From AI brokers to enterprise budgets, 20 VCs share their predictions on enterprise tech in 2025

Whereas AI is lauded by some as the largest technological breakthrough for the reason that industrial revolution, enterprises — arguably the tech’s greatest potential buyer base — have been gradual to undertake AI.

Whereas some traders predicted that 2024 could be the yr we’d begin to see extra AI adoption by enterprises, that didn’t play out as budgets remained constrained and AI tech usually remained within the “experimental” class.

Will that every one begin to change in 2025? Is determined by who you ask.

TechCrunch talked to twenty enterprise capitalists who again startups seeking to promote to enterprises about their predictions for 2025. They instructed us what they anticipate concerning enterprise budgets, traits value following, and what it would take to boost a Sequence A in 2025, amongst different issues. Right here’s what they mentioned.

SC Moatti, managing accomplice, Mighty Capital: I’m actually wanting into this theme — AI adoption hinges on higher knowledge. As enterprises transition from AI experiments to large-scale deployment, the demand for high-quality knowledge intensifies.

Aaron Jacobson, accomplice, NEA: Code brokers for app improvement modernization are underhyped. Anticipate to see AI getting used to re-platform mainframe apps to the cloud and improve older codebases.

Molly Alter, accomplice, Northzone: A key focus of mine is on areas that have been traditionally untouchable by enterprise funds as a result of their enterprise fashions demanded excessive COGS or OpEx. We’re seeing AI automate a lot behind-the-scenes work that sectors like accounting providers, or income cycle administration, or white-glove authorized providers can now command software-like margins.

Marell Evans, founder and basic accomplice, Distinctive Capital: Understanding traits in enterprise gross sales cycles — what’s the period sure organizations are trialing instruments for earlier than making choices about inner adoption? As well as, understanding the completely different pricing fashions of AI [in relation to] conventional SaaS, consumption-based and/or outcome-based.

Mike Hayes, managing director, Perception Companions: An unappreciated metric and one thing that I feel will achieve traction in 2025 is TTFV, or time-to-first-value. I see this as a proxy for ease-of-implementation, so quicker TTFV options ought to have a much bigger benefit going into [the] new yr.

What areas are you seeking to put money into?

Liran Grinberg, co-founder and managing accomplice, Team8: Enterprise resilience, whether or not in entrance of operational faults or malicious insider or outsider threats. The CrowdStrike software program replace incident demonstrated how fragile our digital world is, not solely because of cyberattackers but in addition simply errors.

Jonathan Lehr, co-founder and basic accomplice, Work-Bench: Information sovereignty as a service. Organizations are more and more investing in knowledge sovereignty options pushed by regulatory necessities and geopolitical issues. We’re exploring startup alternatives that allow firms to keep up full management over their knowledge’s location, storage, processing, and governance whereas guaranteeing compliance with native regulatory frameworks.

Mark Rostick, vice chairman and senior managing director, Intel Capital: One space we’re taking a look at is firms that concentrate on task-specific fashions. Whereas the foundational fashions are effectively established, I discover fashions that excel at particular capabilities significantly intriguing, particularly when mixed with brokers constructed on prime of them. As well as, we’re intently monitoring the event of options to transformers and any doable options to cut back the necessity for the large quantity of computing capability now required to coach LLMs and use them in manufacturing.

Mike Hayes, managing director, Perception Companions: Enterprises have traditionally considered expertise as both driving income or lowering value, however that’s shortly altering in favor of expertise that drives enterprise worth whereas concurrently lowering enterprise friction. I search for options that remedy distinctive, orthogonal challenges for enterprises — areas the place conventional options have fallen brief; this consists of vertical and persona-specific workflows reimagined with GenAI or agentic automation and safety improvements that not solely determine and alert, but in addition remediate.

Jason Mendel, enterprise investor, Battery Ventures: Just a few attention-grabbing areas the place I feel AI can add vital worth, and which I’m enthusiastic about, embrace observability / incident response, IT service administration, demand technology and gross sales engagement, offensive safety, software program improvement, and the SOC workflow.

Ed Sim, founder and basic accomplice, Boldstart Ventures: We predict in second order results. So if we assume that sooner or later, that means the following two to a few years, we might stay in a world the place every of us has dozens or a whole bunch of brokers doing work for us, we want to consider all the infrastructure that must be constructed to assist these new digital staff. Who’s going to offer the safety infra to offer entry management? Who’s going to handle these? Is there a platform to handle disparate brokers and safe them? What a couple of runtime system for Claude’s MCP, which looks like a dockerized, safe sandbox for brokers to do work.

What applied sciences, sectors, firms, and so forth., are you discovering attention-grabbing that aren’t AI?

Liran Grinberg, co-founder and managing accomplice, Team8: Quantum computing continues to be promising. Cybersecurity isn’t going anyplace as effectively, with attackers leveraging AI and an elevated complexity in defending our digital infrastructure.

Nina Achadjian, accomplice, Index Ventures: We’ve seen a resurgence in fintech, SaaS, and e-commerce, which have been scorching sectors that noticed a slowdown within the final couple of years. Past that, we anticipate cyber and gaming to proceed to be attention-grabbing this yr, with cyber accelerating additional because the IPO market opens up and laws and disclosure guidelines round safety improve.

Aaron Jacobson, accomplice, NEA: There’s a ton of hype round securing AI, however the larger alternative helps enterprises apply ”Cybersecurity 101” at scale in a means that doesn’t impede consumer productiveness. Key areas of specific curiosity are imposing least privilege entry, sustaining a safe knowledge posture, and stopping ransomware. I’m additionally excited to put money into expertise that facilitates multi-cloud deployments for enterprises.

Molly Alter, accomplice, Northzone: I’m actually enthusiastic about firms addressing the general public sector. The fiscal surroundings for presidency contracting is flush; complete federal company contracts reached $774 billion in 2023. Know-how adoption and modernization are key to driving the efficiencies that the brand new administration is committing to, and there’s a rising ecosystem of firms which might be tackling this head-on.

Andrew Ferguson, vice chairman, Databricks Ventures: We’re spending a major period of time with our system integrator accomplice ecosystem. These firms are doing the exhausting work of serving to enterprises take their knowledge and AI methods and switch them into real-world implementations.

Janelle Teng, vice chairman, Bessemer Enterprise Companions: We’re transferring past the fashionable knowledge stack. The info infrastructure panorama is present process a large transformation, fueled by numerous components, together with the rise of lakehouse structure and convergence towards particular open desk format requirements.

Raviraj Jain, accomplice, Lightspeed Enterprise Companions: Power is a big sector to put money into given growing demand for vitality for knowledge facilities and the challenges with grid failures throughout the nation. We’ll see continued curiosity in nuclear — each fusion and fission.

On the subject of AI, how are you figuring out that an organization has a moat?

Cathy Gao, accomplice, Sapphire Ventures: I give it some thought in a “5D framework”: design, knowledge, area experience, distribution, and dynamism. Since early this yr, we at Sapphire have used this framework to judge firms constructing purposes with AI.

SC Moatti, managing accomplice, Mighty Capital: An AI moat is constructed on proprietary knowledge, cutting-edge algorithms, and scalable infrastructure, enabling distinctive and superior options.

Scott Beechuk, accomplice, Norwest Enterprise Companions: The deepest moats might be created by giant proprietary datasets. The businesses with the best long-term potential are these constructing their very own distinctive datasets to excel of their specific, verticalized channels — usually by both coaching or fine-tuning their very own fashions.

Jonathan Lehr, co-founder and basic accomplice, Work-Bench: As a pureplay seed fund, we’re focusing most of our vitality in vertical AI alternatives tackling business-specific workflows that require deep area experience and the place AI is principally an enabler of buying beforehand inaccessible (or extremely costly to accumulate) knowledge and cleansing it in a means that might’ve taken a whole bunch or hundreds of man-hours.

Raviraj Jain, accomplice, Lightspeed Enterprise Companions: Query to ask is, As fashions grow to be higher, does this firm get threatened or strengthened?

What does it take to boost a Sequence A as an enterprise startup in 2025?

Liran Grinberg, co-founder and managing accomplice, Team8: With a powerful founder-market match, and an formidable imaginative and prescient to construct an enormous firm, one can increase a stable $15 [million to] $25 million Sequence A spherical with just a few $100Ks in ARR.

Molly Alter, accomplice, Northzone: Profitable Sequence A enterprise startups will present sturdy topline traction (>100% YoY) with low burn multiples; gone are the times of 2021 when it was all about progress in any respect prices. Extra importantly, these companies will present a transparent long-term differentiation technique that can set them aside from the host of different choices trying to boost cash and promote into the identical enterprise buyer base.

Kirby Winfield, founding basic accomplice, Ascend: Go from zero to $1 million in two quarters with an A-plus staff in a large market with a differentiated answer having created overwhelming demand.

Andrew Ferguson, vice chairman, Databricks Ventures: Should you’re constructing an AI-first product, an all-star technical staff and early product market traction ($2 [million to] $5 million ARR) could be the Sequence A expectation. The time from product launch to $5 million ARR is materially quicker within the AI period than it was within the conventional SaaS period. I anticipate that the Sequence B bar might be a lot greater — and it stays to be seen if this early ARR is high-quality and sturdy.

Jonathan Lehr, co-founder and basic accomplice, Work-Bench: We’re listening to from downstream friends that the bar is round $1.5 million with the power to develop 3x from there sequentially to boost a stellar Sequence A.

Jason Mendel, enterprise investor, Battery Ventures: Repeatability. Startups which might be fixing an actual ache level in a big market the place there may be clear urgency from a purchaser/consumer perspective must be well-positioned to boost a Sequence A in 2025.

Do you expect enterprises will improve their tech budgets for 2025? Will they lower them?

Aaron Jacobson, accomplice, NEA: Inside AI, we’ll see finances allotted away from “chatbots” to brokers. Enterprises will transfer past the low-hanging fruit of “GPT wrappers” to deploy digital employees that may purpose and take motion to make an actual enterprise influence.

Scott Beechuk, accomplice, Norwest Enterprise Companions: Tech budgets throughout many industries will improve in 2025, pushed by leaders’ need to attain two targets — which is able to generally be at odds with one another. The primary aim is consolidation. The second is growing top-line progress and enhancing operational effectivity, each of that are achievable with AI-based software program purposes. Patrons will buy startup options on this class regardless of their need to consolidate.

Kathleen Estreich, accomplice, Pear VC: In 2024, we anticipated to see extra enterprise adoption of AI. However that hasn’t panned out, primarily as a result of we haven’t but discovered use instances which might be tightly scoped sufficient and the instruments to cut back hallucinations and validate outputs haven’t gotten strong sufficient. In 2025 I anticipate to see extra enterprise adoption because the mannequin suppliers lengthen their stack upward. Each enterprise will want an AI tech technique. Should you don’t undertake, you received’t sustain. This can even create a variety of false alerts on the income facet for AI startups as experimental budgets might be excessive, however true product-market match might be more durable to see at first look.

Kirby Winfield, founding basic accomplice, Ascend: Enterprises will improve AI budgets in 2025. The query isn’t whether or not they’ll make investments however how they’ll sort out pricing, testing, and knowledge safety. Corporations like Salesforce and Smartsheet have already dedicated to AI adoption and can push more durable to leverage their knowledge belongings to remain aggressive.

Susan Liu, accomplice, Uncork Capital: In all probability the identical for the primary half, after which because the economic system improves and income/income enhance, we’ll see a rise in tech budgets within the second half.

Mike Hayes, managing director, Perception Companions: Primarily based on what I’m listening to from our enterprise companions, they’re more likely to marginally improve their tech budgets in 2025, with a deal with areas that ship measurable ROI and clear KPIs. I anticipate stress from boards and CXOs to place AI use instances into manufacturing to extend and obtain discretionary finances. I additionally anticipate continued enterprise funding in cybersecurity and cloud optimization. Mentioned otherwise, the fitting rising applied sciences shouldn’t have bother touchdown because of tech budgets.

Jason Mendel, enterprise investor, Battery Ventures: I’m optimistic about 2025 and anticipate to see firms improve their IT budgets with a powerful deal with rising applied sciences. Heading into the 2025 budgeting season, we at Battery Ventures polled 100 CXOs, collectively representing over $35 billion in annual expertise spend, and 74% of them anticipated to extend their expertise spend in 2025.

Will there be extra AI adoption?

Paul Drews, managing accomplice, Salesforce Ventures: Sure, basically all enterprise workflows might be optimized with AI — particularly agentic AI. We’re seeing actual demand for AI and ML instruments that may make underlying fashions 50% extra environment friendly whereas delivering improved outcomes. AI is experiencing froth, however from a bigger market perspective (not simply Silicon Valley), AI continues to be new and everyone seems to be attempting to determine easy methods to use it, worth it, and buy it.

Mark Rostick, vice chairman and senior managing director, Intel Capital: For the second, it’s clearly simpler to undertake AI via software distributors than attempting to construct your personal platform on condition that the marketplace for enterprise platform instruments continues to be very, very fragmented. I do assume there may be pent-up demand for some kind of platform answer, so I consider we’ll see many founders attempting to deal with that drawback this coming yr.

Raviraj Jain, accomplice, Lightspeed Enterprise Companions: It’s a consensus view however AI adoption will proceed to speed up in 2025 as (1) mannequin capabilities enhance, (2) enabling infrastructure is constructed out, and (3) stronger AI-first merchandise come to market.

What sorts of firms in your portfolio are seeing the strongest progress? Do you expect that can change in 2025?

Marell Evans, founder and basic accomplice, Distinctive Capital: Pressing ache factors for AI-ready prospects are producing shorter enterprise gross sales and procurement cycles and due to this fact quicker traction and scale. As we see AI adoption extra broadly, we may even see enterprises could have higher urge for food to strive not simply fixing for the pressing issues but in addition planning forward to keep up aggressive edge with “good to have” or extra future-forward and strategic options.

Kathleen Estreich, accomplice, Pear VC: We’re seeing nice traction in vertical brokers with a transparent understanding of the distinctive wants of their prospects. I feel vertical SaaS is a big alternative in 2025 to personal the end-to-end workflows with custom-built brokers for the duties to be accomplished.

Janelle Teng, vice chairman, Bessemer Enterprise Companions: A lot of Bessemer’s AI protection tech firms skilled large progress this yr. One in all our observations earlier within the yr is that the protection neighborhood shouldn’t be sitting idly by because the AI revolution sweeps the patron and business industries by storm. The [Department of Defense] mapped and released its formal AI adoption strategy last year, and we predicted that developments and purposes of ML might be embraced as important for the nationwide agenda and the protection neighborhood’s day-to-day work. This prediction proved prescient because the yr continued.

Mark Rostick, vice chairman and senior managing director, Intel Capital: One other sturdy section of the portfolio focuses on the infrastructure layer of software program and providers firms. Anyscale is a improbable instance. With their software program, builders can construct, run, and scale AI purposes immediately. There’s additionally RunPod, a digital cloud service supplier (CSP) for inference. It will possibly bridge the hole between {hardware} and software program stacks, which permits for seamless operation throughout numerous server suppliers, addressing a present problem within the AI area.

Ed Sim, founder and basic accomplice, Boldstart Ventures: No. This is likely one of the best platform shifts I’ve seen in 29 years of being a enterprise capitalist and IMO this can solely speed up.

What are your predictions for the exit surroundings subsequent yr?

Cathy Gao, accomplice, Sapphire Ventures: I predict M&A exercise will improve as giant firms search to accumulate AI experience. Strategic acquirers will deal with startups with domain-specific AI capabilities or excessive knowledge moats. The IPO market will stay cautious, however high-growth firms with profitability metrics would possibly take a look at the waters.

Nina Achadjian, accomplice, Index Ventures: I anticipate extra liquidity in 2025, each for M&As and the general public markets.

Aaron Jacobson, accomplice, NEA: With the change of administration, I anticipate the return of mega M&A offers. We’re going to see a multi-billion and even decacorn M&A final result for a number one AI firm.

Marell Evans, founder and basic accomplice, Distinctive Capital: We anticipate exits to select up barely subsequent yr, probably extra acquisitions and IPOs. Though, given the most recent fed assembly, exit quantity is perhaps slower than we anticipated.

Kirby Winfield, founding basic accomplice, Ascend: I predict new FTC management underneath the incoming administration will make hyperscalers extra acquisition-friendly for tech and expertise. However the IPO market will seemingly stay sluggish, given the frothy valuations some firms can command from the non-public market.

Andrew Ferguson, vice chairman, Databricks Ventures: 2025 could lastly be the yr that we see an uptick in tech M&A exercise, as extra favorable macro and (doubtlessly) much less onerous regulatory oversight make bigger firms much less skittish about M&A. Most strategic M&A might be targeted round wonderful technical founders and expertise, quite than on scaled enterprise, particularly ones that matured throughout the ZIRP period the place the expansion/profitability metrics should not pencil out for strategic acquirers. It’s doable that personal fairness or progress fairness traders make a play to consolidate that class of belongings into broader platforms.

Paul Drews, managing accomplice, Salesforce Ventures: The seemingly emphasis on authorities effectivity and decrease regulation will spur progress, investments, and exits. The general public markets are hovering, however there continues to be hesitation across the IPO course of from a non-public firm perspective. We’ve seen glimmers of hope within the IPO markets, which pre-IPO companies ought to take as a very good signal, however there may be nonetheless some disconnect between the final non-public valuation and the place the general public market will worth companies.

Mike Hayes, managing director, Perception Companions: I feel enterprises will look to strengthen their inorganic progress via acquisition extra in 2025 than in 2024. So far as the IPO market, I do assume that enterprises specializing in mission-critical options with predictable income could have alternatives in 2025. I’m optimistic and energized for 2025.