales shows up after the product has already proven its value inside the account.

So the GTM tools used in PLG + PLS:

• Clay — still used, but instead of cold prospecting, it enriches accounts where product usage is already happening
• Outreach — still used, but reps are reaching out to warm accounts (people already using the product) not cold ones
• Gong — still used for enterprise expansion calls
• Salesforce — still the CRM, but now fed by product data not just sales activity
• 6sense — less critical (you have first-party usage data instead of relying on third-party intent)
• ZoomInfo — less critical for top-of-funnel, more useful for finding the economic buyer inside a company already using the product

The key metrics change completely:

The PQL (Product Qualified Lead) is the PLG equivalent of the SQL. It's a signal that says "this user/account has used the product enough that a sales conversation would be welcome and likely to convert." Defined differently per company — might be "team of 5+ active users on free tier for 21+ days" or "used the API 1,000+ times this month."

PLG is fundamentally a distribution strategy before it's anything else. The best PLG products have a viral loop baked in:

• Collaboration virality (Figma, Notion, Slack)
• Network virality (Calendly, Loom, DocuSign) → Every time you use the product with someone outside your company, they see it and want it
• Integration virality (Zapier, Clay)
• Community virality (GitHub, Hugging Face)

Each of these loops means the product generates its own distribution without paying for ads or hiring SDRs. This is why PLG companies have structurally lower CAC than sales-led companies and can grow faster with less capital.

PLG and the DataFlow Example

Let's bring it back to DataFlow, the company from our earlier walkthrough. Imagine DataFlow adds a PLG motion:

Before (pure sales-led): DataFlow identifies FinCo through ZoomInfo → runs ABM ads → Sarah fills out a demo form → AE closes $120K deal in 94 days

After (PLG layer added): DataFlow launches a free tier — connect up to 2 data sources, process up to 1M rows/month. A junior data engineer at FinCo signs up on a Tuesday afternoon without telling anyone. She builds a pipeline. It works great. She invites 2 teammates. Within 3 weeks, 8 FinCo engineers are active on the free tier.

Now the GTM motion changes:

• Salesforce gets a PQL alert: "FinCo — 8 active users, hitting free tier limits, 3 data sources connected"
• Clay enriches FinCo: 800 employees, Series C, matches ICP perfectly
• AE reaches out to Sarah — but now the conversation isn't "let me tell you what DataFlow does," it's "I see your team has been using DataFlow — how's it going? Here's how enterprise unlocks the limits you're hitting"
• Deal closes in 31 days instead of 94, because the product already proved its value

The PLG motion shortened the sales cycle by 63 days and eliminated the entire top-of-funnel marketing cost for this account. That's the compounding power of PLG at scale.

The most sophisticated companies in 2026 run both — a PLG motion for bottom-up adoption and a sales-led motion for top-down enterprise deals — with the two reinforcing each other. PLG fills the top of the funnel for free; sales converts the highest-value accounts at the bottom. The GTM tool stack serves both, with Salesforce as the connective tissue where product signals and sales activity meet.

Product examples

Hokeystack

The Specific Distribution Channels They Used

• Founder-Led LinkedIn Content (Primary Channel)
• Data-Driven Benchmark Reports (Content Distribution): They published reports analyzing over $640M in pipeline, $119M in revenue, and $255M in ad spend to reveal B2B SaaS benchmarks. These were genuinely useful data that no one else had. Every CMO and VP Marketing wanted to know how their numbers compared. That's a form of content-as-distribution
• Y Combinator (Platform Distribution)
• Direct Sales (Not PLG)

Why HockeyStack Couldn't Use PLG (and What That Tells You)
This is the important lesson. PLG only works when:

1. The product delivers value before the user has to connect all their data
2. A single user can experience the "aha moment" without buy-in from IT, RevOps, and Marketing
3. The product is horizontal enough that many personas at a company find it immediately useful
4. The price is low enough that an individual can expense it

HockeyStack fails all four tests. Their product requires you to connect your CRM, your ad platforms, your marketing automation, your product analytics — before it shows you anything meaningful. There's no "aha moment" in 5 minutes. It's an enterprise data infrastructure play dressed as an analytics tool.

So instead of PLG, they used content as their distribution moat — building an audience of the exact buyers (CMOs, VPs of Marketing, Revenue Operations leaders) who would eventually need their product. By the time those buyers were ready to evaluate attribution tools, HockeyStack was already the trusted name in their LinkedIn feed.

Distribution Strategy Must Match Product
HockeyStack is a good reminder that distribution strategy isn't chosen — it's determined by the nature of your product. The founders couldn't make their product viral, so they made themselves viral instead.

Posthog

Also occupying a niche (web analytics, instead of revenue attribution in hokeystack above), but trying to move into the general big data augmentation

Harvey

Harvey is almost the complete opposite of Manus in every dimension.

• Cold email, enterprise sales led motion
• Prestige led distribution + expand and collapse (build specialized AI agents for narrow legal workflows, then collapse them back into one interface that routes users intelligently)
• Platform distribution: lexisnexis + microsoft

Expand and collapse - The strategy works like this:

1. Expand — build deeply specialized agents for specific workflows (contract analysis, due diligence, litigation research, compliance)
2. Each agent proves ROI in a narrow use case, making it easy to justify to a risk-averse law firm partner
3. Collapse — bring all agents into one unified interface that becomes the lawyer's daily work environment
4. Now you're not a point solution — you're infrastructure

This mirrors what Salesforce did with the CRM, or what HubSpot did with marketing. Start narrow, prove value, expand into a platform. Switching costs compound with every workflow that gets built on top of you.

Vertical AI Has a Different GTM Playbook

Harvey is a canonical example of vertical AI — AI built for a specific industry rather than general use. And vertical AI almost always requires a different GTM approach than horizontal AI:

• The buyer is risk-averse and credential-obsessed (lawyers, doctors, financial advisors)
• The product needs to earn trust before it earns usage
• Distribution runs through professional networks, not social media
• Prestige is a moat — being the tool the best firms use is more defensible than being the cheapest or the fastest
• Expansion within accounts is more valuable than acquiring new accounts

Harvey's CEO has pointed out that GRR (Gross Revenue Retention) is the metric most AI investors are dangerously ignoring — because a vertical AI product that embeds itself into a firm's workflows and doubles seat count within 12 months is a fundamentally different business than one with high churn.

The short version: Manus won through virality and scarcity. Harvey won through trust and prestige. Both got to $100M+ ARR — just through completely opposite routes.

Company valuations

The Full Picture in One Table

The Macro Story: The 2021 Valuation Hangover

The most important context for reading these numbers is the 2021 vintage problem. During 2020-2021, SaaS companies were valued at 50-100x ARR multiples on the back of COVID-driven digital adoption and near-zero interest rates. When rates rose and growth slowed in 2022-2023, public market multiples compressed to 5-15x ARR — and private companies that raised at peak valuations are now trapped.

The pattern across this category:

• ZoomInfo: $20B peak → $1.7B today (91% decline) while revenue grew from ~$700M to $1.25B
• Gong: $7.25B peak → ~$4.5B secondary (38% decline) while revenue grew from $135M to $300M+
• Outreach: $4.4B in 2021, hasn't raised since, likely worth less today

As one analysis put it, Gong was valued at $7.25B in 2021 — that's a 38% drop while revenue tripled. Category leader, Gartner Magic Quadrant, 4,000+ customers, massive ARR — every metric screams "go public" except one: their valuation from 2021 was priced at a 72x revenue multiple that public markets won't support today. Hockeystack

The companies that avoided this trap are the ones that either went public before 2022 (HubSpot), raised at reasonable multiples (Clay's early rounds), or are genuinely AI-native and growing fast enough to justify premium multiples in the current environment (Clay at $3.1B on $100M ARR = 31x, still rich but defensible given growth rate).

The Winners and Losers

Clear winners: Clay (fastest-growing, right-place-right-time for AI GTM), HubSpot (scale, profitability path, AI investments paying off)

Under pressure: ZoomInfo (commoditization of contact data, AI tools replacing their core value prop), Gong (great business, trapped by 2021 valuation), 6sense and Outreach (similar 2021 vintage problem)

Interesting: Salesloft+Clari merger is either a smart consolidation play or two struggling companies combining — the market hasn't decided yet.