Will verticalized sales tools supercharge industrial digitization?

I recently found myself going down the rabbit hole of the future of enterprise resource planning (ERP) in an agentic world- a massive, undefined question sitting at the intersection of AI advancement and the operating systems that underpin nearly every organization. While It’s easier to imagine AI breakthroughs in digitally native sectors, I spend most of my time investing in legacy industries, which haven’t even broadly adopted cloud software, let alone agentic workflows. 

These supply chain-heavy sectors run on tight margins and complex processes, yet many core workflows remain inefficient and undigitized. Commonly, factories and suppliers manage quotes, orders, and inventory with spreadsheets, emails, and phone calls. Surveys show that even among manufacturers who think they’ve “gone digital,” nearly one-third still collect most of their data manually. 

There are a few reasons for this:

• A lack of in-house tech expertise is a major culprit: 46% of manufacturers say they lack the right skills on their team to adopt AI/ML solutions. 
• Cultural resistance also plays a role: many plant managers and sales reps have done things a certain way for decades and are wary of change. 
• Concerns about cybersecurity and data privacy add friction: Notably, manufacturing has become a top target for ransomware, accounting for over 70% of ransomware incidents in one UK 2024 report. 
• Big, general-purpose software (like traditional ERPs) feels too risky to rip-and-replace.

Therefore, while the headlines focus on AI first tools for knowledge workers, I’m more interested in the first principles changes still to come in legacy verticals. If there is a drastic ERP evolution, it’s not starting with a rip and replace play. Instead, it begins with point solutions that solve a real pain point on day one and quietly become the foundation for broader transformation over time. These tools earn trust by delivering immediate value, then expand into adjacent workflows as users grow more comfortable and reliant on them.

The inevitable “why now?”

A few forces at play make a shift possible and timely: Industrial sectors are battling labor shortages and an aging workforce, with nearly half a million U.S. manufacturing jobs sitting unfilled, and as baby boomers retire, that shortfall could hit 1.9 million positions by 2033. As fewer people enter these fields, the younger, smaller cohorts must be more productive than their predecessors, a realistic aspiration given their superior tech savviness having grown up in much more digitally native environments. 

At the same time, advances in AI, cloud infrastructure, and API connectivity have reached a point where even small startups can build very sophisticated solutions quickly. The fact that AI in manufacturing software is growing ~40% annually reflects both supply (new tech capabilities) and demand (industries ready to benefit) explosion. Modern UX and workflow design, influenced by mainstream SaaS, also means these tools are more user-friendly than the clunky software historically used.

Designing a playbook: Start narrow, figure out the workflow, expand into other functions

We have seen countless variations of the “building a true vertical operating system” pitch, and therefore this is where investors tend to click away from the deck. A key question here: which is the right, sticky point solution to run a land and expand playbook?

Digitizing the sales process may be the ideal way to break the inertia: modernizing how a company sells and how it responds to customers’ requests delivers immediate, tangible ROI. Sales remains one of the most underserved functions by legacy systems, both on the commercial and technical fronts, where multiple processes and cross-functional stakeholders breed a perfect storm for friction heavy, long sales cycles.

I recently went deep on this with Sidd Gupta, who’s built a Sales AI tool for chemicals, materials, and advanced manufacturing. On the commercial side, quoting is still a slow, email-heavy process in many industries, largely because older tools weren’t built for seamless collaboration across companies. RFQs often involve CAD files, spec sheets, and multiple supplier clarifications, resulting in long email threads and delays. On the technical sales side, reps depend heavily on overstretched product teams, a challenge made worse by workforce retirements. They're also expected to identify the right products across sprawling SKU catalogs, despite often being trained on only a fraction of them, leading to missed opportunities for cross-sells and upsells across business units.

Once a vertical sales tool gains traction and, in a perfect world, becomes the standard way a factory generates quotes or an airline buys parts, it amasses valuable data and user trust. We have seen how this creates an opportunity to broaden the product into a more end-to-end operating system for that industry, similar to the Salesforce strategy. 

Within manufacturing, a quoting platform can grow upstream, by plugging into engineering and R&D, as well as downstream, by extending into production planning, job scheduling, inventory management, distribution, and quality tracking. 

In practice, this requires integration with many legacy tools (CAD software, existing ERPs, etc.) or replacing them outright. No entrant is immune to the challenges triggered by encroaching on incumbents’ turf, which suggests that successfully navigating this may look like becoming a single pane of glass for users in that vertical, making switching back and forth unnecessary. There are several other elements at play that can cause friction, including IT team dynamics and utilization of CRMs and PIMs, which vary widely depending on customer size and segment. Further, decisioning on the right entry point for an entrant solution will likely be informed by a focus on discrete manufacturing versus process manufacturing, which have nuanced and distinct needs and capital expenditures across sales and R&D.

A new breed of vertical sales automation startups

‍

The above is not an exhaustive list. Dozens of others are emerging across industrial domains, including chemical sourcing platforms and construction bidding software. Even incumbents have noticed the pattern, evident in large enterprise software providers (Salesforce, SAP, etc.) who are adding configure, price, quote (CPQ) modules and industry cloud offerings. 

While I am still thinking about how this all plays into bigger questions associated with the future of ERP in an agentic world, I increasingly find this wedge compelling. Even if, in the long run, this becomes a data collecting play, it could serve as the foundation/ infrastructure on which future autonomous workflows and AI employees can thrive (and perhaps replace ERP as we know it entirely).

If you're building or operating at the edge of manufacturing, data, and intelligent systems, I would love to connect.

‍

A special thank you to- 

Isabelle Styslinger who has been an instrumental thought partner

Sidd Gupta who shared his learning building Nesh

Lauren Roberts and Sruthi Sivanandan for your competitive landscape research and support

‍

Naomi Goez is a Principal Investor at Forum Ventures who is focused on deal origination, evaluation, and execution. Follow her on Linkedin for more market insights and bets.