Profound completes $180 million Series D at a $1.8 billion valuation: the AI marketing platform is moving from an analytics tool to an agent orchestration layer

In the two years during which the investment narrative around generative AI has broadly shifted from the “model layer” to the “application layer,” there are not many niche scenarios that truly prove enterprises are willing to keep paying. Marketing is one of the easiest to validate and also one of the easiest to question: budgets are ample, the path is clear, and results are measurable, but at the same time it is highly fragmented, and the cost of replacing tools is low.

On September 15, 2026, Profound announced the completion of a $180 million Series D at a $1.8 billion valuation, co-led by Sequoia Capital and Kleiner Perkins. The round came less than seven months after its $96 million Series C—when the company was valued at $1 billion. Seven months, a valuation increase of about 80%—that pace itself is worth reading as an industry signal, not just as a piece of financing news.

Industry Background: AI Search Has Reshaped the Entry Point to “Being Discovered”

For the past two decades, the logic of how corporate marketing budgets are allocated has been built on a stable premise: users discover brands and products through search engines, social feeds, and retail shelves. SEO, SEM, content marketing, and e-commerce operations are all roles and toolchains built around this premise.

What generative AI changes is this premise itself. When users begin asking AI assistants directly, “Which one should I buy?” “Which supplier is more suitable?”, brand visibility is no longer determined by keyword rankings, but by whether the model cites it, how it describes it, and where it mentions it when generating answers. This is competition at the semantic layer rather than the ranking layer, and it lacks mature measurement standards.

Profound’s starting point was precisely this measurement gap: providing marketing teams with analysis of “brand visibility in AI answers.” The company says its platform is built on more than 2 billion real user prompts; this scale of data is the foundation of its product’s usability—without a large-scale corpus of prompts and answer samples, it would be impossible to turn “AI mentions” into comparable metrics.

From Analytics to Execution: The Three-Layer Evolution of the Product Structure

To understand this Series D, the key is to understand that Profound’s product has already moved from “observation” to “execution.”

Layer One: AI Search Visibility Analytics. This was the company’s core product when it was founded in 2024, solving the measurement problem of “how a brand is presented in AI answers.”

Layer Two: Context Manager. This component distills unstructured materials such as enterprise knowledge base files, meeting notes, and communication threads into reusable brand context, enabling AI Marketer to continuously adjust its output as brand positioning changes. The industry significance of this layer is often underestimated: one of the biggest friction points in enterprise AI deployment is that the model lacks continuous understanding of dynamic knowledge within the organization, not merely that model capabilities are insufficient.Layer Three: AI Marketer and Ads Studio. AI Marketer is described as an Agent orchestrator that scans brand data, analyzes AI answers, identifies opportunities, and dispatches sub-Agents to carry out specific work. Ads Studio, meanwhile, lets marketers build, launch, and manage AI search ad campaigns in OpenAI, Google, and Meta ad managers, and complete tracking through Profound’s own pixel.

This three-layer structure outlines a clear commercialization path: first measurement (building a data moat), then context (raising switching costs), and finally execution and a budget entry point (tapping into ad spend). Advertising placement is the link closest to budget in the marketing technology stack; once a platform can simultaneously explain “what happened in AI answers” and directly act on “how much to spend on this,” its role shifts from analytics tool to workflow system.

Customer Structure and Commercial Validation

The company’s disclosed customer list spans consumer goods, finance, and technology: Comcast, The Estée Lauder Companies, Walmart, Campari Group, Royal Bank of Canada, as well as Zoom, ServiceNow, Ramp, Cursor, MongoDB, Figma.

Its coverage ratio is the more noteworthy metric: more than one-third of the Fortune 100 and 16% of the Fortune 500. At Series C, the corresponding figures were more than 700 enterprises and over 10% of the Fortune 500, with more than 500 additional customers using Profound Agents daily. By Series D, the number of enterprise customers exceeded 1,000.

In seven months, the customer count grew from more than 700 to over 1,000, and Fortune 500 penetration rose from 10% to 16%, indicating that growth was not driven only by expanding individual accounts; new large customers continued to enter. For a company founded as recently as 2024, this is relatively rare sales efficiency.

Enterprise-side governance capabilities are likewise a procurement threshold. Capabilities listed on Profound’s website include role-based access control (RBAC), SCIM identity provisioning, and SOC 2 compliance, along with support for more than 200 integrations. In large enterprises, these are usually not nice-to-haves but entry requirements.

Fundraising Cadence and Capital Logic| Date | Round | Amount | Valuation | Lead Investors | | --- | --- | --- | --- | --- | | 2024-08-12 | Seed round | $3.5 million | Not disclosed | Khosla Ventures, Saga Ventures, South Park Commons | | 2025-06-18 | Series A | $20 million | Not disclosed | Kleiner Perkins | | 2025-08-12 | Unspecified round | $35 million | Not disclosed | Not disclosed | | 2026-02-24 | Series C | $96 million | $1 billion | Lightspeed Venture Partners | | 2026-09-15 | Series D | $180 million | $1.8 billion | Sequoia Capital, Kleiner Perkins (co-led) |

At the Series C, the company disclosed cumulative funding of more than $155 million, and at that time it was just one day shy of 18 months since founding. With this round’s $180 million, cumulative funding has entered the $300 million range.

It is worth noting the stability of the investors. Khosla Ventures, Saga Ventures, and South Park Commons followed from the seed round through Series D; Kleiner Perkins entered at Series A and co-led Series D; Lightspeed continued to participate after leading Series C; Sequoia was upgraded from a Series C participant to co-lead of Series D. In a cycle where application-layer companies generally face doubts about “valuation inversion and existing shareholders not following on,” this structure of continuous reinvestment says more than any single round’s amount.

Market impact: the three parties are not affected equally

For enterprise customers. Procurement logic is shifting from “buying an analytics tool” to “buying a governable Agent workflow.” This means the evaluation dimensions are changing: no longer whether the dashboard looks good, but the Agent’s permission boundaries, whether context sources are auditable, whether outputs are traceable, whether it can integrate with existing identity systems (SCIM, RBAC), and whether it meets compliance requirements such as SOC 2. Profound putting governance capabilities front and center on its product page reflects exactly what enterprise buyers actually care about in 2026.For investors. This funding round provides a new pricing anchor for whether the AI application layer can sustain high-multiple valuations: from zero to $1.8 billion in about 18 months. It will push up valuation expectations for similar companies and also increase market scrutiny of revenue quality—especially when the customer base includes many Fortune 500 companies, requiring a distinction between pilots and full deployments.

For traditional martech vendors. The pressure comes from a shift in the entry point. When a brand’s AI search ad placements can be managed uniformly across OpenAI, Google, and Meta on one platform, the functions previously handled separately by SEO tools, ad management platforms, and analytics suites within the martech stack begin to show a consolidation trend.

Competitive Landscape: Who Benefits, Who Is Under Pressure, and Who May Follow

The beneficiaries are three types of players. First, application-layer companies with real enterprise customers and governance capabilities, which can use this valuation anchor to raise capital; second, frontier model vendors and cloud infrastructure providers, because agent orchestrators essentially amplify inference calls and token consumption, representing incremental demand at the AI infrastructure layer; third, ad platforms able to capture the new budget category of “AI search advertising,” with the ad managers of OpenAI, Google, and Meta gaining a new intermediary layer.

Those under pressure are mainly tool vendors with narrow functional scope. Products that only provide “AI visibility monitoring” will find it difficult to compete with integrated platforms that have extended downward into context management and upward into ad execution and budget entry points, because switching costs for the latter rise with depth of use.

Possible directions for follow-on moves include three paths: first, traditional SEO and digital experience analytics vendors layer agent execution capabilities on top; second, large marketing clouds (such as Salesforce’s marketing product line) incorporate AI search visibility into their existing enterprise suites, using integrated procurement to compete against point solutions; third, frontier model vendors or ad platforms themselves provide official measurement and ad placement tools, directly capturing the middle layer. The third path poses a long-term structural risk to Profound: part of its value rests on cross-platform neutrality, and neutrality only commands a premium when platform operators do not offer equivalent capabilities.

Enterprise Implications: What Should Enterprises Watch?1. Build AI visibility into formal metrics systems. If customer acquisition paths have partially shifted to AI chat interfaces, then in addition to brand health and search share, metrics for brand presence in AI answers should be established; otherwise, the basis for allocating marketing budgets will be systematically distorted. 2. Governance capability comes before capability ceilings. Before deploying marketing Agents, clarify data sources, permission boundaries, output review, and audit trails. SOC 2, RBAC, and SCIM have become gating requirements because once an Agent is connected to a brand knowledge base and ad budget, the cost of erroneous output is far higher than with traditional software. 3. Distinguish pilots from scale. Vendor-reported customer counts and penetration rates reflect contracted coverage, not full deployment or actual ROI. Buyers should, during evaluation, request deployment depth data by business unit and by scenario, and require quantifiable cost-savings or efficiency-improvement metrics. 4. Preserve cross-platform bargaining power. Whether choosing to build in-house or buy, avoid handing over AI search measurement and campaign execution entirely to a single intermediary, to guard against a passive position after platforms launch official tools. 5. Pay attention to interpretive authority at the model post-training layer. Profound plans to conduct post-training for marketing scenarios and build evaluation benchmarks, which is essentially a contest over the authority to define “what counts as good marketing work.” Once definitional authority is established, it will in turn affect internal performance standards and vendor evaluation methods.

Use of Funds: From Product Company to Research Institution

This round of funding will be used to expand its applied AI labs in New York and San Francisco. According to company disclosures, the team will study how frontier models perform on marketing work, evaluate model capabilities, and conduct model post-training for marketing scenarios.

The composition of this research team is noteworthy: research scientists, International Physics and Informatics Olympiad medalists, and economists. The researchers are developing a benchmark to test how far AI can go on complex marketing tasks—whether models can understand a business, weigh conflicting market signals, and produce usable results across marketing functions. Ali Vaghar, head of data at Profound, said the company is building the technical rigor and first set of evaluation methods to make AI truly useful for marketers.From an industry perspective, this is a classic “application company moving up into the research layer.” The reason is not hard to understand: if product capabilities depend entirely on frontier models that any competitor can call, the moat can only come from proprietary data and workflow depth; but once a company begins post-training for vertical scenarios and defining evaluation standards, it adds a layer of hard-to-replicate assets on top of the model. Sequoia Capital partner Anas Biad said that large brands are rapidly applying AI to marketing teams, and Profound has established its position at the application layer. Kleiner Perkins partner Ilya Fushman described this evolution as expanding from AI search into a broader platform for how modern marketing teams operate.

Future Outlook

Within 12 months. The key thing to watch is the actual adoption rate of Ads Studio—that is, whether customers really entrust their AI search advertising budgets to the platform rather than merely using its analytics features. At the same time, watch whether the benchmark is publicly released and cited by the industry—acceptance of evaluation standards by third parties is a precondition for establishing the power to define them. Competitor follow-through is almost certain: AI search visibility monitoring will quickly become commoditized, and differentiation will concentrate at the two ends of context management and campaign execution.

Within 24 months. The market will most likely stratify: a small number of platform companies will control the budget entry point and evaluation standards, while many point solutions will be consolidated or marginalized. Enterprise procurement will make AI governance capabilities (permissions, auditing, compliance) a hard requirement, and vendors lacking these capabilities will be excluded from large enterprises. At the same time, the likelihood increases that advertising platforms and frontier model vendors will launch official measurement tools, and the middle layer’s value proposition will be forced to shift from “measurement” to “cross-platform orchestration and accountability for outcomes.”

On a three-year horizon. The more substantive change is organizational structure rather than tools. When the discovery entry point for marketing shifts from search rankings to model-generated answers, marketing teams will need Agent systems that can understand the business, weigh signals, and execute multi-functional tasks—not more dashboards. This will redefine the boundaries of the martech stack and will also bundle marketing budgets, data governance, and AI governance into the same procurement decision. For investors, the criterion will shift from “whether it has enterprise customers” to “whether it controls irreplaceable context assets, evaluation standards, or budget entry points.”

Issues Still to Watch

Profound’s growth curve is currently supported by customer count and Fortune 500 penetration, but public information has not yet disclosed revenue scale, net revenue retention, or gross margin structure. The growth expectations implied by this round’s valuation are quite aggressive. In addition, the customer list in the company’s external statements includes a large number of technology companies, and these customers’ willingness to pay may differ from that of large traditional enterprises.For industry observers, the value of this financing does not lie in the amount itself, but in the clear judgment it signals: marketing budgets are migrating toward AI conversational interfaces, and around this migration, the first consolidator with large-scale capital backing has emerged across the three layers of infrastructure—measurement, context, and execution. The next question is not whether this path exists, but whether it can, before the platforms themselves enter the fray, convert neutrality and context assets into sufficiently high switching costs.