A curated directory of modern software for researching markets, creating demand, reaching buyers, operating revenue workflows, and measuring results.Directory release: 43 tools · 14 operating categories Last reviewed: 2026-08-30 Looking for a CRM, a research tool, or a better way to run a campaign? Start with the job below, then compare the products in that category. The list favours newer, AI-native products and modern infrastructure, while keeping a few established options for comparison. Each entry explains where the product fits and what to check before adopting it. If you are still deciding which work to automate, start with AI use-case selection. For newsletters and practitioners, see reading sources. No placement is paid. Lensmor is built and maintained by Ivan Xu, the owner of this repository; its relationship is disclosed in the entry below. All other tools are included independently. Third-party links include
ref=b2b-playbook only so destination sites can identify this repository as the referral source. The parameter is not an affiliate code, does not affect inclusion, and does not change the price you pay.
Favicons are stored locally and used only to identify products. Product names, marks, and icons remain the property of their respective owners.
How this directory uses “AI-native”
An AI-native claim should be supported by observable product behavior: access to relevant business context, multi-step work, tool use or workflow execution, human review or permissions, and a usable output. “AI-powered” copy alone is not enough.
Start with the job
Use this table to find a starting point. When comparing data providers, test them on the same sample of accounts and contacts. For broader platforms, check the specific features you will use rather than the size of the feature list.Tool directory
Every product appears once, grouped by its primary operating job. Use the type definitions above to compare products, then read across each row for fit, differentiation, and adoption risks. On a phone, swipe the tables sideways to read the remaining columns.01 · CRM, customer data & revenue systems
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Attio | AI-native | a GTM team wants a flexible CRM data model with AI research, classification, summaries, workflows, call context, and agent access close to the records. | objects, lists, workflows, AI attributes, embedded intelligence, and MCP make the CRM an active operating layer instead of a passive database. | complex permissions, reporting, migration, integration depth, or large-scale CRM administration matters more than flexibility. |
 Day AI | AI-native | a team wants calls, email, meetings, and relationship context to maintain the CRM and surface next actions with less manual data entry. | its product model starts with customer memory and conversation context, then uses AI to extract structured intelligence with sources and reasoning. | a younger platform must replace a deeply customized CRM; test integrations, field controls, permissions, export, and migration on real workflows. |
 HubSpot | Established benchmark | a team values one broad customer platform across CRM, marketing, sales, and service, plus a large integration and implementation ecosystem. | HubSpot is a useful benchmark for an integrated revenue stack, while Breeze adds contextual assistants, specialized agents, research, and workflow automation. | hub, seat, credit, implementation, and data-model costs may expand faster than the operating value. |
02 · Market, account & buyer intelligence
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Common Room | AI-native | reps and GTM teams need a signal workspace that combines first-party, product, community, web, and social activity—not only an automation layer. | it builds a continuously updated buyer-intelligence layer, resolves identities, and lets AI agents prioritize changes and move insights into GTM action. | source provenance, identity resolution, permissioning, regional coverage, or implementation load needs a controlled sample first. Common Room and Unify are not drop-in substitutes: compare who will live in the workspace versus who will run plays. |
 Unify | AI-native | a team wants intent signals, account research, enrichment, agent qualification, and outbound execution in one signal-driven workflow. | audiences and plays connect first-party, third-party, and AI-discovered signals to prospecting and multichannel action instead of stopping at an intent score. | credit consumption, signal provenance, mailbox ownership, suppression logic, or automated outreach permissions are not yet governed. Do not treat it as a simple replacement for Common Room’s workspace. |
 RB2B | Modern infrastructure | a B2B team wants to turn anonymous website visits into person- or company-level intent signals and route qualified profiles into CRM, Slack, or outbound workflows. | a lightweight tracking script, page-level context, ICP filtering, and native routing integrations turn web activity into timely sales signals without requiring a form submission. | person-level identification is US-only, match rates vary, and shared accounts or devices can cause misattribution. This category is noisy; benchmark match rate, ICP precision, and reachability on your traffic before buying. Validate geography, consent, cookie disclosures, suppression, and outreach rules. |
03 · Contact data, enrichment & verification
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Clay | AI-native | Marketing or RevOps needs GTM middleware: multi-provider enrichment, web research, classification, scoring, routing, and personalized data at scale. | Claygent combines agents, data providers, tables, workflows, and reasoning traces so teams can express research that a fixed database filter cannot. | the learning curve is steep and it is a poor self-serve tool for line reps. Nobody owns prompts, provider selection, credit spend, quality sampling, errors, and write-back. Extra HTTP sources should be hit-rate tested on your ICP, not standardized by brand. Pair with a sales UI when reps need lists, sequences, or a dialer. |
 pipe0 | AI-native | a team wants Clay-like search, enrichment, and automation as an API, MCP server, Slack bot, or sheet—without running the waterfall only inside Clay’s UI. | searches find people and pipes enrich them as the same primitives across API, spreadsheet, and agents; waterfalls can bill only completed answers. | coverage versus Clay’s provider catalog, credit cost at volume, and write-back ownership still need a sample. It does not replace a sales engagement UI. |
 Crustdata | AI-native | agents, internal tools, or Clay/n8n workflows need live people and company search, enrichment, and watchers through a versioned REST API or MCP—not a static contact file. | one people-and-company graph exposes search, enrich, signals, and an MCP server so agents can query fresh records instead of a stale export. | live and specialized endpoints are plan-gated. The product also serves recruiting and investing; test GTM fields, freshness, regional coverage, and lawful use on your own ICP before replacing another data layer. |
 FullEnrich | Modern infrastructure | a team already has people or accounts and needs a waterfall across multiple providers to find work emails or phone numbers. | one enrichment layer can query multiple data vendors through a UI, CRM workflow, API, or agent-facing infrastructure. | regional coverage, field-level accuracy, provider provenance, data rights, or unit economics are central; test a representative sample. |
 MoltSets | AI-native | an agent, script, Clay table, or n8n workflow needs people and company search, contact enrichment, email validation, mobile numbers, or reverse lookup through API or MCP. | MoltSets is designed without a traditional dashboard around agent-readable tools, consistent request shapes, risk scores, API access, and an MCP server for Claude, Codex, and other agent workflows. | a beta service, current US-first coverage, fair-use rules, phone-token economics, data provenance, privacy obligations, or production uptime needs deeper review. Benchmark a representative list before replacing other providers. |
 Apollo | Established benchmark | a sales team needs lists, sequences, a dialer, and a commercial UI—often next to Clay or another enrichment layer rather than instead of it. | it is a useful bundled baseline for database, engagement, and embedded AI against which narrower data and outreach products can be compared. | data accuracy, regional privacy obligations, source transparency, niche coverage, or avoiding an all-in-one dependency matters. It does not replace governed enrichment logic in Clay or an API. |
04 · Outbound & sales engagement
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Salesforge | AI-native | a team wants human-led or agent-assisted email and LinkedIn sequences alongside optional prospecting, mailbox, warm-up, and deliverability infrastructure. | the Forge stack spans data, infrastructure, outreach, replies, AI SDR execution, APIs, and MCP without requiring every component. | domain ownership, mailbox control, sender reputation, approval steps, or local outreach rules are not explicitly assigned. |
 Instantly | Modern infrastructure | cold-email infrastructure, sequencing, inbox management, lead data, and deliverability operations are already part of a qualified outbound motion. | outreach, sending accounts, warm-up, replies, CRM functions, and an AI copilot sit in one focused operating surface. | sending volume is being used to compensate for weak ICP evidence, generic copy, fragmented suppression lists, or uncertain compliance. |
 Nooks | AI-native | an outbound team needs parallel or power dialing, account research, live talk tracks, call automation, coaching, and a shared sales-floor workflow. | AI connects prospect prioritization, phone data, dialing, live context, notes, follow-up, and coaching around the moment a person answers. | telemarketing law, consent, number reputation, contact-data quality, rep readiness, or the economics of higher call volume are unresolved. |
05 · Content, brand & creative
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 AirOps | AI-native | a content team wants repeatable research, SEO/AEO, creation, refresh, quality review, and publishing workflows grounded in brand knowledge. | workflows and agents combine models, search, data, knowledge bases, human review, and CMS integrations instead of treating content as one prompt. | no subject-matter expert owns factual accuracy, source quality, editorial differentiation, and the risk of scaled low-value content. |
 Gamma | AI-native | a team needs a fast first version of a presentation, visual document, one-pager, social asset, or lightweight website. | one AI design surface can generate, reshape, brand, share, export, and programmatically create several communication formats. | factual review, brand precision, accessibility, PowerPoint export fidelity, collaboration controls, or workspace data settings are critical. |
 Publer | Modern infrastructure | a small B2B team needs one calendar to schedule LinkedIn and other social posts without buying an enterprise social suite. | network coverage, bulk scheduling, and small-team pricing cover the operating job that Buffer-class tools usually own. | LinkedIn personal-versus-company API limits, approvals, and analytics depth still need a trial. Listening, advocacy, and influencer modules belong in a larger suite; scheduling will not invent a point of view. |
06 · Website, SEO & conversion
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Framer | AI-native | a marketing or design team wants to generate and refine pages, CMS structures, copy, visuals, code components, and site improvements in one canvas. | its agents work directly on editable website and CMS state, while humans can inspect, change, and publish the result. | portability, complex CMS modeling, localization, accessibility governance, custom back-end behavior, or enterprise publishing controls dominate. |
 Relume | AI-native | a team wants to turn a company brief into a sitemap, wireframes, copy direction, and a style guide before detailed design or development. | AI is constrained by a real component system and can export structured work into Figma, Webflow, or React workflows. | the team expects a finished strategy or production website; information architecture, proof, copy, accessibility, and final implementation still need owners. |
 Mintlify | AI-native | product, developer-relations, support, or marketing teams need documentation and knowledge sites that help both human users and AI agents understand and adopt a product. | Git- and MDX-based publishing, API documentation, preview deployments, cited AI answers, agent-assisted maintenance, analytics, APIs, MCP, llms.txt, and Markdown export connect content creation, discovery, and agent consumption. | content ownership, source freshness, AI credit and plan limits, authentication, search and answer quality, accessibility, migration, export, and support escalation require explicit owners. |
07 · Paid demand, ABM, personalization & buyer enablement
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Mutiny | AI-native | marketing and sales need account-specific pages, decks, business cases, or other customer-facing assets connected to brand and GTM context. | one agent can use brand material and account data to produce personalized experiences and assets across campaigns and deals. | identity matching, data permissions, brand approval, proof selection, or incremental conversion lift cannot be measured. |
 Metadata | Modern infrastructure | a B2B demand team wants one execution layer to plan, build, review, experiment with, and optimize paid campaigns across supported channels. | it focuses automation on campaign operations and pipeline-oriented demand work instead of supplying another reporting-only dashboard. | channel coverage, account matching, minimum spend, creative control, attribution assumptions, or platform access must remain fully in-house. |
 Qwilr | Modern infrastructure | a sales, marketing, customer-success, or agency team needs interactive proposals, quotes, personalized buyer pages, electronic acceptance, payment, and engagement analytics. | Qwilr turns a static proposal into a branded web-based buying experience, while CRM integrations, templates, AI creation, conditional content, and a smart proposal workflow reduce repetitive production. | complex CPQ, legal approval, procurement redlines, CRM source-of-truth rules, accessibility, offline delivery, or buyer-tracking consent requires a more controlled document process. |
08 · Events, webinars & field marketing
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Lensmor | AI-native · Owner product | a B2B sales, field, ABM, partnership, or event team needs to choose relevant trade shows and identify ICP-fit exhibitors, attendee signals, available visitor data, and decision-makers before outreach. | event discovery, source-aware company and people data, ICP matching, contact enrichment, AI-assisted research, and export into the existing CRM or outreach stack are connected in one vertical workflow. Disclosure: Lensmor is built and maintained by Ivan Xu, the owner of this repository. It is included under the same evidence and limitation standard and is not claimed to be universally best. | coverage is assumed to be complete or every person is assumed to be an organizer-confirmed registrant. Depth varies by event, source, region, and timing; review source labels, privacy, and outreach permissions. |
 Bizzabo | Established benchmark | a field or event team runs in-person, hybrid, or conference programs and needs registration, onsite, networking, and CRM-connected operations in one event OS. | portfolio-scale event operations—not a single webinar tool—sit on an open integration surface. | enterprise contracting, attendee-data ownership, and whether webinar or video repurposing actually needs a separate platform (see Goldcast) remain open. |
 Goldcast | AI-native | a B2B marketing team wants webinars, virtual or field events, recording, attendee experience, and post-event video repurposing in one platform. | events, Content Lab, recording, search, highlight reels, and agentic video workflows connect a live program to an ongoing content engine. | event format, registration ownership, integrations, media rights, retention, attendee data, or enterprise contracting require a narrower platform. It is not a substitute for a full in-person conference OS. |
 iCapture | Established benchmark | booth teams need one badge- or card-capture app across shows that writes qualified leads into CRM or MAP fields instead of a rented scanner spreadsheet. | badge-provider integrations, offline capture, and CRM sync turn floor interactions into reportable records. | show coverage, field mapping, enrichment add-ons, and price versus event volume still need a quote. Scanning is worthless if booth fields never land in the CRM. |
09 · Partnerships, affiliate & ecosystem
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Crossbeam | Modern infrastructure | companies need to identify overlapping customers, prospects, and opportunities with partners while controlling what underlying data is shared. | ecosystem data, account mapping, intelligence, AI chat, and MCP can bring partner context into seller and agent workflows. | CRM sync scope, matching rules, privacy, sharing policies, partner adoption, and ownership of co-sell actions are not agreed. |
 PartnerStack | Modern infrastructure | a B2B SaaS company needs to recruit, onboard, manage, attribute, and pay affiliates, influencers, resellers, or co-sell partners. | program operations and automated payouts are connected to a B2B partner network rather than treated as an internal spreadsheet. | the program lacks partner economics, recruitment ownership, enablement, fraud controls, tax review, or a clear distinction between management and demand creation. |
10 · Lifecycle, customer marketing & advocacy
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Loops | Modern infrastructure | a SaaS team wants marketing, product, lifecycle, and transactional email connected to product events and contact properties without a large marketing suite. | a focused data model, API, workflows, experiments, reusable themes, and agent-assisted editing keep email close to the product. | complex enterprise lead scoring, many brands, deep sales-led nurture, broad channel orchestration, or advanced governance requires a larger MAP. |
 Featurebase | AI-native | product, support, customer-success, and GTM teams want one customer-facing system for support, feedback, roadmaps, updates, help content, tickets, surveys, and in-product communication. | customer and company context connects the inbox, feedback, help center, changelog, workflows, Fibi AI, public API, and MCP so support signals can become product and lifecycle action. | customer-data sync, AI answer sources, escalation, permissions, support SLAs, multi-brand governance, localization, or migration from a specialized help desk needs a controlled pilot. |
 UserEvidence | AI-native | product marketing, customer marketing, sales, or advocacy teams need to capture, approve, organize, and activate verified customer proof. | AI helps ask better follow-ups, surface quotes and outcomes, turn approved evidence into assets, and recommend relevant proof without inventing customer claims. | customer permission, approval workflow, selection bias, source traceability, or the difference between a testimonial and representative evidence is not governed. |
11 · Workflow automation, GTM engineering & AI agents
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 n8n | Modern infrastructure | a technical operator needs deterministic automation and AI agents with APIs, code, human approval, error paths, model choice, and a self-hosting option. | traditional workflows, agent nodes, MCP, memory, guardrails, monitoring, and hundreds of integrations can coexist without forcing every step through an LLM. | nobody owns credentials, retries, observability, deployment, upgrades, model cost, and recovery from partial failure. |
 Gumloop | AI-native | a team wants no-code AI workflows and agents for web research, data extraction, documents, enrichment, triggers, and connected business tools. | visual workflows can become tools for adaptive agents, while APIs, webhooks, MCP, and reusable nodes connect the result to other systems. | external-site changes, credit consumption, prompt drift, permissions, monitoring, or auditability could make an autonomous workflow brittle. |
 Zapier | Established benchmark | a non-technical team needs the fastest route to connect common applications, build straightforward trigger-action workflows, or test an agent using a broad integration catalog. | its ecosystem is the benchmark for accessible business automation, and Zapier Agents extends that action layer into agent workflows. | task-based cost, branching complexity, error handling, data residency, agent permissions, or silent failures could become material. |
12 · Product, marketing & revenue analytics
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 HockeyStack | AI-native | a B2B team needs buyer journeys, multi-touch attribution, lift analysis, account intelligence, and plain-language analysis across marketing and CRM data. | Odin works as a scoped AI analyst over the connected GTM model, with structured, inspectable outputs and recommended next actions. | tracking coverage, identity stitching, CRM hygiene, funnel definitions, implementation scope, or attribution is being mistaken for causal proof. |
 Dreamdata | Modern infrastructure | a B2B company needs account-level customer journeys, revenue attribution, campaign performance, and audience activation across a long buying cycle. | it joins marketing, web, advertising, and revenue data to make channel and content contribution traceable back to the underlying journey. | account matching, data completeness, lookback windows, model choice, offline touches, or attribution results are being treated as experimental causality. |
 PostHog | AI-native | product, growth, and marketing teams want one event model for web and product behavior, funnels, retention, paths, session replay, feature flags, experiments, surveys, and revenue signals. | autocapture, SQL and HogQL, replay, flags, experiments, AI-generated analysis, self-driving regression reports, APIs, and MCP keep quantitative evidence close to product and growth decisions. | event design, identity resolution, consent, privacy controls, data residency, ingestion cost, sampling, retention, and instrumentation ownership need governance; observed correlations and AI-generated explanations are not causal proof. |
13 · Privacy, consent & data governance
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Transcend | Modern infrastructure | legal, privacy, engineering, data, and marketing teams need one layer for consent, preferences, data inventory, data-subject requests, lineage, and AI data-use decisions. | policy-as-code and system connectors can turn “may we use this data?” into an enforceable, logged decision across customer-data and agent workflows. | no qualified owner has mapped legal requirements, purposes, systems, jurisdictions, exceptions, and implementation responsibility. The product does not replace legal advice. |
 Vanta | Modern infrastructure | a B2B company needs continuous compliance evidence, risk workflows, a customer-facing trust center, or faster security-questionnaire handling during sales. | integrations, evidence collection, trust sharing, questionnaire automation, and AI-assisted review connect security operations to customer trust. | the required framework, product tier, audit scope, evidence quality, human verification, or buyer-specific security process is unclear. Automation does not guarantee certification. |
14 · Conversation intelligence & coaching
| Tool | Type | Use when | Why it stands out | Validate before adopting |
|---|---|---|---|---|
 Fathom | AI-native | a team needs recording, transcripts, summaries, and searchable notes—and is not buying forecast, deal inspection, or a revenue OS. | bot or no-bot capture, instant summaries, and ask-across-calls keep the job at “what was said” without forcing an enterprise RI suite. | coaching scorecards, CRM write-back, recording consent, and whether managers will actually inspect calls still need owners. It is not a substitute for Gong-class forecast and enablement. |
 Gong | Established benchmark | sales and CS leaders need conversation intelligence plus forecast, coaching, and enablement on the same revenue graph—not only a notetaker. | capture, deal inspection, forecast, and enablement sit on one interaction graph so managers can coach from what buyers said. | implementation load, seat cost, and the temptation to treat CI as a second CRM or a second forecast. If the only job is notes, start with a lighter tool. Split this job from cadence in MarTech governance. |
Copyright © 2026 Ivan Xu. All rights reserved. See the copyright and reuse terms. Canonical source: github.com/weilun88313/B2B-Playbook