Despite AI being praised by some as the most significant technological advancement since the industrial era, businesses — potentially AI’s largest customer base — have been gradually integrating AI.

Although some investors anticipated that 2024 would mark increased AI adoption by businesses, that expectation was unmet as financial allocations remained limited and AI technologies often stayed within the “trial” phase.

Will things start shifting in 2025? It varies depending on who is asked.

TechCrunch conversed with 20 venture capitalists who support startups aiming to market to businesses about their forecasts for 2025. They shared their predictions on enterprise budgets, noteworthy trends to observe, and the requirements to secure a Series A in 2025, among various insights. Here’s their input.

Which business-related trends will capture your attention most in 2025?

SC Moatti, managing partner, Mighty Capital: My interest is geared toward this topic — AI integration relies on superior data. As companies transition from AI trials to extensive application, the necessity for top-tier data escalates.

Aaron Jacobson, partner, NEA: Code agents for updating application development are underappreciated. Expect increased AI utilization in transitioning mainframe apps to the cloud and enhancing older codebases.

Molly Alter, partner, Northzone: A major focus of mine is on sectors that were traditionally inaccessible by venture funds due to their business models requiring high COGS or OpEx. We’re witnessing AI automating numerous back-end tasks allowing sectors like accounting services, revenue cycle management, or premium legal services to achieve software-like profit margins.

Marell Evans, founder and general partner, Exceptional Capital: Grasping trends in business sales procedures — how long do certain companies test tools before finalizing decisions about in-house adoption? Additionally, recognizing various AI pricing models compared to traditional SaaS, consumption-oriented, and/or result-oriented models.

Mike Hayes, managing director, Insight Partners: An overlooked measure and a concept I believe will gain popularity in 2025 is TTFV, or time-to-first-value. I view this as an indicator for ease-of-implementation, suggesting that faster TTFV solutions may have a more significant edge going into the new year.

Which sectors are you aiming to invest in?

Liran Grinberg, co-founder and managing partner, Team8: Business resilience, whether facing operational failures or harmful insider or outsider threats. The CrowdStrike software update incident illustrated how vulnerable our digital environment is, not only due to cyber attackers but also simple errors.

Jonathan Lehr, co-founder and general partner, Work-Bench: Data sovereignty as a service. Companies are increasingly invest in data sovereignty solutions driven by regulatory demands and geopolitical factors. We are investigating startup prospects that empower companies to maintain full control over their data’s location, storage, processing, and governance while ensuring adherence to local regulations.

Mark Rostick, vice president and senior managing director, Intel Capital: We are interested in companies that specialize in task-specific models. While foundational models are well-established, I find models that excel in specific functions particularly engaging, especially when coupled with agents built atop them. In addition, we are carefully watching the evolution of transformer alternatives and any potential solutions to reduce the vast computational power necessary to train LLMs and employ them in production.

Mike Hayes, managing director, Insight Partners: Traditionally, technology was considered by businesses as either a revenue driver or cost reducer, but this perception is rapidly shifting in favor of technology enhancing enterprise value while concurrently minimizing business friction. I seek out solutions that address distinct, unconventional challenges for businesses — areas where traditional solutions have been inadequate; this encompasses vertical or persona-specific workflows reimagined using GenAI or agentic automation and security advancements that not only identify and warn, but also resolve issues.

Jason Mendel, venture investor, Battery Ventures: Several compelling areas where I believe AI has the potential to significantly add value, and which intrigues me, include observability / incident response, IT service management, demand creation and sales engagement, offensive security, software development, and the SOC workflow.

Ed Sim, founder and general partner, Boldstart Ventures: We approach with consideration of secondary effects. If, in the upcoming two to three years, we anticipate living in a world where each of us is assisted by dozens or hundreds of agents, we need to deliberate on the entire infrastructure required to support these new digital assistants. Who will offer the security infrastructure for access control? Who will manage these agents? Is a platform available to manage distinct agents and secure them? What about a runtime system for Claude’s MCP, which resembles a dockerized, secure sandbox for agents to execute tasks?

Are there technologies, sectors, companies, etc., capturing your interest that aren’t AI?

Liran Grinberg, co-founder and managing partner, Team8: Quantum computing remains exciting. Furthermore, the cybersecurity field continues to be critical, as threats are now leveraging AI with increasing frequency.

Safeguarding our online systems entails complexity.

Nina Achadjian, partner, Index Ventures: There’s been a revival in financial technology, Software as a Service, and online shopping sectors, which were popular but experienced a deceleration in recent years. Furthermore, we anticipate cybersecurity and gaming fields to maintain their interest this annum, with an increased pace in cybersecurity as the initial public offering market expands and security-related regulations along with disclosure mandates grow.

Aaron Jacobson, partner, NEA: There is substantial excitement regarding AI security, yet the larger prospect lies in assisting organizations to apply “Cybersecurity 101” broadly in a manner that doesn’t obstruct user effectiveness. Significant areas of focus involve reinforcing minimal privilege access, sustaining secure data postures, and preventing ransomware attacks. I’m also eager to back technologies facilitating cross-cloud deployments for businesses.

Molly Alter, partner, Northzone: I am truly enthusiastic about enterprises addressing public sector challenges. The government contracting economic landscape is flourishing; aggregate federal agency contracts amounted to $774 billion in 2023. Adoption of technology and modernization are crucial for driving effectiveness that the new leadership is devoted to, and a growing network of firms is tackling this directly.

Andrew Ferguson, vice president, Databricks Ventures: We invest considerable time with our system integrator partner network. These organizations are undertaking the challenging task of assisting enterprises in transforming their data and AI strategies into practical applications.

Janelle Teng, vice president, Bessemer Venture Partners: We are progressing beyond the modern data stack. The data infrastructure scene is experiencing a considerable shift, propelled by diverse factors, such as the emergence of lakehouse architectures and convergence towards specific open table format protocols.

Raviraj Jain, partner, Lightspeed Venture Partners: Energy presents a major investment sector, considering the surging demand for energy to power data centers and the nationwide grid failure challenges. There will be sustained interest in nuclear energy — both fusion and fission.

Regarding AI, how do you ascertain that a firm possesses a competitive edge?

Cathy Gao, partner, Sapphire Ventures: I evaluate it using a “5D framework”: design, data, domain knowledge, distribution, and dynamism. Since the onset of this year, we’ve employed this framework at Sapphire to assess firms constructing AI-driven applications.

SC Moatti, managing partner, Mighty Capital: An AI competitive edge is founded on exclusive data, state-of-the-art algorithms, and scalable frameworks, enabling distinctive and superior solutions.

Scott Beechuk, partner, Norwest Venture Partners: The most profound competitive edges will be formed by extensive proprietary datasets. The firms with the highest long-term potential are those constructing their unique datasets to thrive in their distinct, vertical channels — typically via training or refining their custom models.

Jonathan Lehr, co-founder and general partner, Work-Bench: As a dedicated seed fund, we allocate most of our focus on vertical AI prospects targeting business-specific operations that demand profound industry knowledge, where AI chiefly enables access to previously unattainable (or extremely costly to obtain) data and refining it in a manner that would have demanded hundreds or thousands of working hours.

Raviraj Jain, partner, Lightspeed Venture Partners: The inquiry to pose is, As models advance, does this entity face a threat or gain strength?

What is required to secure a Series A as an enterprise startup in 2025?

Liran Grinberg, co-founder and managing partner, Team8: With a notable founder-market alignment and a bold vision for building a substantial company, one can successfully raise between $15 [million to] $25 million in a Series A round with just a few hundred thousand dollars in ARR.

Molly Alter, partner, Northzone: Successful Series A enterprise launches will demonstrate robust top-line growth (>100% YoY) paired with minimal burn multiples; the era of 2021 with growth at any expense is over. Crucially, these ventures will display a clear, long-term differentiation plan that will distinctly separate them from the multitude of other offerings striving to attract investors and engage the same enterprise client base.

Kirby Winfield, founding general partner, Ascend: Progress from zero to $1 million swiftly, within two quarters, supported by an exceptional team in a vast market offering a distinctive solution that has generated considerable demand.

Andrew Ferguson, vice president, Databricks Ventures: If you’re creating an AI-centered product, having a stellar technical team and early product market traction ($2 [million to] $5 million ARR) might be the Series A expectation. The journey from product launch to $5 million ARR is significantly faster in the AI age compared to the traditional SaaS times. I predict that the Series B benchmark will be considerably higher — and it remains uncertain if this initial ARR is of high quality and sustainability.

Jonathan Lehr, co-founder and general partner, Work-Bench: We’re receiving insights from downstream peers that the threshold is around $1.5 million with the potential to multiply threefold sequentially, in order to secure an outstanding Series A.

Jason Mendel, venture investor, Battery Ventures: Predictability. Startups addressing a genuine pain point in an extensive market where there is a clear urgency from a buyer/user viewpoint should be well-positioned to raise a Series A in 2025.

Will enterprises expand their technology budgets for 2025, or will they shrink them?

Aaron Jacobson, partner, NEA: In the realm of AI, funds will shift from “chatbots” to agents. Businesses will progress beyond the basic “GPT wrappers” to implement digital workers capable of reasoning and performing tasks to yield a meaningful impact on business operations.

Scott Beechuk, associate, Norwest Venture Partners: The tech spending across several sectors is anticipated to rise in 2025, motivated by executives’ aspirations to meet two objectives — which may occasionally clash. The initial objective is consolidation. The other is to elevate top-line growth while enhancing operational effectiveness, achievable through AI-driven software solutions. Despite their consolidation goals, buyers will invest in startup solutions within this field.

Kathleen Estreich, associate, Pear VC: In the year 2024, there was an anticipation of higher enterprise AI adoption. However, it hasn’t materialized, mainly because properly defined use cases haven’t been established, and tools to mitigate hallucinations and verify outputs lack the necessary strength. I anticipate broader enterprise adoption in 2025 as model providers elevate their stack. An AI tech plan will be crucial for all businesses; failing to adopt may result in falling behind. This will generate numerous misleading signals on AI startup revenues since experimental budgets may soar, yet finding a true product-market match might prove initially challenging.

Kirby Winfield, initial general partner, Ascend: In 2025, organizations are set to augment their AI funding. The discussion isn’t centered around if they’ll invest but rather how they’ll approach pricing, trials, and data safeguarding. Entities like Salesforce and Smartsheet have already shown commitment to AI adoption and are poised to intensify their efforts in utilizing their data assets to sustain competitiveness.

Susan Liu, associate, Uncork Capital: The situation will likely be consistent in the first section of the year, but as the economy and financial outcomes improve, a boost in tech expenditures should happen in the latter portion.

Mike Hayes, managing director, Insight Partners: Insights from our corporate partners suggest a slight boost in tech budgets in 2025, targeting domains that provide quantifiable ROI and definitive KPIs. I foresee mounting board and CXO pressure to operationalize AI use cases, thus securing discretionary finance. Additionally, enterprise investments in cybersecurity and cloud enhancements seem on course to continue. To put it differently, suitable emerging technologies shouldn’t face barriers in obtaining financial investment due to tech budgets.

Jason Mendel, venture investor, Battery Ventures: I’m positive about 2025 and foresee firms enhancing their IT budgets with a marked emphasis on emerging technologies. As the budgeting cycle for 2025 approaches, our Battery Ventures study involving 100 CXOs, who collectively manage technology expenditures exceeding $35 billion annually, revealed that 74% anticipate an increase in technology spending in 2025.

Is greater AI adoption on the horizon?

Paul Drews, principal partner, Salesforce Ventures: Indeed, practically all business processes can be refined using AI — particularly agentic AI. There’s a tangible demand for AI and ML solutions that can enhance underlying models by 50% in efficiency while providing superior outcomes. Although AI facing volatility, on a broader scope (beyond Silicon Valley), remains in its infancy, with everyone exploring ways to utilize, price, and acquire it.

Mark Rostick, VP and senior principal director, Intel Capital: Currently, opting for AI via application vendors is noticeably more straightforward compared to developing an individual platform, given the substantial fragmentation in the enterprise platform tools market. I perceive a pent-up requirement for some platform solution, leading to numerous innovators attempting to resolve this challenge in the upcoming year.

Raviraj Jain, associate, Lightspeed Venture Partners: It’s widely agreed that AI acceptance will continue to gather speed in 2025 as (1) model abilities elevate, (2) foundational infrastructure expands, and (3) more robust AI-first offerings emerge on the scene.

Which businesses in your portfolio are currently experiencing the most robust expansion? Do you foresee any changes in 2025?

Marell Evans, originator and chief partner, Exceptional Capital: Pressing challenges for AI-ready clients are resulting in quicker sales and procurement cycles within companies, thereby accelerating momentum and scaling. As AI acceptance becomes more widespread, corporations might display a greater willingness to explore solutions beyond urgent issues, planning ahead with “nice to have” options or pursuing more futuristic strategies.

Kathleen Estreich, associate, Pear VC: Significant momentum is observed in specialized agents comprehending their clients’ distinctive requirements. In my view, vertical SaaS represents a substantial prospect in 2025 to manage complete workflows by creating specialized agents for necessary tasks.

Janelle Teng, VP, Bessemer Venture Partners: A number of Bessemer’s AI defense tech enterprises recorded remarkable growth this year. Earlier in the year, our insight was that the defense sector wasn’t passively witnessing the AI revolution transforming consumer and commercial spaces. The [Department of Defense] crafted its formal AI acceptance roadmap last year, and we anticipated that evolutions in ML applications would be integrated as vital elements for the national agenda and routine defense operations. This forecast turned out to be accurate as the year evolved.

Mark Rostick, VP and senior principal director, Intel Capital: Another potent segment in the portfolio zeroes in on the foundational layer of software and service companies. Anyscale epitomizes this. With their solutions, developers are empowered to craft, execute, and escalate AI applications promptly. Additionally, there’s RunPod, a virtual cloud service provider (CSP)for deduction. It can connect the distance between hardware and software tiers, facilitating smooth functionality across diverse server suppliers, addressing an ongoing issue in the AI domain.

Ed Sim, originator and primary partner, Boldstart Ventures: No. This is one of the most significant platform transformations I’ve observed in nearly three decades as a venture financier and IMO this will only gain speed.

What forecasts do you have for the exit circumstances next year?

Cathy Gao, collaborator, Sapphire Ventures: I foresee M&A endeavors will rise as major corporations aim to procure AI proficiency. Strategic acquirers will target startups with specialized AI capabilities or substantial data advantages. Caution will prevail in the IPO sphere, but enterprises showing strong growth and profitability metrics might venture into it.

Nina Achadjian, collaborator, Index Ventures: I expect increased liquidity in 2025, for both M&As and the public markets.

Aaron Jacobson, partner, NEA: With the new administration in place, I foresee mega M&A transactions resurging. We are poised to witness multi-billion and even decacorn outcomes in M&A for prominent AI enterprises.

Marell Evans, founder and primary partner, Exceptional Capital: We anticipate a slight increase in exits next year, possibly through more mergers and IPOs. However, the recent federal meeting suggests that exit volume might not rise as swiftly as anticipated.

Kirby Winfield, originating general partner, Ascend: I project that new FTC leadership under the forthcoming administration will make large-scale acquirers more receptive to acquiring tech and talent. Yet, the IPO market will likely stay slow due to the high valuations that some firms can achieve in private transactions.

Andrew Ferguson, vice president, Databricks Ventures: 2025 could be the year we finally witness an increase in tech M&A pursuits, as more favorable macro conditions and potentially less stringent regulations may make bigger companies less hesitant about M&A. Most strategic M&A will zero in on exceptional technical founders and innovations, rather than on extensive firms, especially those that grew during the ZIRP era where their growth/profit ratios might not be appealing to strategic acquirers. There’s a possibility that private or growth equity investors might attempt to integrate that category of assets into wider platforms.

Paul Drews, managing partner, Salesforce Ventures: The probable focus on government efficacy and reduced regulation will catalyze development, investments, and exits. Although public markets are thriving, there remains reluctance concerning the IPO process from a viewpoint of private enterprises. We’ve noticed hints of promise in the IPO markets, which pre-IPO firms should view as encouraging, but a gap persists between prior private valuations and expected public market pricing.

Mike Hayes, managing director, Insight Partners: I predict that in 2025, companies will aim to fortify their non-organic expansion through acquisitions more than in 2024. Regarding the IPO market, I do believe that corporations focusing on essential solutions with stable revenue will have chances in 2025. I am hopeful and invigorated for 2025.