AI Sales and Marketing Tools Market: October 2026
The AI layer of go-to-market sells on volume and prices on credits, and the proof sits almost entirely with vendors. No research firm sizes AI SDR agents as a category.
What did we find?
- Outbound agents hit a ceiling that inbox providers set. Gmail, Yahoo, and Microsoft cap complaint rates and reject unauthenticated bulk mail. An AI SDR priced per contact reached earns more as it sends more, and the inbox rules punish exactly that volume.
- Retention evidence for AI SDR tools is thin and contested. The one third-party investigation found, from 11x employees, "$14 million" claimed ARR against about "$3 million" past trial. 11x answered with a 79% retention rate. Nothing since has settled it.
- Seats are giving way to credits, with the seat kept as a gate. Clay and ZoomInfo's GTM.AI price without seat limits or seat fees. HubSpot lets buyers swap seats for credits mid-contract. Outreach, Gong, and Salesforce keep the seat and meter agent work on top.
- Outcome pricing exists, and it is narrower than the label. HubSpot charges $0.50 per resolved conversation and $1 per lead recommended for outreach. None of seven AI SDR pricing pages read this month charges per meeting booked or per qualified lead.
- Suites are buying the agent layer. Salesforce bought Qualified (closed 2026-04-01) and agreed to buy Momentum and Braze. HubSpot agreed to buy Warmly. Adobe closed Semrush. Clari merged into Salesloft and Highspot into Seismic. Clay raised at $7.1B. Money flows to the data layer and to the suites.
- Write access through MCP is now normal, and its price is the open question. Outreach, Apollo, Clay, Common Room, Klaviyo, and HubSpot let outside agents write records. Only ZoomInfo publishes a dollar price per credit for agent use, and Lusha a credit rate per MCP call.
The state of the AI layer of go-to-market in October 2026, from no vendor’s side. The scope is everything outside core CRM. That means AI SDR and BDR agents, data enrichment and orchestration, sales engagement and revenue intelligence, and AI content and marketing agents. CRM vendors appear only where they ship SDR or marketing agents, because CRM and sales force automation sit outside this report.
The question this edition tracks is claims against proof. Vendor claims are marked (vendor claim) and are unverified. Third-party evidence is marked (third-party evidence). Arithmetic on published figures is labeled as such.
What happened in October 2026. This is the first edition, so it sets the baseline.
- Salesforce agreed to buy Braze for about $5.66B. The 2026-10-20 deal pays $47.50 a share in cash and closes in Salesforce’s fiscal 2028. It is the largest deal in this edition’s timeline (section 12).
- Gartner re-rated revenue orchestration. The 2026 Revenue Action Orchestration Magic Quadrant (2026-10-13) evaluated 13 vendors and named four Leaders, with Gong and Outreach confirmed. Writer claimed the top capabilities score in IDC’s agentic marketing MarketScape for B2C enterprises, where SAP, Adobe, and Insider One claim Leader (section 2).
- HubSpot moved buyers from seats toward credits, and cut 7% of staff. On 2026-10-15 it added default spend caps and self-serve seat-to-credit swaps. On 2026-10-06 it announced about 660 job cuts, “not driven by AI-related efficiencies” (sections 8 and 12).
- Agents got budgets before they got proof. ZoomInfo added a monthly credit budget per Agent Team (2026-10-20) and Salesforce added hard spending limits to Agent Fabric (2026-10-27). No vendor in the Agent Pricing Tracker publishes a refund for a wrong agent action (sections 8 and 9).
- The month’s proof was vendor-written. Outreach’s Forrester study claims 244% ROI and was commissioned by Outreach. No third party published AI SDR retention data in the window (section 11).
- Agent write access spread, and its price stayed unpublished. Amplemarket added MCP tools (2026-10-15) and Writer put MCP on existing seats (2026-10-27). Salesforce’s plan to meter agent MCP calls in Flex Credits still has no public rate (section 9).
- Regulation moved by calendar, not by new rules. No October FCC order on AI calls was found. The EU AI Act’s chatbot disclosure duty has applied since 2 August, and machine-readable marking for older systems falls due 2026-12-02 (section 10).
1. Summary
The AI layer of go-to-market sells on volume and prices on credits, and the proof sits almost entirely with vendors. No research firm sizes AI SDR agents as a category. The adjacent markets that are sized run from about $3.7B (sales intelligence) to $33B (AI in marketing), on definitions that do not agree. The public companies in scope report slowing or negative net retention in data and steady retention in lifecycle marketing. Every AI SDR outcome number found this month comes from the vendor that sells the agent.
Seven things define the market in October 2026.
- Outbound agents hit a ceiling that inbox providers set. Gmail, Yahoo, and Microsoft cap complaint rates and reject unauthenticated bulk mail. An AI SDR priced per contact reached earns more as it sends more, and the inbox rules punish exactly that volume (section 11).
- Retention evidence for AI SDR tools is thin and contested. The one third-party investigation found, from 11x employees, “$14 million” claimed ARR against about “$3 million” past trial. 11x answered with a 79% retention rate. Nothing since has settled it (section 11).
- Seats are giving way to credits, with the seat kept as a gate. Clay and ZoomInfo’s GTM.AI price without seat limits or seat fees. HubSpot lets buyers swap seats for credits mid-contract. Outreach, Gong, and Salesforce keep the seat and meter agent work on top (section 8).
- Outcome pricing exists, and it is narrower than the label. HubSpot charges $0.50 per resolved conversation and $1 per lead recommended for outreach. None of seven AI SDR pricing pages read this month charges per meeting booked or per qualified lead (section 8).
- Suites are buying the agent layer. Salesforce bought Qualified (closed 2026-04-01) and agreed to buy Momentum and Braze. HubSpot agreed to buy Warmly. Adobe closed Semrush. Clari merged into Salesloft and Highspot into Seismic. Clay raised at $7.1B. Money flows to the data layer and to the suites (section 12).
- Write access through MCP is now normal, and its price is the open question. Outreach, Apollo, Clay, Common Room, Klaviyo, and HubSpot let outside agents write records. Only ZoomInfo publishes a dollar price per credit for agent use, and Lusha a credit rate per MCP call (section 9).
- Public evidence has limits. Review sites, analyst pages, and several vendor pricing pages blocked automated reads. This edition measures what vendors publish, not what they win or keep (section 14).
2. Categories and terms
Eight labels describe the space. Each maps to a different budget line and shortlist.
| Label | Who uses it | Scope |
|---|---|---|
| AI SDR or AI BDR | Vendors (11x “digital workers”, Artisan “AI BDR”, AiSDR) | Agents that find, research, and contact prospects and book meetings. No analyst category or sizing found |
| Sales intelligence | Mordor Intelligence, Precedence Research | B2B contact and company data with intent signals (section 5) |
| GTM orchestration or GTM engineering | Clay, Unify | Workflow tools that chain data, AI research, and outreach. Clay’s CEO spoke on the “GTM engineer” role at TechCrunch Disrupt 2026 (TechCrunch, 2026-10-02, article URL not kept) (third-party evidence) |
| Revenue action orchestration | Gartner | Sales engagement, conversation intelligence, and forecasting in one category. Replaced the sales engagement framing |
| Revenue orchestration platforms | Forrester, IDC (“unified revenue orchestration”) | Same territory as Gartner’s category |
| Multichannel marketing hubs | Gartner | B2C lifecycle messaging (Braze, Salesforce Marketing Cloud, Adobe) |
| Agentic marketing | IDC (2026, three editions: B2C enterprises, midmarket, small businesses) | Marketing agents. The first analyst category with “agentic” in its name in this market |
| AI in marketing | Precedence Research and other sizing firms | Any AI spend in marketing. The widest and least comparable label |
The label decides the shortlist. A RevOps leader buying Clay searches “GTM engineering”. A CRO buying Outreach reads Gartner’s revenue action orchestration MQ. An SDR manager buying an AI SDR finds no analyst coverage at all and relies on vendor content and peer referral.
Analyst placements, read 2026-10-30. Analyst pages returned 403 to fetches, so placements come from vendor announcements and third-party summaries, labeled.
| Report | Date | Placements |
|---|---|---|
| Gartner Magic Quadrant for Revenue Action Orchestration 2026 (Rietberg, Gottlieb, Cruz) (third-party evidence, via vendors) | 2026-10-13 | 13 vendors evaluated, four Leaders. Gong (“one of only four vendors named a Leader”, second year, Gong, 2026-10-16) and Outreach (second year, Outreach, 2026-10-15). The other two Leaders are not named in any source read, including two third-party summaries (Rework, Industry Lens) |
| Gartner Critical Capabilities for RAO 2026 (third-party evidence, via vendor) | 2026-10-14 | Outreach claims first for the “Acquire New Customers” use case |
| Gartner RAO MQ, first edition (third-party evidence) | December 2025 (2025-12-15 per a Gartner abstract in a search index) | Clari Leader and Salesloft Visionary, per Clari and Rework (updated 2026-05-30). Gong and Outreach announced Leader placements |
| Forrester Wave, Revenue Orchestration Platforms for B2B, Q3 2026 (third-party evidence, via vendor) | 2026-09-24 | Outreach, top Current Offering score per Outreach. Others not named |
| IDC MarketScape, Unified Revenue Orchestration Platforms 2026 (third-party evidence, via vendor) | 2026-09-21 | Outreach a Leader. Others not checked |
| Gartner Magic Quadrant for Multichannel Marketing Hubs (third-party evidence, via vendor) | 2026-09-22 | Braze a Leader for the fourth year, per Braze’s release of 2026-09-24 |
| IDC MarketScape, Worldwide Agentic Marketing for B2C Enterprises 2026 (US54113426) (third-party evidence, via vendors) | September 2026 | SAP (SAP, 2026-09-30), Insider One (Insider One, 2026-09-29), and Adobe (report excerpt) claim Leader. Writer links the same report and claims “the highest score in capabilities of any vendor”, not Leader (Writer, 2026-10-12) |
| IDC MarketScape, Agentic Marketing for Midmarket (US54841126) and Small Businesses (US54113526) (third-party evidence, via vendor) | September 2026 | ActiveCampaign a Leader in both (ActiveCampaign, 2026-10-06) |
| Forrester Wave, Experience Optimization Solutions, Q3 2026 (third-party evidence, via vendor) | 2026-08-17 release | Optimizely a Leader |
Salesloft’s homepage makes no 2026 analyst claim (read 2026-10-30), a notable absence for a former first-edition Leader through Clari. No Gartner B2B Marketing Automation MQ, Forrester Conversation Intelligence Wave, October Forrester Wave in sales or marketing, or AI SDR placement was found. That gap is a search-coverage limit, not proof of absence (section 14).
3. The process and where value sits
Buyers evaluate the AI layer against nine steps, composited from vendor product pages read this month. The steps are target account selection, data and enrichment, signal detection, outbound outreach (email, LinkedIn, phone), inbound conversion (web chat and voice), meeting booking, deal execution and forecasting, content creation, and lifecycle messaging.
The evidence ranks the steps unevenly.
- Inbound conversion holds the clearest measured outcomes, and they are still vendor-measured. Salesforce reports its own use of Qualified’s Piper at 6% conversion and “more than 60 meetings per week on average” since April 2026 (Salesforce, 2026-07-29) (vendor claim). The inbound visitor already chose to arrive, so inbox rules do not bind.
- Outbound is where the claims are largest and the constraint is hardest. The ceiling is deliverability, not copy quality (section 11).
- The data layer captures spend. Clay claims 4x revenue growth in 2025 and “More than 17k customers” (Clay, 2026-09-09) (vendor claim). ZoomInfo’s revenue grew 1.2% in Q2 2026 (section 5). The money moved from the static database to the workflow on top of it.
- Deal execution has the system of record for conversations. Gong claims ARR above $500M growing more than 55% (Gong, 2026-05-12) (vendor claim). Globes notes that ARR is not an accounting figure (Globes) (third-party evidence).
- Content creation has the weakest independent proof. Jasper cites a Forrester study with “342% ROI” that Jasper commissioned (2025-09-10) (vendor claim, vendor-commissioned). No independent ROI study for marketing content agents was found.
- Lifecycle messaging has audited retention. Braze reports dollar-based net retention of 110% (section 5), the only audited retention figure among the marketing agent vendors.
4. Buyers and triggers
| Persona | Buys on | Evidence |
|---|---|---|
| CRO or VP Sales | Pipeline per rep, meetings booked, forecast accuracy | Gartner’s RAO MQ anchors on this buyer (section 2) |
| RevOps or GTM engineer | Data coverage, workflow control, credit cost per run | Clay sells unlimited seats and meters actions, a model built for a small technical team (section 8) |
| SDR manager | Meetings per month at lower cost than a hired SDR | AI SDR pricing pages quote prospects or leads per month (section 8) |
| CMO or marketing ops | Content volume, campaign speed, lifecycle revenue | HubSpot’s Campaign, Nurture, and Content agents launched in public beta at UNBOUND 2026 (martech.org, 2026-10-16) (third-party evidence) |
| CFO or procurement | Predictable spend on credits | HubSpot’s CCO, quoted by CMSWire: “Nobody should have to miss out on getting the most out of AI because they don’t know what the bill will be.” (CMSWire, 2026-10-15) (third-party evidence) |
Buying triggers, strongest evidence first:
- Sales headcount is falling at the vendors themselves. ZoomInfo cut about 600 roles, roughly 20% of staff, in May 2026, and its CEO’s email cited a shift toward consumption pricing (8-K accepted 2026-05-11, read through StockTitan) (vendor claim, filed). HubSpot’s October cut says the opposite about cause (section 12). A buyer cannot read AI substitution from either.
- Spend control is now a buying criterion. Three vendors shipped caps in one month (section 8). Spend caps sell to the CFO, which says the CFO now sits in the deal.
- Enterprise adoption claims are vendor-reported. 11x cites a Ramp spend report in which “40% of enterprises buying in this category chose 11x” (11x, 2026-05-14) (vendor claim). The post links no Ramp source, and the report was not found.
How buyers score AI is changing (speculative, analysis). Expect buyers to stop scoring “has an agent” and score cost per qualified meeting, inbox placement after 90 days, and whether the agent stops inside a budget. The first evidence is this month’s spend caps and HubSpot’s move to per-outcome units.
5. Size and structure
No firm sizes the AI SDR category. The sized adjacent markets disagree by more than their growth rates.
| Firm (read 2026-10-30) | Market | Base | Forecast | CAGR |
|---|---|---|---|---|
| Mordor Intelligence (updated 2026-01-22) | Sales intelligence | $4.42B (2025), $4.99B (2026) | $9.15B (2031) | 12.89% |
| Precedence Research (updated 2025-06-30) | Sales intelligence | $3.65B (2025) | $9.02B (2034) | 10.54% |
| Grand View Research (updated July 2026) | Marketing automation software | $6.7B (2024) | $15.6B (2030) | 15.3% |
| Precedence Research (updated 2026-09-09) | Marketing automation software | $9.80B (2025) | $36.97B (2035) | 14.20% |
| Precedence Research (updated 2026-09-16) | AI in marketing | $32.73B (2026) | $217.33B (2034) | 26.7% |
The two sales-intelligence figures differ by about 21% on the same base year. The two marketing-automation figures differ by about $3B across adjacent years, more than either growth rate explains, so the firms count different things. MarketsandMarkets returned 403. The AI-in-marketing figure counts any AI spend, so it does not size the agent vendors.
Public company anchors are the harder numbers. Each comes from an SEC filing (vendor claim, filed).
| Company | Period | Revenue | Growth | Retention or scale |
|---|---|---|---|---|
| HubSpot (8-K) | Q2 2026 | $911.7M | 20% (17% constant currency) | 306,446 customers, $11,800 average subscription revenue per customer. Q3 guide $924 to 925M |
| ZoomInfo (8-K) | Q2 2026 | $310.4M | 1.2% | Net revenue retention 89%. 1,891 customers above $100K. FY2026 guide $1.207 to 1.217B |
| Klaviyo (8-K) | Q2 2026 | $370.6M | 26% | More than 205,000 customers. 4,477 above $50K ARR, up 36% |
| Braze (8-K) | Q2 FY27 (to 2026-07-31) | $227.2M | 26.2% | Dollar-based net retention 110%. 2,789 customers |
| Adobe (8-K) | Q3 FY2026 | $6.76B | 13% | Total ARR $27.50B. “AI-first ARR grew more than 150%”, no dollar base |
Private vendors disclose ARR on their own terms. Gong claims above $500M (vendor claim). Clay announced $100M ARR on 2025-12-08 (vendor claim), and Sacra estimates about $200M in September 2026 (third-party evidence, estimate). Outreach claims AI ARR grew 480% year over year in its fiscal Q2 and AI credit use grew 12x in H1 2026, with no base (Outreach, 2026-09-09) (vendor claim). A growth rate with no base cannot be sized.
Structure. Three layers hold the market. CRM suites (Salesforce, HubSpot, Microsoft) bundle agents into seats and credits. Data incumbents (ZoomInfo, Apollo) hold contact records and face AI-native workflow tools on top. Point agents (11x, Artisan, AiSDR, Regie.ai) sell a replacement for an SDR hire. The suites are absorbing the point layer by acquisition (section 12).
6. Segments
| Segment | Who sells there | What decides deals |
|---|---|---|
| Enterprise CRM-native | Salesforce (Agentforce, Piper, Hunter), Microsoft Dynamics 365 Sales, HubSpot | Installed CRM, credit bundles, governance controls |
| Mid-market outbound AI SDR | 11x, Artisan, AiSDR, Regie.ai, Amplemarket | Meetings booked per month, deliverability, contract terms |
| GTM engineering and orchestration | Clay, Unify, Common Room | Data coverage, workflow flexibility, credit cost per run |
| B2B contact data | ZoomInfo, Apollo, plus Cognism, Lusha, People Data Labs (not researched this edition) | Coverage, accuracy, compliance with data broker law |
| Revenue orchestration | Outreach, Salesloft (with Clari), Gong, Backstory (formerly People.ai) | Analyst standing, CRM write-back, forecast accuracy |
| B2C lifecycle messaging | Braze, Klaviyo, Salesforce Marketing Cloud, Adobe | Send volume, decisioning, channel breadth |
| Content and brand agents | Jasper, Writer, Adobe GenStudio, HubSpot Content Agent | Brand control, approval flow, enterprise security |
| Self-hosted open source | Mautic | Cost and data control. No AI features in the 7.2 release notes (GitHub, read 2026-10-30) (third-party evidence) |
Many deals run through the CRM already owned. HubSpot’s prospecting agent ships inside Sales Hub, and Salesforce’s Hunter and Piper ship inside Agentforce. A point AI SDR competes against a line item the buyer already pays for. HubSpot now sells its agents as Agent Hub, “included in all Starter, Professional, and Enterprise editions”, and custom agents built in Agent Builder “consume HubSpot Credits” (HubSpot, read 2026-10-30) (vendor claim). Third parties report Agent Hub and Agent Builder as renames of Breeze Agents and Breeze Studio, launched as a Professional and Enterprise beta (Resolve247, AI Weekly) (third-party evidence). HubSpot’s own pages date the launch 2026-07-23, state no rename, and keep Breeze only in Breeze Assistant (vendor claim).
7. Drivers and restraints
Drivers, ranked by evidence quality.
- Consumption pricing lowers the cost of a trial (strongest, from pricing pages). AiSDR charges $1.50 per person prospected with no monthly fee, and ZoomInfo’s GTM.AI gives 1,000 data and 1,000 AI credits free (section 8) (vendor claim).
- Agent access through MCP (real, recent). Outside agents in Claude, ChatGPT, and Copilot now read and write GTM records in at least nine products (section 9).
- Cost pressure on sales teams. Vendors in scope cut their own staff this year (section 12). The cause each states differs, so the driver is real and its link to AI is unproven.
- Analyst categories with agent language. IDC’s three agentic marketing MarketScapes are the first of their kind in this market (section 2).
Restraints.
- Deliverability. The binding constraint on outbound volume (section 11).
- Platform restriction on scraping. LinkedIn removed the Apollo and Seamless.AI company pages on 2025-03-06. Apollo’s CEO said the company was “actively working with LinkedIn to understand the nature of our brand page restriction” (MarTech, 2025-05-08) (third-party evidence). LinkedIn gave no public statement in that report. Secondary sources say the Apollo page was later restored, unverified.
- Churn and contract lock-in. The 11x terms make credits non-refundable and payments “non-cancelable” (11x terms, June 2026) (vendor claim). Retention evidence is contested (section 11).
- Data broker law. California’s DROP platform has required brokers to process deletion requests since 2026-08-01 (section 10). ZoomInfo, Apollo, Clay, and Amplemarket are registered brokers.
- Runaway spend. Credit meters let a verbose agent cost more than planned. Vendors answered with caps rather than with lower prices (section 8).
- Trust in autonomous sending. AiSDR’s site says the product is not for buyers who want “set and forget” (AiSDR, read 2026-10-30) (vendor claim). HubSpot’s deal agent uses “one click approval”. The autonomous default is a choice vendors still hedge.
8. Pricing and deal models
Four meters compete, and the seat survives as the gate. Vendors price on what they can count.
| Model | Vendors | What the buyer carries |
|---|---|---|
| Flat monthly plan sized by volume | 11x (prospects, calls, chats per month), Artisan (leads per month), Qualified (web sessions and active email contacts) | Pays for capacity whether or not meetings result |
| Per contact or per lead | AiSDR ($1.50 per person prospected), HubSpot Prospecting Agent ($1 per lead recommended) | Pays for reach, not for replies |
| Seat plus credits | Outreach, Gong (credits for Enrich), Regie.ai, Unify, Apollo, Amplemarket, Common Room, Salesforce, Microsoft | Pays per user and again per action |
| Credits without seats | Clay (unlimited seats), ZoomInfo GTM.AI (“no seat fees”) | Carries volume and design risk. A verbose workflow costs more |
| Per outcome, by vendor definition | HubSpot Customer Agent ($0.50 per resolved conversation), Salesforce ($2 per conversation, $2 per Help Agent resolution) | The vendor decides what counts as the outcome |
Published agent prices, read 2026-10-30. All (vendor claim) unless marked.
| Vendor | Product | Price | Unit |
|---|---|---|---|
| 11x | Alice Growth (page) | From $3,750 a month, billed annually | Up to 5 users, 2,000 new prospects a month. About $1.88 per new prospect (arithmetic) |
| 11x | Julian Growth Voice and Chat (page) | From $5,333 (voice) and $2,417 (chat) a month, annual | 2,000 calls or 3,000 chats a month |
| Artisan | Ava (page, HTML read) | Not published | Plans “sized to your lead volume, mailboxes, and dialer seats”, about 2,500 or 6,000 leads a month |
| AiSDR | Pay as you go (page) | $1.50 | Per person prospected, no monthly fee. Team plans quoted at up to 40% less per contact |
| Regie.ai | RegieGO Pro and Team (page) | $41 to 49 a month, and $125 to 150 per seat a month | 5,000 and 15,000 credits a month per seat |
| Qualified | Piper (page) | Not published | Web sessions plus active email contacts |
| Salesforce | Agentforce (page) | $500 per 100,000 Flex Credits. $0.10 per action, $0.15 per voice action. $2 per conversation | Action or conversation |
| Salesforce | Agentforce add-on, User License, Coworker License | $125, $5, and $50 per user a month | User |
| Salesforce | Sales Core, Advanced, Max (page) | $195, $395, and $550 per user a month. Max includes 2.75M Flex Credits per org per year | User |
| HubSpot | Credits (catalog) | $0.010 per credit, $10 per 1,000 | Credit |
| HubSpot | Customer, Prospecting, Content, Nurture agents (same catalog) | 50, 100, 1,000, and 10 credits | Resolved conversation, lead, content piece, email |
| Microsoft | Dynamics 365 Sales Premium (page) | $150 per user a month with 1,000 Copilot Credits | User. No per-credit rate published |
| Clay | Launch and Growth (page) | $167 and $446 a month on plan cards | 15K actions plus 3K data credits, and 40K actions plus 6K data credits. Unlimited seats |
| ZoomInfo | GTM.AI pay as you go (page) | $0.38 per data credit at the $1,000 tier, $0.05 per AI credit | Credit, no seat fees |
| Common Room | Essential (page) | $2,500 a month, billed annually | 5 seats, 5K research and 2.5K Prospector credits |
| Unify | Base and Pro (page) | $20 and $60 per seat a month | 800 and 2,400 credits per seat a month |
| Amplemarket | Startup (page) | $600 a month, annual | 2 users plus per-user credits |
| Jasper | Pro (page) | $69 monthly or $59 annual | Seat. Agents run on credits at an unread rate card |
| Writer | Palmyra X6 (blog) | $2 input and $8 output | Per million tokens |
| Klaviyo | Free tier (page) | “$5 of Composer usage/month” | Dollars of AI usage |
| Outreach, Gong, Salesloft, Braze | Platforms | Not published | Outreach and Gong price per user plus credits. Salesloft is “Talk to Sales” |
Pages contradict themselves, which is a finding. Clay’s pricing page shows Launch at $167 on the card, “Starts at $60/mo” elsewhere, and $185 in the FAQ, with two data credit counts. Clay posted its move from Starter, Explorer, and Pro to action-plus-credit plans on 2026-03-11, and PricingSaaS recorded it the next day (PricingSaaS) (third-party evidence). HubSpot’s newsroom lists the outcome-priced agents on Pro and Enterprise, while the product page lists Starter too. ZoomInfo’s credit expiry reads “expire 12 months after they are granted” and “never expire while your account is active” on one page.
The shift from seats to credits. HubSpot made it explicit on 2026-10-15. “They can swap seats for credits (or for another seat type) themselves without waiting for their contract to renew” (HubSpot) (vendor claim). The same page sets an account credit limit “up to 50% above a customer’s monthly credits” by default. A pricing blog reports that new EMEA customers now buy one platform edition per seat plus credits, with a start date that conflicts with HubSpot’s own, unverified. Salesloft removed its Advanced and Elite packages on 2026-09-28 (UsagePricing) (third-party evidence).
Outcome pricing is narrower than the label. HubSpot moved two agents on 2026-04-14. “You pay when it works, full stop” (HubSpot, updated 2026-04-13) (vendor claim).
The prospecting unit is a lead recommended for outreach, not a reply or a meeting. The previous unit was a monthly charge per enrolled contact. No vendor among 11x, Artisan, AiSDR, Regie.ai, Qualified, Salesforce, and HubSpot charges per meeting booked or per qualified lead on its pricing page (all read 2026-10-30). Vendors outside that set were not searched.
Spend controls arrived in one month. HubSpot’s default cap (2026-10-15), ZoomInfo’s “Set a Monthly Credit Budget on Each Agent Team” and a default cap of 100 audience-trigger runs a day (2026-10-20, release notes), and Salesforce’s FinOps Dashboard and Model Wallet, which give “hard limits that stop requests before costs exceed budget” (2026-10-27, Salesforce) (vendor claim). Cross-market pricing comparisons live in the Agent Pricing Tracker 2026.
Contract terms carry the risk the meter hides. 11x’s terms say “All Credits purchases are non-refundable” and fees are “non-cancelable and non-pro-ratable for partial months” (§4.1 and §4.3, terms, June 2026) (vendor claim). AiSDR runs team plans on quarterly contracts and pay as you go with no contract. A buyer who doubts retention claims can choose the contract shape that limits exposure.
9. The agentic shift
Vendor messaging ran from copilot, to named agent, to “digital worker”, to always-on agents. Third-party verification of autonomy is almost absent. Claim inflation shows in how vendors describe the same function. Outreach’s CPO says “Autonomous agents handle the execution”, while its Deal Agent “sends AI-suggested deal-field updates to Slack so reps can accept or edit” (Business Wire via 01net, 2026-04-27) (vendor claim). Gong’s CEO says “Now, AI wins deals for you”, while its triggered Custom Agents are “rolling out through the rest of the year” (Gong, 2026-09-30) (vendor claim).
The ladder mirrors the AR edition’s five rungs, adapted so that rung 5 means a published unit price with spend control.
| Rung | Definition | Who is there (October 2026) |
|---|---|---|
| 5. Priced agent economics | Agents sold at a published unit price, with spend caps or outcome units | HubSpot (per-outcome agents, default cap, seat-to-credit swaps). Salesforce (per action and per conversation, Agent Fabric hard limits). ZoomInfo (GTM.AI credit prices, budget per Agent Team) |
| 4. Commercially proven agents | Named agents with usage or customers disclosed by a listed company, an acquirer, or a third party | Qualified Piper (Salesforce’s own use, acquired 2026-04-01). Klaviyo (Composer above 138,000 users, Customer Agent above 1.3M messages, Klaviyo, 2026-09-09). Braze (Decisioning Studio, now being acquired). Clay (Claygent passed 1 billion runs, 2025-06-05) |
| 3. Shipped agents that act on their own | Agents that send, update, or book without per-item review, outcomes vendor-claimed | 11x (Alice “Autopilot activated”, Julian voice). Artisan (Ava “Autonomous by default, adjustable by design”). AiSDR (sends after launch review). Regie.ai (“Always-On Pipeline”). Amplemarket (Duo). Apollo (scheduled agents). Gong (AI Data Extractor fills CRM fields “with no rep input required”). HubSpot Prospecting Agent (fully autonomous mode) and Nurture Agent |
| 2. Drafting and research with approval | Research, summaries, drafted copy, a human sends or approves | Outreach (Deal Agent suggestions, Omni), Salesloft (about 15 research and drafting features), Writer (playbook approvals), Common Room (RoomieAI research), Unify, Microsoft Sales Development agent (preview), Optimizely Virtual Teammates (“built-in oversight”) |
| 1. AI marketing ahead of product | Agent language with no shipped scope, or scope not verifiable | Salesforce Hunter (in pilot, GA planned November per a partner blog, Incepta, 2026-10-02) (third-party evidence). Clay, which at Sculpt (2026-10-08) showed “a new way to use Clay, led by an agent that plans, runs, and improves your plays” for attendees, with early access “in the coming weeks”, and previewed an Inbound SDR Agent by waitlist (Clay) (vendor claim). Jasper GEO Agent (“autonomous execution”, no publishing scope stated). Adobe (GenStudio and Agent Orchestrator pages timed out) |
Mautic sits off the ladder. Its 7.2 releases carry no AI features, and its “AI Manifesto” page returned 404.
MCP write access is common, and almost nobody prices it.
| Vendor | Agent access (read 2026-10-30) | Write | Price for agent use |
|---|---|---|---|
| Outreach | MCP server GA 2026-02-24, full suite 2026-06-02 | Yes. Outside agents “sequence prospects, update records, and run workflow automations” | Not stated |
| Apollo (docs) | About 60 actions | Yes. “Destructive actions, such as bulk delete, aren’t supported” | Enrichment and search draw credits. Most CRM and email actions do not |
| Clay (MCP) | Claude, ChatGPT, Codex | Yes. “push to sequences”, “all governed by Ops-defined write rules” | Not stated |
| Common Room | MCP and CLI | Yes. “write-backs on contacts, organizations, activities, segments, and notes” | Included “across all plans” |
| Klaviyo | ”260+ MCP tools and capabilities and 490+ APIs” | Yes. Agents can “read, write, and go live on its own” | Not stated |
| HubSpot (MCP) | Remote server, OAuth | Lists read and write objects, while the FAQ says “read-only access to standard records” | Not stated |
| ZoomInfo (docs) | Account research, contact search, conversation intelligence | Not stated | Credits. “Some older subscriptions only carry recurring monthly credits, which MCP can’t use” |
| Braze | Operator Connect, beta, acts under the user’s identity | Yes, in beta | Not stated |
| Salesloft | MCP gives agents “full context into your live pipeline” | Not stated | Not stated |
| Gong | MCP live through Microsoft 365 Copilot (2026-07-01) | Not stated | Not stated |
| Writer | MCP on existing seats (2026-10-27) | Not stated | Included in the seat |
| Amplemarket | ”Prospect, research, and take action in Claude or ChatGPT”, new tools 2026-10-15 | Yes, per the page | Tier not shown |
| Lusha (credits, 2026-10-15) | API, MCP, Claude, ChatGPT | Not stated. “Export to CRM” costs 1 credit per record | 1 credit per search call returning 1 to 25 results. Reveals 1 credit for an email, 5 for a phone. No dollar rate on the page |
| Cognism (blog, 2026-10-15) | Claude, ChatGPT, Copilot as a custom connector | No. “controlled and read-only” | Standard and Pro plans. Existing credits apply |
| Salesforce | Agentforce is an MCP client. “Claudeforce” lets users “update records, and trigger workflows” from Claude (2026-09-15) | Yes | Not stated. No Salesforce page read this month gives a rate or start date for agent MCP calls |
| 11x, Artisan, AiSDR, Regie.ai, Qualified | No public MCP server found on homepages, pricing pages, or sitemaps | n/a | 11x lists “Custom MCP Integration” as a Julian Enterprise line item |
Jasper shipped an MCP server on 2025-09-23 with scope unstated, and Backstory says it works with Claude, ChatGPT, Copilot, and Slack. The pattern holds across categories. The incumbents with a system of record opened write access, and the point AI SDRs did not.
10. Regulation
Outbound AI touches four rule sets: telephone, email, data protection, and AI transparency. None changed in October. The calendar did the work.
US federal
| Item | State (read 2026-10-30) | Why it matters for AI SDR and marketing tools |
|---|---|---|
| TCPA, AI voices | The FCC’s Declaratory Ruling of 2024-02-08 (FCC-24-17) holds that “AI technologies that resemble human voices and/or generate call content using a prerecorded voice” fall under the TCPA’s “artificial or prerecorded voice” rules (Kelley Drye) (third-party evidence). fcc.gov returned 403 | AI voice outbound (11x Julian, Salesforce Agentforce Voice) needs prior express consent, and written consent for telemarketing |
| FCC AI-call disclosure NPRM | FCC-24-84, adopted August 2024, reply comments closed 2024-10-25 (Wiley) (third-party evidence). It proposes AI disclosure when a caller obtains consent (Wiley) and at the start of a call (Perkins Coie) (third-party evidence). Law-firm trackers and 2026 open-meeting agendas show no final order and no October 2026 order or NPRM (searched 2026-10-30) | A final rule adds a script duty to every AI voice agent |
| One-to-one consent rule | Vacated by the Eleventh Circuit (No. 24-10277, opinion file dated 2025-01-24, court PDF) (third-party evidence, primary, text not parsed) | Lead-generation consent shared across sellers survives at the federal level |
| CAN-SPAM | Opt-out, sender identity, and no deceptive headers. The last confirmed FTC inflation adjustment (effective 2025-01-17) is cited at $53,088 per email by secondary sources. A Federal Register document reports OMB memo M-26-11 cancelling 2026 penalty inflation adjustments (FR 2026-18853) (third-party evidence, not parsed) | Each AI-sent email is a separate potential violation |
US states
| Item | State (read 2026-10-30) | Why it matters |
|---|---|---|
| Texas SB 140 | Signed 2025-06-20, effective 2025-09-01, extends private rights of action for “solicitation-related communications” (Texas Legislature) (third-party evidence, primary) | Text outreach to Texans now carries private suit risk |
| California Delete Act (DROP) | “Data brokers are required to begin processing these requests on August 1, 2026”, at least once every 45 days (CPPA) (third-party evidence, primary) | Contact databases shrink as California residents delete in one request |
| California data broker registry | Lists ZoomInfo Technologies, Zenleads (Apollo), Clay Labs, Tagis (Amplemarket), People Data Labs, Cognism, Lusha, Seamless, LeadIQ, Warmly, and 6sense. Common Room and Unify not found (CPPA registry CSV, downloaded 2026-10-30) (third-party evidence, primary) | Registration puts these vendors inside DROP |
| Utah SB 226 | Effective 2025-05-07. A supplier using generative AI in a consumer transaction must disclose it “if the individual asks or otherwise prompts the supplier”. Disclosure “at the outset” and “throughout the interaction” earns a safe harbor. Fines up to $2,500 per violation. The AI Policy Act now repeals 2027-07-01 (enrolled bill) (third-party evidence, primary) | A chat or voice agent selling to Utah consumers must answer truthfully when asked. Whether B2B outreach counts as a “consumer transaction” was not checked |
| Colorado AI Act (SB24-205) | SB25B-004, signed 2025-08-28, moved the requirements to “June 30, 2026” (Colorado Legislature) (third-party evidence, primary). Any 2026 change is unverified | High-risk AI duties. GTM scoring tools are unlikely to qualify, analysis |
European Union
| Item | State (read 2026-10-30) | Why it matters |
|---|---|---|
| EU AI Act, Article 50 | The chatbot disclosure duty (Art. 50(1)) and deepfake and public-interest text labels apply from 2026-08-02. The Digital Omnibus moved Art. 50(2) machine-readable marking to 2026-12-02 for systems on the market before 2 August (McCann FitzGerald, 2026-07-02, Silicon Canals, 2026-07-18) (third-party evidence). Fines up to €15M or 3% of turnover. The Official Journal text did not render | A website chat agent (Piper, HubSpot Customer Agent, Braze Conversational Agents) must tell the visitor it is AI at first contact. Writer’s watermarking ships “by Dec 2”, matching the marking date |
| High-risk AI rules | Pushed to 2027-12-02 for Annex III systems by the Omnibus (Silicon Canals) (third-party evidence) | Outside most GTM tools |
| GDPR and ePrivacy | B2B cold email runs on legitimate interest in most member states, with prior consent rules in some. National rules and enforcement were not re-read this edition | Apollo turned off French do-not-call screening and moved consent responsibility to customers (Releasebot, 2026-09-01) (third-party evidence, aggregator) |
The CNIL fine on Kaspr for scraping LinkedIn data (€240,000, December 2024) is cited from memory. Both CNIL URLs tried returned 404, so it is unverified this edition.
11. Deliverability and the economics of outbound volume
The AI SDR prices on contacts reached, and the inbox providers punish contacts reached. This section carries the bulk-sender rules, how vendors work around them, and the retention evidence.
The rules
| Provider | Rule | Threshold | Effective | Source |
|---|---|---|---|---|
| Gmail | SPF and DKIM, DMARC at p=none minimum, one-click unsubscribe on marketing mail | ”close to 5,000 messages or more to personal Gmail accounts” in 24 hours. Keep spam below 0.1% and never reach 0.3% | February 2024. “Starting November 2025, Gmail is ramping up its enforcement on non-compliant traffic”, with “temporary and permanent rejections” | Google (third-party evidence, primary) |
| Yahoo | SPF and DKIM, DMARC at p=none minimum that “must pass”, From-domain alignment, unsubscribes honored “within 2 days" | "Keep your spam rate below 0.3%”. The page states no bulk volume number. Some guides cite 5,000 a day | February 2024, unsubscribe from June 2024, per secondary guides | Yahoo Sender Hub, read 2026-10-30 (third-party evidence, primary) |
| Microsoft Outlook.com | SPF and DKIM pass, DMARC pass with alignment | 5,000 or more to consumer mailboxes from one From domain (support article) or “more than 5,000” a day (Postmaster page) | 2025-05-05. Junk first, then rejection with “550 5.7.515” | Microsoft Support (third-party evidence, primary) |
At Gmail’s 0.3% ceiling, 5,000 messages a day allow 15 complaints (arithmetic). An agent that sends to 2,000 new prospects a month (11x Alice Growth) from one domain stays below the bulk tier. The same agent at enterprise volume (10,000+ prospects a month) does not, unless it spreads mail across domains.
How vendors work around it
Every outbound vendor sells the workaround as a line item.
| Vendor | Mailbox or domain product | Source |
|---|---|---|
| 11x | Managed Gmail mailboxes in Alice Growth | Pricing page (vendor claim) |
| Artisan | Sends “in your reps’ names, from your own sending domains”. Plans sized by mailboxes | Pricing page (vendor claim) |
| AiSDR | Warmed mailbox $33 | Pricing page (vendor claim) |
| Unify | Managed mailboxes $25 per mailbox a month | Pricing page (vendor claim) |
| Apollo | MCP can buy sending domains. Catch-all domain protection shipped 2026-09-01 | Docs (vendor claim), Releasebot (third-party evidence) |
| Clay | Sequencer GA 2026-09-01 on Launch, Growth, and Enterprise. Domains and inboxes bought with credits, about three weeks of warming advised | Clay (vendor claim) |
Analysis. Mailbox and domain rotation keeps each domain under the complaint ceiling while total volume grows. The rules count complaints per domain, so the workaround scales the vendor’s revenue and the buyer’s domain count together. The provider rules aim at the sending identity, and the products multiply identities. Expect the inbox providers to aggregate across related domains, which the Gmail rule already does for “the same primary domain”. That is speculative and is a checkpoint (section 13).
Retention evidence
The only investigative evidence on AI SDR retention is 19 months old and unrebutted by data.
| Source | Finding | Tag |
|---|---|---|
| TechCrunch, 2025-03-24 | Employees said 11x “might say it had $14 million in annual recurring revenue” while contracts past the three-month break clause “totaled only about $3 million”. One said “We were losing 70-80% of customers that came through the door.” ZoomInfo said “we are not a customer” | (third-party evidence) |
| 11x, in the same article | ”retention rate is currently 79%”, highest churn in “initial cohorts in late 2023”, reports contracted ARR | (vendor claim) |
| Sifted, via Refonte (2026-08-12) | Alleged retention of about 20 to 30% for part of 2024. 11x called it “inaccurate”. Sifted not read directly | (third-party evidence, secondary) |
| Rework, updated 2026-06-03 | ”churn rates run at roughly 50 to 70 percent annually, according to 2026 sales-tech analyses”. Early churners “cited inbox placement and domain health”. No primary source named | (third-party evidence, weak) |
The two 11x figures do not reconcile. A 79% retention rate and a 70 to 80% loss rate describe different cohorts, different definitions, or a contradiction. Neither side published the cohort. 11x’s founder stepped down as CEO in May 2025 (TechCrunch) (third-party evidence). On 2026-10-14 its new CEO repositioned the company toward voice agents, with no ARR, funding, or customer figures (11x) (vendor claim).
Audited retention exists only at the public companies, and it splits by layer. ZoomInfo’s net revenue retention is 89%, so its installed base shrinks before new sales. Braze’s is 110%. Neither sells an AI SDR. The data layer under outbound agents loses revenue per customer while the lifecycle layer gains it. That is consistent with buyers moving data spend into workflow tools (section 3), and the method cannot prove the cause.
Vendor proof in the month was commissioned or self-reported. Outreach’s Forrester Consulting study, commissioned by Outreach, models 244% ROI and payback “in less than six months” for a composite organization (2026-10-07) (vendor claim, vendor-commissioned). Gong cites “roughly 32% larger deal sizes and 26% higher win rates” for early adopters (vendor claim). Customer outcomes on AI SDR sites are all (vendor claim): Leica Biosystems “9.7% reply rate” (11x), Medisafe “29 meetings in 30 days” (AiSDR), Emburse “3X more meetings in 6 months” (Qualified), Jotform 26 qualified deals from 1,000 contacts (HubSpot).
12. Consolidation and funding
Deal timeline. Salesforce/Qualified (agreed December 2025, closed 2026-04-01, price undisclosed) · Clari and Salesloft (merger announced August 2025, closed 2025-12-03) · Salesforce/Momentum (agreed 2026-02-18, close expected in Salesforce’s fiscal Q1 2027, price undisclosed, Momentum) (vendor claim) · HubSpot/Warmly (agreed 2026-06-30, price undisclosed, Warmly) (vendor claim) · Seismic/Highspot (completion announced 2026-08-18, terms undisclosed, Highspot) (vendor claim) · Adobe/Semrush ($1.9B, announced 2025-11-19, closed about 2026-04-28) · Klaviyo/Agency team and technology (2026-08-05) · ZoomInfo/DoubleO.ai (date unverified) · ZoomInfo/Workbounce (October 2026, date unverified) · Salesforce/Braze (agreed 2026-10-20, about $5.66B).
What the timeline shows:
- The suites buy the agent and the channel. Salesforce now owns an inbound SDR agent (Piper) and has agreed to own conversation capture (Momentum) and a B2C lifecycle engine (Braze). HubSpot’s Warmly deal puts a registered data broker (section 10) inside a CRM suite. “$47.50 in cash per share”, “approximately $5.66 billion in enterprise value”, closing in Salesforce’s fiscal 2028 after shareholder and regulatory approval (Salesforce, Braze 8-K) (vendor claim, filed). Braze’s stock fell about 25.8% after its September results, per a secondary report (third-party evidence).
- Revenue orchestration and enablement consolidated. Seismic and Highspot now run as Seismic under CEO Rob Tarkoff and showed a combined roadmap at Seismic Shift, 12 to 15 October (vendor claim). Clari and Salesloft trade as Salesloft, clari.com/press redirects to salesloft.com, and Steve Cox is CEO (Salesloft) (vendor claim). Andy Byrne, named as future CEO at announcement, is not on the leadership page (read 2026-10-30). Reported layoffs of 76 roles and about $450M combined ARR come from a competitor with no source, unverified.
- Adobe bought search visibility, not an agent. Semrush deregistered on 2026-05-08 (Form 15-12G, EDGAR) (third-party evidence, primary). Adobe’s CEO changes on 2026-12-01, when Anil Chakravarthy succeeds Shantanu Narayen (vendor claim).
- Vendors are cutting their own staff for stated reasons that conflict. ZoomInfo cut about 600 roles (about 20%) in May 2026 and cited consumption pricing (section 4). HubSpot announced on 2026-10-06 that it will “reduce the size of our team by ~7%” and that “This is not driven by AI-related efficiencies” (8-K) (vendor claim, filed). HubSpot expects “approximately $65 million to $75 million” in charges, mostly in Q4 2026 (8-K) (vendor claim, filed), and reorganizes product teams “around these outcomes instead of hubs”.
Venture money concentrates in the data and workflow layer.
| Company | Round | Date | Source |
|---|---|---|---|
| Clay | $115M Series D at $7.1B, led by Wellington | 2026-09-09 | Clay (vendor claim) |
| Clay | Employee tender at $5B | 2026-01-28 | Clay (vendor claim) |
| Clay | $100M Series C at $3.1B, led by CapitalG | August 2025 | Sacra (third-party evidence) |
| Unify | $40M Series B led by Battery Ventures, with OpenAI Startup Fund | Date not shown | Unify (vendor claim) |
| Artisan | $25M Series A led by Glade Brook, with HubSpot Ventures | Page date not extracted | Artisan (vendor claim) |
| 11x | $50M Series B led by a16z, about $350M valuation | October 2024 | Secondary only (third-party evidence) |
| Writer | $200M Series C at $1.9B | November 2024 | (third-party evidence). No 2026 round found |
Clay’s valuation more than doubled in thirteen months, from $3.1B to $7.1B. This edition verified no 2026 round for any AI SDR point vendor in scope. Funding for 11x, AiSDR, Regie.ai, Apollo, and Amplemarket in 2025 and 2026 is unverified, not absent (section 14).
13. Dated checkpoints
| Date | Event | Why it matters |
|---|---|---|
| 2026-11-04 | ZoomInfo Q3 2026 results, after market close | Whether net revenue retention moves off 89% as consumption pricing and Agent Teams land |
| Early November (Nov 4, 5, or 11 by calendar, unconfirmed) | HubSpot Q3 2026 results | First quarter with seat-to-credit swaps and spend caps. Any credit revenue disclosure |
| 2026-11-05 (third-party calendar, unconfirmed) | Klaviyo Q3 2026 results | Composer and Customer Agent usage against 138,000 users and 1.3M messages |
| November 2026 (partner blog) | Salesforce Hunter GA inside Agentforce Coworker | Moves Hunter off rung 1 if it ships |
| Fall 2026 (Clay’s wording) | Clay Workflows leave beta. Agent-first Clay reaches early access | Moves Clay off rung 1 if the agent ships beyond Sculpt attendees |
| By 2026-12-31 | Gong triggered Custom Agents finish rollout | Gong’s first agents that run on events without a user prompt |
| 2026-12-01 | Adobe CEO transition | Strategy signals for GenStudio and Semrush |
| 2026-12-02 | EU AI Act Art. 50(2) marking for pre-existing systems. Writer watermarking due | First marking deadline for content agents in the EU |
| Early December (date not announced) | Braze Q3 FY27 results and proxy for the Salesforce vote | Retention against 110% and the shareholder approval timeline |
| 2026-12-31 | End of Clay’s legacy-plan trial window | Whether legacy Starter, Explorer, and Pro customers churn or migrate |
| Unscheduled | FCC final order on AI-call disclosure | A script duty on every AI voice agent |
| About October 2027 | Next Gartner RAO MQ | Whether any AI-native SDR vendor enters the category |
14. Evidence limits
- Web search was limited. Most reads in this edition are direct fetches of known pages, which favors vendor pages over third-party coverage. Absence of an event means “not found in the pages read”, not “did not happen”.
- Outside-in only. Public research measures what vendors publish, not what they win, keep, or lose.
- Analyst and review pages blocked fetches. Gartner, G2, Forrester, and MarketsandMarkets returned 403 or empty pages. Placements in section 2 come from vendor announcements and third-party summaries, labeled.
- Primary regulator pages blocked fetches. fcc.gov returned 403, two Federal Register PDFs did not parse, the EUR-Lex omnibus page did not render, and CNIL URLs returned 404. Each affected claim in section 10 is labeled with the source used instead.
- Several pricing pages hide prices from automated reads. Artisan, Apollo, Outreach, Gong, Salesloft, Qualified, and Braze publish no readable price. Apollo’s prices come from a third party.
- Every AI SDR outcome number is vendor-claimed. No third-party study of AI SDR meetings, reply rates, or retention was found in the window.
Unverified items.
- The two unnamed Leaders in Gartner’s 2026 RAO MQ (two searches, Gartner abstract 403), and the first edition’s exact day.
- Salesforce Hunter’s November 2026 GA, and the Qualified deal price ($1.2B and about $500M both appear in third-party snippets).
- Salesforce’s Flex Credit rate and start date for agent MCP calls. No Salesforce page read this month states the plan.
- HubSpot’s MCP price, GA status, and whether MCP calls draw credits. Its EMEA per-seat-plus-credits switch and start date.
- HubSpot’s Q3 date (IR page 403, not in its 8-Ks) and Klaviyo’s Q3 date (IR page 404).
- Why Clay’s plan prices conflict on one page.
- ZoomInfo’s DoubleO and Workbounce deal dates, whether GTM.AI MCP can write, and its Snowflake, Exa, and n8n announcements. Its October release notes name none of them.
- Common Room’s ownership. Its site navigation carries a Zoom product group and a “Zoomtopia ‘26 recap”, and neither Common Room’s site nor Zoom’s October newsroom names a deal.
- Salesloft’s February 2026 layoffs and the combined ARR, and Andy Byrne’s current role.
- Funding in 2025 and 2026 for 11x, AiSDR, Regie.ai, Apollo, and Amplemarket. 11x’s Series B figures. Unify’s and Artisan’s round dates.
- The Ramp spend report 11x cites. The UserGems attribution of the 50 to 70% AI SDR churn figure. Sifted’s 11x retention report, read only through a secondary source.
- Braze’s acquisition of OfferFit and the Braze deal premium.
- Whether Microsoft has moved from Junk routing to blanket rejection.
- Whether Salesforce/Momentum and HubSpot/Warmly have closed. Neither vendor page read says so.
- The 2026 CAN-SPAM penalty figure after the reported OMB cancellation of 2026 inflation adjustments.
- Any 2026 change to the Colorado AI Act date after 2026-06-30 (the bill page lists none). Any state AI-disclosure law with a 2026 effective date. California’s SB 942 dates were not read (leginfo returned 403). The CNIL Kaspr fine. Whether LinkedIn restored Apollo’s company page.
- Adobe GenStudio and Agent Orchestrator scope and pricing, Adobe MAX 2026 dates, and generative credit rates.
- Jasper’s 2026 layoffs, valuation, and ARR. Writer’s 2026 funding.
- Iterable’s fall agents (Monitoring, Analytics, Nova) and their autonomy, from a release post only (2026-09-24).
How was this study done?
- Method
- Public sources only. Desk study of AI SDR agents, data enrichment, sales engagement, and AI marketing agents. Research window 2026-10-01 to 2026-10-30, with earlier background where the structure needs it. The study is refreshed monthly, and this edition is the first, the October 2026 edition.
- Sample
- Desk study. 99 public sources: vendor pages, pricing pages and docs, filings and earnings material, analyst and press coverage, and government pages.
- Confidence
- Every claim comes from a public source and carries a link. Vendor claims are marked (vendor claim) and are unverified. Third-party evidence covers analyst reports, surveys, filings, trade press, and government pages. Web search was limited during the research window, and several analyst, review, and vendor pages blocked automated reads, so some claims rest on press releases, wire copies, or search summaries, and each is labeled. The evidence shows what vendors publish and how analysts and the press describe it. It cannot show what vendors win or what agents do in production. The final section lists the items that stay unverified.
Sources
- Gong
- Outreach
- Rework (resources.rework.com)
- Industry Lens (industry-lens.com)
- Outreach
- SAP (news.sap.com)
- Insider One
- report excerpt (business.adobe.com)
- Writer
- ActiveCampaign
- Salesforce
- Clay
- Gong
- Globes (en.globes.co.il)
- martech.org
- CMSWire
- 11x
- Mordor Intelligence
- Precedence Research
- Grand View Research
- Precedence Research
- Precedence Research
- 8-K (sec.gov)
- 8-K (sec.gov)
- 8-K (sec.gov)
- 8-K (sec.gov)
- 8-K (sec.gov)
- Outreach
- GitHub
- HubSpot
- Resolve247
- AI Weekly
- MarTech
- 11x terms
- AiSDR
- page (11x.ai)
- page (11x.ai)
- page (artisan.co)
- page (aisdr.com)
- page (regie.ai)
- page (qualified.com)
- page (salesforce.com)
- page (salesforce.com)
- catalog (legal.hubspot.com)
- page (microsoft.com)
- page (clay.com)
- page (gtm.ai)
- page (commonroom.io)
- page (unifygtm.com)
- page (amplemarket.com)
- page (jasper.ai)
- blog (writer.com)
- page (klaviyo.com)
- PricingSaaS
- HubSpot
- UsagePricing
- HubSpot
- release notes (zoominfo.com)
- Salesforce
- Business Wire via 01net
- Gong
- Klaviyo
- Incepta (inceptasolutions.com)
- Clay
- docs
- MCP (clay.com)
- MCP (developers.hubspot.com)
- docs (gtm.ai)
- credits (lusha.com)
- blog (cognism.com)
- Kelley Drye
- Wiley
- Perkins Coie
- court PDF (media.ca11.uscourts.gov)
- FR 2026-18853 (public-inspection.federalregister.gov)
- Texas Legislature (capitol.texas.gov)
- CPPA
- enrolled bill
- Colorado Legislature
- McCann FitzGerald
- Silicon Canals
- Google (support.google.com)
- Yahoo Sender Hub (senders.yahooinc.com)
- Microsoft Support
- Clay
- TechCrunch
- Rework (resources.rework.com)
- TechCrunch
- 11x
- Momentum
- Warmly
- Highspot
- Salesforce
- Braze 8-K (sec.gov)
- Salesloft
- 8-K (sec.gov)
- 8-K (sec.gov)
- Sacra
- Artisan
Published .
Cite this
Couch, C. (2026). AI Sales and Marketing Tools Market: October 2026. Ends at Cash. https://chriscouch.ai/research/ai-sales-marketing-tools-market-2026/
BibTeX
@misc{couch2026aisalesmarket,
author = {Couch, Chris},
title = {AI Sales and Marketing Tools Market: October 2026},
year = {2026},
howpublished = {Ends at Cash},
url = {https://chriscouch.ai/research/ai-sales-marketing-tools-market-2026/}
}