Where Value Accrues In The Modern Revtech Stack

For the last year, we’ve spent a lot of time thinking about the cracks forming in B2B go-to-market (GTM).

Cold outbound has become harder, CRMs are being asked to do things they were never really designed to do, and while AI investment has poured into prospecting and sales execution, some of the most valuable parts of the customer lifecycle are surprisingly still underbuilt.

The data has been telling this story for over a year now. The Bridge Group’s 2025 study of 351 B2B companies found that only 60% of sales development representatives (SDRs) were hitting quota, the lowest level in the study’s history. Belkins, looking across 7.5 million cold emails, found unique reply rates of just 0.45%.

But we don’t think the interesting takeaway is simply that outbound is getting harder, rather we see a structural change happening. AI has already begun collapsing what used to be a series of discrete GTM stages into a more continuous process.

On the Seed team at Alumni Ventures (AV), that shift has changed how we think about investing in RevTech, and we’re increasingly less focused on who builds the best standalone tool and more focused on which parts of the stack become more valuable as these systems start working together.

Ahead of our upcoming SF Tech Week GTM event, we wanted to share how we’re currently thinking about where value is likely to accrue across the RevTech stack.

The Funnel Was Never Really a Funnel

For decades, GTM software has reflected the organizational structure around it: marketing generates leads, sales converts them into opportunities, customer success inherits the account, support handles problems, and revenue operations (RevOps) tries to make the systems talk to one another.

Software used to follow very similar boundaries: a lead lived in one system, an opportunity lived in another workflow, product usage might live somewhere else entirely, and billing, support, and customer health each developed their own systems and owners.

While that created what would seem like a seamless funnel, customers usually don’t move through companies in clean stages. For example, a customer can submit a support ticket, increase product usage, visit the pricing page, and bring a new division into a trial in the same week. Depending on which system you are looking at, the exact same account could simultaneously look like a churn risk, a support issue, and a major expansion opportunity.

We heard a real-world case of this recently from a chief revenue officer (CRO) in our network. They said one of their largest customers filed a support ticket around a missing integration. Two days later, the account hit 80% of its usage limit. At roughly the same time, usage of a newer AI feature had tripled over the prior month.

Historically, those were three different signals living in three different systems and probably owned by three different teams. In an AI-native GTM stack, however, those signals can be interpreted together, and while that sounds like a small change, we believe it is a fundamental one. The funnel was partly a product of the limitations of the software underneath it, and as those limitations disappear, some of the organizational boundaries built around them start to look much less permanent.

From Funnel to Feedback Loop

We think the emerging RevTech stack can be simplified into three layers:

These three layers are becoming much more tightly connected. As that loop starts happening continuously, the question becomes less about whether each layer matters and more about where the most value is likely to accrue.

When we look at the market through that lens, we think the economics of the three layers are starting to diverge.

While Signal Is Valuable, Owning It Might Not Be

The first layer has received a lot of attention, which isn’t exactly a surprise as better signals improve almost everything downstream. That has made signal collection and enrichment one of the most crowded areas of the RevTech stack.

We’ve seen Apollo build a contact and intent graph, Gong turn sales conversations into structured intelligence, product-led-growth platforms capture usage signals that CRMs largely missed, and community and intent platforms surface customer behavior that historically remained invisible.

But one of the most interesting things to happen in RevTech recently has been how many first layer standalone companies have stopped being standalone. Apollo acquired Pocus, Zoom acquired Common Room, Salesforce acquired Bluebirds, HubSpot acquired Dashworks and Warmly, and Salesloft and Clari combined.

This pattern we’re seeing matters because it shows signal is valuable, but it often becomes much more valuable when it is connected directly to systems that act on it. Which raises a more important question for companies: do you own proprietary intelligence, or are you simply collecting data that a larger platform could eventually integrate itself?

Those are very different businesses, and we are much more interested in companies that create unique customer context through their own product and become harder to replicate as they learn from more usage and outcomes.

Execution Is Getting Cheap

We see a different dynamic beginning to emerge in execution, where AI is lowering the cost of many GTM activities that once required much more software spend, human attention, or, in many cases, both. Account research, sequence generation, personalized outreach, follow-ups, and support can now be done faster and with far less manual effort than before.

Right now we are seeing hints of that pressure in pricing: Regie.ai moved from a $499-per-user monthly product earlier this year toward a $49 self-serve offering and Unify introduced a $20 entry point as it broadened its product.

And that leads us to the part of RevTech we currently find most interesting.

Good Judgment Is Scarce

Imagine an account where usage falls 15%, the CFO visits the pricing page, the internal champion files a support ticket, a second business unit starts a trial, and the renewal is 70 days away.

In that situation, generating an email is probably the easiest part. The harder question is whether anyone should reach out, who should do it, and what that response should look like. Maybe the account should receive an automated message, maybe a human should step in, maybe the company should push an upgrade or offer a discount, or maybe the best decision is to leave the customer alone.

These are decisioning problems, and as execution becomes cheaper we think decisioning solutions, like what should happen, when it should happen, and who should do it, become much more valuable.

That is why we are particularly interested in the middle layer, where the value comes from learning which decisions actually lead to better outcomes and using that history to improve the next one.

As these systems take on more of that responsibility, the challenge becomes earning enough trust to influence important decisions. That raises one of the biggest questions in RevTech over the next several years: who earns the right to decide?

What We Look For

If the value of the middle layer comes from making better decisions over time, then the most interesting products collect proprietary data while also building proprietary outcomes. They can see what happened after a decision was made, whether the customer renewed, the deal closed, usage increased, or an intervention actually reduced churn, and use that result to improve the next decision. We believe a product that only observes activity can build a better picture of the customer, while a product that connects decisions to outcomes has the chance to get better every time it is used.

Why Startups Could Still Win in Decisioning

If decisioning becomes the most valuable layer, the obvious question is why an existing platform like Salesforce or HubSpot would not just own it. Incumbents clearly have advantages around distribution and the amount of customer data already sitting inside their products, but having more data does not necessarily mean making better decisions, especially when those decisions depend on context that sits across several different systems.

That is where we think startups can still have an advantage. A decisioning product used across many companies in the same market can see how similar decisions play out in different environments, giving it a broader set of outcomes to learn from and a better understanding of what tends to work across a category. At the same time, because it can sit across product usage, support, pricing, and sales activity, it is not limited to the context inside any one incumbent platform.

Over time, that combination of cross-system context and cross-customer learning can become a real source of defensibility, allowing the product to improve its decisions and earn more trust and workflow over time.

One More Area We’re Watching

One additional place we think this architecture remains particularly underbuilt is post-sale.

Most of the visible AI investment in GTM has attacked acquisition first. Which makes sense since sales is large, measurable, and expensive, and companies will pay for more pipeline.

The economics of the customer base tell a somewhat different story. At the median, roughly 40% of net-new annual recurring revenue (ARR) now comes from expansion. And according to Benchmarkit, generating an expansion dollar requires roughly $0.80 of sales and marketing spend, compared with $1.63 for a new-customer dollar.

This is a big gap, yet in many organizations, customer success still relies on health scores, periodic check-ins, and dashboards built around lagging indicators. We think that is a mismatch with the rest of the stack, especially because post-sale may be where the feedback-loop model becomes most valuable. Expansion and retention depend on understanding a wide range of customer signals together, not in isolation.

That’s why we think some of the most interesting RevTech companies may be the ones that bring post-sale into the same decisioning loop as the rest of the revenue engine, rather than treating customer success as a separate workflow.

This doesn’t mean the CRM is any less important, but its role has just consolidated around recording what has already happened, while we expect the next generation of RevTech to be defined by systems trusted to decide what should happen next.

We’ll be exploring many of these questions at our upcoming SF Tech Week event with GTM operators, founders, and investors, focusing on where AI is actually changing the way companies sell, where human judgment still matters, and what that means for the next generation of RevTech.

That last question is one we’ll come back to in Part II.

About Authors

Interested in Seeing Curated Venture Deals?

Sources

1. https://belkins.io/blog/cold-email-response-rates
2. https://www.bridgegroupinc.com/research/2025-sdr-models-metrics-report-the-bridge-group
3. https://regie.ai/pricing
4. https://www.benchmarkit.ai/2026-saas-ai-native-metrics
5. https://webflow.unifygtm.com/blog/announcing-unifys-next-chapter

This communication is from Alumni Ventures, a for-profit venture capital company that is not affiliated with or endorsed by any school. It is not personalized advice, and AV only provides advice to its client funds. This communication is neither an offer to sell, nor a solicitation of an offer to purchase, any security. Such offers are made only pursuant to the formal offering documents for the fund(s) concerned, and describe significant risks and other material information that should be carefully considered before investing. For additional information, please see here. Achievement of investment objectives, including any amount of investment return, cannot be guaranteed. Co-investors are shown for illustrative purposes only, do not reflect all organizations with which AV co-invests, and do not necessarily indicate future co-investors. Example portfolio companies shown are not available to future investors, except potentially in the case of follow-on investments. Venture capital investing involves substantial risk, including risk of loss of all capital invested. Diversification cannot prevent investment loss; it is a strategy to mitigate investment risk. This communication includes forward-looking statements, generally consisting of any statement pertaining to any issue other than historical fact, including without limitation predictions, financial projections, the anticipated results of the execution of any plan or strategy, the expectation or belief of the speaker, or other events or circumstances to exist in the future. Forward-looking statements are not representations of actual fact, depend on certain assumptions that may not be realized, and are not guaranteed to occur. Any forward-looking statements included in this communication speak only as of the date of the communication. AV and its affiliates disclaim any obligation to update, amend, or alter such forward-looking statements, whether due to subsequent events, new information, or otherwise.