GTM Teams Are Unlocking Value with AI
The current state of AI for B2B SaaS GTM teams is no longer “experimental but interesting.” Its unevenly adopted, but already operationally meaningful to the reps and teams who take full advantage of it. Across industries, 88% of organizations report regular AI use in at least one business function, yet only about one-third say they have begun scaling AI across the organization. Marketing and sales remain the functions where revenue gains from AI are most commonly reported, and the companies seeing the most value are the ones redesigning workflows rather than merely layering AI onto old processes. Smaller companies are less likely than larger ones to have reached that scaling phase according to Mckinsey.
For PE firms that focus on middle-market and growth B2B SaaS companies, that means the opportunity is to help them move from scattered copilots and ad hoc prompting into a repeatable operating model built on cleaner data, fewer tools, role-specific workflows, and a clear build versus buy architecture. Salesforce’s latest sales research points in the same direction: 54% of sales teams already use AI agents, another 34% expect to within two years, and 94% of sales leaders with agents say they are critical for meeting business demands. At the same time, only about one-third of sales teams run on an all-in-one platform, the rest use an average of eight tools, and 84% of teams without an all-in-one platform say they plan to consolidate.
Our POV in one sentence: buy the execution surfaces and trusted systems of record, then build the intelligence and orchestration layer on top of them with Claude and MCP-compatible data/context connectors. That conclusion follows from both the market evidence on workflow redesign and the architecture trend now visible across HubSpot, ZoomInfo, Crossbeam, Outreach, Gong, and Anthropic’s own MCP tooling.
The framework below roughly captures the operating model we see emerging across high-performing GTM teams.
This is consistent with how Salesforce, Gartner, LinkedIn, Gong, Clari, and customer success platforms describe successful AI usage: AI handles research, signal monitoring, summarization, structured recommendations, and execution support; humans remain differentiated in empathy, judgment, confidence-building, and value framing.
What’s True Now in GTM AI
Three truths define the market right now.
- AI adoption is broad, but scaled transformation is still rare. The McKinsey survey referenced in the first section of this post found that 23% of respondents reporting that their organizations are scaling an agentic AI system somewhere in the enterprise, while 39% are still only experimenting. The same survey found that high performers are far more likely to redesign workflows, have direct senior-leadership ownership, and scale AI across multiple functions.
- The GTM value pool is real and concentrated in practical workflows. LinkedIn’s 2025 B2B sales research found that 56% of sales professionals now use AI daily. Among sellers using AI, 38% said AI-assisted lead and company research saves them more than 1.5 hours a week, sellers who improved response rates with AI reported an average 28% lift, and 69% of sellers using AI said AI shortened sales cycles by an average of one week. Gartner reinforces the same idea from the leadership angle: sales organizations that provide AI-enabled next best actions are 2.6 times more likely to achieve commercial growth, and organizations that prioritize upskilling sellers on AI are 2.4 times more likely to achieve strong revenue growth.
- The bottleneck is increasingly not model quality. It is context, data quality, tool sprawl, and operating discipline. Salesforce reports that reps spend more than half their time on non-selling work, including prospecting and data entry. It also found that 42% of reps are overwhelmed by too many tools, 51% of sales leaders with AI say tech silos delay or limit AI initiatives, and 74% of sales teams with AI are prioritizing data hygiene to support it. McKinsey found the same basic issue across the enterprise: many organizations are still in pilot mode, especially smaller ones, and workflow redesign is one of the strongest contributors to meaningful value.
That point is especially important for middle-market and growth companies. McKinsey found that nearly half of respondents from companies with more than $5 billion in revenue had reached the AI scaling phase, compared with 29% of respondents from companies with less than $100 million in revenue. Anthropic has also noted that adoption among smaller businesses has lagged larger enterprises and often “stops at the chat window,” which is precisely the trap many growth-stage software companies fall into.
There is also a buyer-side reality that keeps the human seller relevant. Gartner’s 2026 B2B buyer research found that 69% of buyers prefer to validate AI-generated insights with sales reps. In a related survey, buyers were 28 percentage points more likely to say a rep helped them advance to the next step in the purchase process than GenAI, 32 points more likely to say a rep made them feel confident in the purchase decision, and 39 points more likely to say a rep understood their needs. The implication is clear: AI is now excellent at research, messaging support, signal monitoring, and next best actions. It is not the system of trust that closes complex B2B deals.
A compact summary of the market evidence from LinkedIn, Salesforce, McKinsey, and Gartner is below.
| Market evidence | What it means for Plative’s POV |
|---|---|
| 88% of organizations use AI in at least one function, but only about one-third have scaled it. | The market is past “why AI,” but still early on “how to operationalize AI.” |
| Marketing and sales are the functions where revenue gains from AI are most commonly reported. | GTM is the best first beachhead for portfolio-wide AI value capture. |
| 54% of sales teams use AI agents now and 34% expect to within two years. | AI agents are already mainstream in sales planning assumptions. |
| 42% of reps feel overwhelmed by too many tools, and non-platform teams average eight tools. | Tool sprawl is a direct inhibitor of AI value. Consolidation is strategic, not cosmetic. |
| 69% of AI-using sellers report sales cycles shortened by about a week, and AI-personalized outreach users report a 28% response-rate lift. | The clearest near-term ROI comes from research, outreach, and execution efficiency. |
| Buyers still turn to humans for confidence and validation. | The goal is AI-augmented sellers, not seller replacement. |
GTM AI by User Level
We maintain that the most useful way to think about GTM AI in a B2B SaaS company is by role, because the maturity of off-the-shelf software varies meaningfully by user and workflow.
| User level | What off-the-shelf AI already does well | What should often be built or customized with Claude |
|---|---|---|
| Growth marketing and outbound ops | Prioritizes in-market accounts, interprets buying group and intent signals, personalizes campaigns, and coordinates programs across channels. 6sense positions itself as a signal foundation for every team, tool, and AI agent. Demandbase similarly emphasizes AI-driven prioritization of signals, accounts, and buying-group activity across marketing, sales, RevOps, and customer success. Mutiny focuses on account-personalized customer-facing content. HubSpot Breeze embeds AI across marketing, sales, and service. | Custom signal-to-play orchestration that blends proprietary product usage, field events, partner data, win-loss insights, and ICP nuance. This is especially useful when a company’s best growth signals are not available in a single commercial platform. |
| Sales development and prospecting | Researches leads, sources contacts, prioritizes accounts, drafts outreach, and automates follow-up. HubSpot’s Prospecting Agent can research enrolled contacts, execute outreach strategies, and use buying signals. Apollo, ZoomInfo Copilot, Regie, and Unify all pitch a combination of prospecting, personalization, and follow-up automation. Clay and Common Room increasingly serve as flexible signal and enrichment layers. | Account dossier generation that combines CRM history, web research, product telemetry, partner overlaps, and internal notes into a seller-specific or segment-specific outreach brief. Built workflows also help when tone, guardrails, and segmentation are highly company-specific. |
| Account executives | Delivers meeting prep, automates CRM hygiene, summarizes calls, flags risk, suggests next steps, and supports forecast inspection. Gong’s AI agents extract deal data, generate structured briefs, and identify themes from conversations. Salesloft highlights account research, drafting emails, managing follow-ups, highlighting at-risk deals, and surfacing key coaching moments. Outreach positions AI agents across prospecting, deal management, coaching, forecasting, and expansion. Clari focuses on deal inspection, trend analysis, forecasting, and revenue cadence. Salesforce positions sales AI around prospecting, conversation optimization, and decision support. | Customized deal strategy copilots, business-case generators, mutual action plan generators, competitive POV writers, solution-summary builders, and deal desk agents that reflect the company’s own sales methodology, packaging logic, and product positioning. |
| Account management and customer success | Surfaces health and risk signals, automates digital coverage, drafts follow-ups, and supports renewal and expansion workflows. Gainsight Atlas centers on retention, renewals, and scale without added headcount. ChurnZero AI emphasizes embedded next best actions, knowledge-grounded agents, and workflow execution across onboarding, adoption, renewal, and expansion. Vitally centers the AI-powered workspace for CSMs, with meetings, health, docs, dashboards, and automation. Planhat positions itself as an AI-enabled customer lifecycle platform. Crossbeam’s ecosystem data also supports retention and expansion. | QBR pack generation, stakeholder map refresh, adoption-to-expansion hypothesis generation, executive business review drafts, renewal war rooms, and account plans that combine product telemetry, support history, commercial history, conversation data, and partner context. These often need company-specific health-score logic and packaging rules. |
| Sales management, RevOps, and commercial leadership | Forecasting support, pipeline inspection, next best actions, coaching recommendations, and structured planning are now mature categories. Clari’s deal inspection and trend analysis agents are designed to flag slipping deals and recommend remediation. Gong predicts revenue outcomes from hundreds of conversation-derived signals. Gartner found AI-enabled next best actions strongly associated with commercial growth and predicts that by 2027, 95% of sellers’ research workflows will begin with AI. Salesforce reports that 91% of sales pros say AI benefits sales planning. | Portfolio benchmark agents, board-pack generators, whitespace estimators, quota and territory scenario tools, operating cadence copilots, and cross-portfolio pattern mining. These are ideal build candidates because the value sits in combining proprietary portfolio data and Plative’s operating model. |
A useful simplification is this: SDRs and marketers benefit most from AI that finds and works signals, AEs benefit most from AI that reduces prep and gives deal intelligence, CS teams benefit most from AI that scales coverage and turns customer telemetry into action, and leaders benefit most from AI that turns fragmented signals into predictable operating cadence. That pattern shows up repeatedly across the survey data and the vendor landscape.
Off-the-Shelf Tools That Work Now
The off-the-shelf market is increasingly coherent if you group it into layers rather than brands.
Signal and context layer: 6sense, Demandbase, Clay, Common Room, Unify
This is where modern GTM AI begins. 6sense describes itself as a GTM intelligence platform that turns every signal into intelligence every team, tool, and AI agent can act on. Demandbase uses AI to analyze signals, accounts, and buying groups and then activate coordinated programs across GTM. Clay, Common Room, and Unify occupy the more flexible, growth-oriented end of the market, where enrichment, buying signals, and workflow orchestration are bundled into operational systems for outbound and signal-based selling.
Execution layer: Apollo, Outreach, Gong, Salesforce
Now, those signals become seller behavior. Salesforce Sales Engagement is a strong example for growth and middle-market companies because it packages AI assistant features, autonomous agents, and embedded AI across the platform, while its Prospecting Agent can research accounts, execute outreach, and act on buying signals. Apollo offers a lighter-weight, consolidated AI sales platform for sales and marketing teams. Outreach, Unify, and Gong increasingly compete around who can turn signal plus CRM plus conversation data into prioritized execution.
Post-sale and retention layer: Gainsight, ChurnZero, Vitally, Planhat
TSIA’s customer success research suggests AI adoption in CS has lagged sales and marketing, but is accelerating as companies shift toward digital-led models, more predictive adoption metrics, and proactive churn prevention. Gainsight, ChurnZero, Vitally, and Planhat are all pushing versions of the same thesis: AI will let CSM organizations cover more accounts, spot risk earlier, and convert post-sale truth into pre-sales and renewal advantage.
Partner and ecosystem layer: Euler, PartnerStack, Crossbeam
The partner layer is no longer niche for B2B SaaS. Euler for example brings partner and ecosystem intelligence directly into Salesforce, HubSpot, Gong, and Outreach. Crossbeam cites partner-supported deals as associated with a 53% higher close rate and a 46% faster sales cycle, and its MCP server lets AI agents access ecosystem data on demand. Euler presents itself as an AI-native PRM platform, while PartnerStack AI automates partner recruiting. For software companies that rely on agencies, SIs, cloud partners, referral ecosystems, or co-sell influence, this category should now be considered part of the core AI GTM stack rather than an optional add-on.
With these AI-native or enabled applications, the time-to-value is faster, governance is clearer, and change management is easier than building those surfaces from scratch.
When To Build With Claude
Claude is best understood as three different implementation paths, each suited to different GTM jobs.
| Claude path | Best-fit GTM use cases | Why it fits |
|---|---|---|
| Claude Chat, Team, and Enterprise Search | Executive research, win-loss synthesis, deal reviews, customer summaries, QBR drafting, account planning, internal enablement content, and cross-functional knowledge search. | Claude can search the web with citations, connect to Google Workspace, and, on Team and Enterprise plans, run a dedicated organization-wide enterprise search project that searches across connected sources with optimized instructions. |
| Claude API | Embedded GTM copilots, structured output generators, human-in-the-loop workflow agents, proposal/RFP assistants, renewal-risk summarizers, and applications that need predictable JSON outputs or direct integration into existing software. | The Claude API supports web search, Files API, structured outputs, prompt caching, and direct remote MCP connections through the Messages API. Prompt caching can reduce repetitive processing cost and time, and files can be uploaded once and reused across workflows. |
| Claude Code, Agent SDK, and Managed Agents | Internal tools, multi-step GTM automation, ops bots, portfolio-level reporting automations, and workflow orchestration where code, files, commands, or tool use matter. | Agent SDK gives production agents the same loop and tools that power Claude Code, including reading files, running commands, web search, and editing code. Managed Agents adds secure sandboxing and built-in tools. Claude now supports up to 1M-token context windows on current Sonnet and Opus 4.6 models, which is useful for long GTM playbooks, large research packs, and complex operating materials. |
The emerging architecture trend in GTM software strongly favors building selective intelligence layers on top of bought systems. Anthropic’s MCP connector enables direct remote MCP server connections from the Messages API. HubSpot’s MCP server is now generally available and allows any MCP-compatible AI tool to read and write CRM records through natural language. ZoomInfo MCP connects AI models to structured company, contact, and technographic workflows. Crossbeam’s MCP server allows AI agents to access ecosystem context. Outreach has launched Outreach MCP, and Gong has introduced MCP support as well. Put differently, the GTM software market is becoming an AI-composable market.
That leads to a very actionable build thesis for Plative: buy the systems that create or hold GTM truth, build the workflows that combine those truths into differentiated decisions and outputs. This is an inference, but it is directly supported by the market’s shift toward MCP, organization knowledge search, and AI-native context layers.
The highest-confidence build opportunities for a B2B SaaS portfolio look like this:
| Build candidate with Claude | Why it is high value | Why it is usually better built than bought |
|---|---|---|
| Account research copilot that merges CRM history, ZoomInfo, Crossbeam, web research, product usage, and recent call notes into one brief | It directly improves SDR quality, AE prep time, and management visibility. LinkedIn and Gartner both show research efficiency and next best actions are among the highest-return AI workflows. | The best version depends heavily on proprietary ICP logic, product telemetry, partner context, and internal account history. Standard tools rarely blend all of those elegantly. |
| Deal strategy and business-case agent for AEs and SEs | It turns calls, emails, product usage, and internal playbooks into concrete next steps, mutual action plans, value narratives, and risk summaries. | Existing tools summarize and score, but the company-specific value framework, pricing logic, and competitive positioning are differentiated IP. |
| Renewal and expansion copilot for AM and CS | It supports QBRs, health reviews, adoption summaries, renewal risk explanations, expansion hypotheses, and executive stakeholder briefings. | Health logic, adoption thresholds, and renewal playbooks are highly company-specific even when the underlying data sits in Gainsight, ChurnZero, or Vitally. |
| Portfolio operating agent for CROs, CEOs, and PE operating partners | It converts pipeline, forecast, churn, expansion, hiring, and productivity data into weekly operating narratives and board-ready summaries. McKinsey and Gartner both point to workflow redesign and planning as key value levers. | There is no off-the-shelf product that naturally reflects a PE sponsor’s cross-portfolio benchmarks, value-creation plan, and Plative methodologies. |
| Partner motion assistant that identifies intro paths, co-sell plays, and retention/expansion influence | Crossbeam already surfaces ecosystem intelligence, and partner-supported deals can materially outperform. | The differentiator is combining ecosystem data with each portfolio company’s own stage, vertical, customer base, and target-account strategy. |
There are also clear governance implications. Prompt caching is eligible for Zero Data Retention, while Anthropic’s direct MCP connector is not. Team and Enterprise plans give organizations control over access, while Enterprise can expose compliance APIs for eDiscovery and DLP needs. The right architecture therefore depends on what is being passed to the model, where the authoritative data lives, and whether the portfolio company needs read-only analysis, human approval, or autonomous action.
Portfolio Rollout Blueprint
For a private equity firm, the best program is not “roll out one AI tool to every company.” The best program is a staged operating model.
One-pager for the PE sponsor
| What matters most | Plative recommendation |
|---|---|
| Where to start | Start in GTM, because marketing and sales are where enterprises most commonly report revenue gains from AI, and because the role-level use cases are already mature enough to be operationalized. |
| What to fix first | Data hygiene, tool consolidation, and workflow design. Salesforce and McKinsey both show these are prerequisites for scaled value. |
| Where to buy | CRM-native AI, sales engagement, conversation intelligence, forecasting, customer-success platforms, ecosystem intelligence. |
| Where to build | Cross-system decision support, portfolio benchmarking, account research, deal strategy, renewal intelligence, and partner orchestration. These rely on proprietary context more than generic UI. |
| What not to do | Do not fund isolated pilots with no data owner, no workflow owner, and no adoption plan. Gartner and McKinsey both warn that value comes from redesign and enablement, not tool spray. |
A practical rollout path framework we’ve used for other B2B SaaS companies looks a bit like this:
The strongest signal in the research is that AI value comes from operating-model change. High performers redesign workflows, simplify technology, own data quality, and upskill their teams. In sales specifically, Gartner found that organizations prioritizing AI upskilling are far more likely to achieve strong revenue growth.
A sensible first-year KPI framework for portfolio companies should focus less on generic “AI usage” and more on workflow outcomes: SDR research time, meetings booked per rep, reply rates, CRM completeness, meeting-prep time, pipeline inspection lag, forecast variance, renewal coverage ratio, QBR production time, expansion identification rate, and tool-count reduction. That emphasis matches the specific areas where the survey and vendor data show measurable value, namely research, outreach, planning, forecasting, retention, and signal-to-action speed.
Two Parting Questions
Two questions should be answered company by companies before they commit to an implementation plan.
First, where does the ground-truth GTM context actually live today, CRM, product telemetry, call data, marketing automation, partner systems, support systems, or spreadsheets?
Second, which workflows are standardized enough to buy immediately, and which are strategically differentiated enough to justify building on Claude with APIs and MCP? The answer will vary by portfolio company maturity, GTM motion, and data quality. The market evidence suggests that getting those two questions right matters more than choosing the “best model” in the abstract.