Gong launches ‘Mission Andromeda’ with AI gross sales teaching, chatbot and open MCP connections to rivals

Gong, the income intelligence firm that has spent a decade turning recorded gross sales calls into information, at the moment launched what it calls Mission Andromeda — its most bold platform launch thus far, bundling a brand new AI-powered teaching product, a sales-focused chatbot, unified account administration instruments, and open interoperability with rival AI methods by means of the Mannequin Context Protocol.

The discharge arrives at a pivotal second. The income expertise market is consolidating at a tempo that will have been unthinkable two years in the past, and Gong — nonetheless a personal firm with roughly $300 million in annual recurring income — finds itself on the heart of a class that Gartner solely formally outlined three months in the past. Mission Andromeda is Gong's reply to a primary query going through each enterprise AI vendor in 2026: Can you progress past surfacing insights and truly change how folks work?

"The entire present, Andromeda, is principally a group of very vital capabilities that take us an enormous step ahead," Eilon Reshef, Gong's co-founder and chief product officer, instructed VentureBeat in an interview forward of the launch. He described it as an effort to make income groups "extra productive as people" and to provide leaders "higher selections" — positioning the discharge not as a characteristic dump, however as an working system improve.

The brand new merchandise goal each layer of the gross sales workflow, from teaching to account administration

Mission Andromeda incorporates 4 essential elements, every focusing on a special layer of the gross sales workflow.

The headliner is Gong Allow, a brand-new product with its personal pricing tier — Reshef described it as "within the tens of {dollars} per seat monthly" — that assaults what the corporate sees as a gaping gap in most gross sales organizations: the disconnect between coaching and efficiency. Highspot and Seismic introduced their intent to merge in February 2026, making a mixed enablement large, and Gong is now transferring instantly onto their turf.

Gong Allow has three items. The primary, AI Name Reviewer, analyzes accomplished buyer calls and grades reps based mostly on their group's personal methodology. When requested whether or not this operates in actual time, Reshef was direct: "For that exact agent, it's post-call, as a result of clearly you wish to grade the entire name as an entire — perhaps you didn't do something in minute one, minute 30." The second piece, AI Coach, lets reps follow high-stakes conversations — pricing objections, renewal danger situations — in opposition to AI-generated simulations constructed from the corporate's personal profitable name patterns. The third, Initiative Monitoring, hyperlinks teaching applications to income metrics so leaders can see whether or not new behaviors truly present up in reside offers.

Past Allow, the launch contains Gong Assistant, a conversational AI chatbot purpose-built for income groups that lets customers ask questions on buyer calls contained in the platform. The discharge additionally introduces Account Console and Account Boards, which unify buyer exercise, danger indicators, and subsequent steps right into a single view for gross sales and post-sales groups. And rounding out the bundle is built-in assist for the Mannequin Context Protocol, the open commonplace initially developed by Anthropic, enabling Gong to trade information with AI methods from Microsoft, Salesforce, HubSpot, and others.

Gong makes use of 4 completely different LLM suppliers and says the true moat is the info, not the fashions

In a market the place each firm desires to assert proprietary AI supremacy, Reshef described a notably pragmatic method to the fashions powering the brand new options. Gong makes use of each inside fashions and basis fashions from exterior suppliers, he mentioned, noting that "4 out of the 5 main AI corporations, LLM, are principally Gong prospects."

The corporate picks fashions process by process. "Primarily based on the product or process at hand, we choose the suitable mannequin," he mentioned. "We’d generally swap out and in a mannequin if we really feel it's greatest for our prospects they usually get an increasing number of energy." Reshef drew a transparent line between what wants a big language mannequin and what doesn’t: "Our income prediction fashions are usually not utilizing LLMs, however type of the core interplay chatbots — in fact, you're going to make use of the muse mannequin."

This method contrasts with rivals which have hitched their wagon to a single AI supplier. It additionally displays a philosophical selection: Gong's actual moat, Reshef steered, will not be the fashions themselves however the information beneath — what the corporate calls the Income Graph, its proprietary layer that captures telephone calls, Zoom conferences, emails, textual content messages, WhatsApp conversations, and extra, stitching them collectively right into a related intelligence layer.

Recording each gross sales dialog raises apparent privateness questions, and Gong says it has spent a decade answering them

Storing and analyzing each buyer dialog a gross sales crew has raises apparent questions on privateness and information governance. Reshef was keen to handle them head-on.

"We've been across the block for a protracted whereas — somewhat bit over a decade — with AI first," he mentioned. "Through the years, we've developed precisely these capabilities which might be probably the most boring items of AI, which is: how do you gather the suitable information? How do you handle it? How do you handle permissions about it, retention insurance policies, proper to be forgotten?"

On the delicate query of whether or not Gong trains its AI on buyer information throughout accounts, Reshef drew a agency boundary. Coaching, he defined, occurs per buyer: "The vast majority of the coaching occurs based mostly on every buyer's information." He pointed to massive accounts like Cisco, which he mentioned has 20,000 Gong customers — sufficient information to coach the AI Coach from inside their very own surroundings. "AI Coach can go mine what's working of their surroundings. It may not work of their competitor's surroundings — perhaps their advantages are completely different, their objections are completely different."

Cross-customer coaching, he mentioned, occurs "solely in very, very uncommon circumstances, very protected based mostly — like transcription. However we don't do it for business-specific processes."

MCP offers Gong an open door to rival platforms, however safety stays an unsolved downside throughout the trade

Gong's assist for Mannequin Context Protocol is probably probably the most strategically vital piece of the launch. The corporate now presents built-in consumer and server assist for MCP, enabling organizations to attach Gong with different AI methods whereas sustaining clear controls over information entry, utilization, and provenance. Gong first introduced MCP assist in October 2025 at its Rejoice convention, the place it revealed preliminary integrations with Microsoft Dynamics 365, Microsoft 365 Copilot, Salesforce Agentforce, and HubSpot CRM. At this time's launch builds on that basis.

However Reshef didn’t sugarcoat MCP's limitations. "MCP could be very immature in relation to safety," he instructed VentureBeat. The protocol lets enterprise AI methods share information and context, however belief stays the enterprise's duty. He defined a two-sided mannequin: Gong can pull information from companions like Zendesk by means of licensed integrations, and concurrently makes its personal MCP server obtainable in order that instruments like Microsoft Copilot can question Gong's information. "It's as much as the corporate which connections they really really feel are safe sufficient," he mentioned. "The most secure ones are those that we've type of like licensed in a means. However MCP is an open protocol. They’ll join it to their very own methods. We’ve got no management over this."

That candor issues. As MCP adoption accelerates throughout the enterprise software program stack, safety groups are scrambling to grasp what occurs when agentic AI methods begin speaking to one another with out people within the loop. Gong seems to be betting that transparency concerning the protocol's immaturity will construct extra belief than advertising bravado.

Early prospects report sooner ramp occasions and better win charges, however the latest options are nonetheless days previous

When requested for exhausting numbers, Reshef supplied a mixture of platform-wide outcomes and measured candor concerning the latest options. Present Gong prospects report roughly a 50 p.c discount in gross sales rep ramp time and 10 to fifteen p.c enhancements in win charges, he mentioned.

However on Gong Allow particularly, he acknowledged the product remains to be model new. "The coach has been available in the market for actually, you realize, days, every week," he mentioned. "I might most likely mislead you if I mentioned, 'Hey, we're already seeing folks crushing it after taking three or 4 programs.'" For the sooner model of Allow that features the AI Name Reviewer, nonetheless, he mentioned prospects are "positively seeing a really excessive type of ability enchancment" and are attributing will increase in win charges and quota attainment to these beneficial properties — although he conceded that "it's all the time exhausting to do 100% attribution."

Morningstar, considered one of Gong's early adopters, supplied a pre-launch endorsement. Rae Cheney, Director of Gross sales Enablement Expertise at Morningstar, mentioned in an announcement that Gong Allow helped the agency "spend much less time on standing updates and extra time on the work that truly strikes offers."

Reshef insists AI nonetheless wants a human operator, placing Gong at odds with the autonomous agent hype

One of many extra attention-grabbing threads in Reshef's remarks involved his view of AI autonomy — or somewhat, its limits. He pushed again on what he referred to as a "frequent misperception about AI" — that it operates fully autonomously.

"There needs to be an individual within the center, which I name operator," he mentioned. "It may very well be RevOps. It may very well be enablement. Within the case of coaching, it may very well be analysts. Typically it may very well be even enterprise leaders." These operators, he argued, are chargeable for a "repeatable means of AI doing one thing, measuring the AI" and adjusting over time.

This philosophy extends to the AI Name Reviewer's suggestions. Gong doesn’t dictate what the system trains on — enablement leaders select. "We don't determine what they wish to prepare on. We allow them to select," Reshef mentioned. "You iterate, you optimize, you see the way it goes, and there needs to be any person within the group who's chargeable for ensuring this aligns with the enterprise wants."

That stance places Gong at odds with the extra aggressive "autonomous agent" rhetoric rising from some rivals, and it might resonate with enterprise consumers who stay cautious about letting AI run unsupervised in revenue-critical workflows.

A wave of mega-mergers is reshaping the income AI market, and Gong is racing to remain forward of the mixed giants

Mission Andromeda doesn’t exist in a vacuum. The income AI panorama has been reshaped by a outstanding wave of consolidation over the previous six months.

In a category-defining transfer, Clari and Salesloft merged in December 2025 to type what they referred to as a "Income AI powerhouse," combining roughly $450 million in ARR underneath new CEO Steve Cox. Simply two weeks in the past, Highspot and Seismic signed a definitive settlement to merge, making a mixed entity price greater than $6 billion targeted on AI-powered gross sales enablement — the exact same territory Gong is now invading with Allow.

In the meantime, Gong was named a Chief within the inaugural 2025 Gartner Magic Quadrant for Income Motion Orchestration, printed in December. The corporate positioned highest among the many 12 distributors evaluated on each the "Capacity to Execute" and "Completeness of Imaginative and prescient" axes and ranked first in all 4 evaluated use circumstances in Gartner's companion Crucial Capabilities report.

In his interview, Reshef didn’t title rivals instantly, however he drew a transparent distinction. "We've constructed a product from the bottom up. It's all natural," he mentioned. "All the different gamers within the discipline have type of stitched collectively instruments. And clearly you may't simply get it to be a coherent product when you simply sew collectively instruments. A few of them even have a number of logins." That could be a thinly veiled shot on the merged Clari-Salesloft entity, which Forrester has described as presenting a "bifurcated method" — Salesloft serving frontline customers whereas Clari helps administration insights.

Reshef additionally pointed to development as a aggressive weapon. "We're rising on the high deck aspect when it comes to SaaS corporations," he mentioned, including that Gong employed roughly 200 R&D workers this 12 months and plans to rent one other 200. "It's type of a flywheel the place we will make investments extra in R&D, we make the product higher, we get extra capabilities, extra flexibility, extra enterprise prospects."

Gong declines to debate IPO timing, however its structured launch cadence tells a narrative of its personal

Any dialogue of Gong's trajectory inevitably raises the query of a public providing. When requested instantly, Reshef declined to remark: "I wouldn't touch upon IPO at this stage. No."

The corporate has been on a transparent development arc. Gong has raised roughly $584 million thus far, with its final official funding spherical valuing it at $7.25 billion — the end result of a collection of speedy jumps from $750 million in 2019 to $2.2 billion in 2020. The corporate reached an annual gross sales run price of roughly $300 million in January 2025, pushed largely by the adoption of AI, in response to Calcalist.

However that valuation has since slipped. As Calcalist reported in November 2025, Gong is conducting a secondary spherical for firm workers and traders at a valuation of roughly $4.5 billion — effectively beneath its 2021 peak. The providing is being performed by means of Nasdaq's non-public market platform and was in superior phases on the time of the report. It isn’t but clear whether or not the corporate has repriced worker choices, a few of which have been issued at considerably increased valuations than the present secondary spherical. Gong instructed Calcalist that it "frequently receives inquiries from potential traders" however as a personal firm doesn’t "interact in hypothesis."

The structured quarterly launch cadence that Mission Andromeda inaugurates — full with galactic naming conventions and coordinated product narratives — definitely resembles the type of predictable, story-driven method that public market traders reward. Reshef framed it in a different way: "We felt like having quarterly launches with a reputation, a mission, and a narrative round it makes it simpler to work… It's a great way to teach the market regularly."

Gong's 50 p.c productiveness goal reveals the place it thinks the way forward for gross sales is heading

Reshef's most revealing remark got here when he laid out the corporate's long-term thesis: Gong goals to extend productiveness for income professionals by 50 p.c. "We're not there but," he admitted. "I believe we're like at 20 to 30 — no matter, exhausting to measure."

He broke the productiveness acquire into two classes. The primary is making high-complexity human duties — like conducting a reside Zoom gross sales name — higher, by means of teaching, coaching, and evaluate. "I believe there's going to be a protracted whereas, if ever, that Zoom conversations are going to get changed by bots," he mentioned. The second is automating the guide drudgery: name preparation, post-meeting summaries, follow-up emails, account analysis briefs.

The excellence issues as a result of it frames Gong's ambition not as changing salespeople however as making them dramatically simpler — a message calibrated to enchantment to the 1000’s of income leaders who management Gong's shopping for selections. Whether or not that thesis holds will rely on whether or not Gong Allow, the AI Coach, and the remainder of Mission Andromeda can ship measurable beneficial properties in a market that has been burned earlier than by instruments that promise perception however wrestle to alter conduct.

Gong presently serves greater than 5,000 corporations worldwide. The Clari-Salesloft merger has produced a rival with deeper mixed sources. The Highspot-Seismic mixture is assembling a gross sales enablement colossus. And a new Gartner class means each enterprise purchaser now has a framework for comparability buying. The subsequent twelve months will check whether or not Mission Andromeda is the discharge that cements Gong's place on the heart of the income AI class — or the final massive swing earlier than the consolidated giants shut in.

"Our mission is to be on the forefront," Reshef mentioned. "If everyone else is doing 20 p.c, we're going to do 50. If everyone goes to do 50, we're going to do 80."

Within the income AI wars, that type of confidence is simple to undertaking. Delivering on it, with brand-new merchandise nonetheless days previous and a market being remade round you in actual time, is one thing else completely.