Nice to meet you.

Enter your email to receive our weekly G2 Tea newsletter with the hottest marketing news, trends, and expert opinions.

AI and Deep Learning: What’s Next in SaaS Innovation

December 5, 2023

deep-learning

The software-as-a-service (SaaS) sector has crafted a compelling narrative of growth and innovation in the past decade.

With its market value soaring to its peak in 2023, this thriving industry epitomizes the pinnacle of business evolution.

Driven by factors such as cost efficiency, scalability, and universal accessibility, SaaS products have permeated various sectors, fundamentally reshaping business operations.

How the AI hype impacts the SaaS landscape

SaaS organizations, often celebrated as a notable success in business innovation in the last decade, have experienced impressive expansion. 

The global adoption of cloud computing, the prevalence of mobile devices, and the growing range of SaaS solutions in different sectors have contributed to the widespread appeal of SaaS on a global scale. 

Small and medium-sized enterprises prefer SaaS due to its scalability and accessibility, while larger corporations seek to streamline operations and cut IT infrastructure costs. SaaS truly appeals to all.

In the midst of this ever-changing landscape, the excitement surrounding artificial intelligence (AI) is gaining greater prominence, leading technology and product leaders to strategically balance resource allocation while meeting innovation commitments. 

Within the sphere of SaaS, innovation isn't merely a routine practice—it’s a vital requirement.

A deep dive into SaaS innovation

Due to the nature of SaaS being highly adaptable, being at the forefront of innovation, and having many organizations implementing AI in different forms, Panintelligence took a deep dive into innovation priorities for SaaS and amalgamated this information into a report. 

The report includes how SaaS approaches AI technologies, how AI fits into their broader innovation and investment strategies, and the challenges the SaaS sector faces in responsibly and effectively leveraging this technology.

The report also explores the critical role of CPOs as strategic partners to boards in evaluating the best approaches for achieving innovation-driven growth.

Innovation priorities of top SaaS organizations

Panintelligence’s SaaS innovation report, which involved interviews and analysis with 54 top SaaS organizations, found that more than half (55%) said innovation is a major strategic priority. 

These companies dedicate regular board-level attention to product and operational innovation and pour significant resources into developing new products, features, and capabilities.

The research uncovered that the drive for innovation comes from two primary sources:

  •  Maximising SaaS product value: Nearly all SaaS leaders interviewed expressed that their innovation initiatives were geared towards enhancing customer satisfaction and loyalty, distinguishing their offerings, addressing the demand for new functionalities, and creating additional features for upselling opportunities. At least 90% of the surveyed SaaS leaders shared these objectives.
  • Improving the resilience of SaaS platforms: Security and data privacy remain significant priorities for most SaaS companies (91%). A minimum of 80% of the vendors we engaged with strive to enhance their platforms' performance and stability and streamline their internal operations.

The impact of short-term investment on SaaS innovation

The SaaS industry's innovation roadmaps are significantly influenced by short-term investment cycles. 

In an era marked by rapid technological advancements and shifting market dynamics, the pressure to deliver immediate ROI can divert focus from long-term innovation to short-term gains.

Here are key insights into why short-term investment cycles affect innovation roadmaps in the SaaS sector:

 Immediate ROI pressure

The expectation for quick returns in short-term investment cycles can lead SaaS companies to prioritize projects and features that promise immediate revenue, potentially sidelining transformative, long-term innovations. The necessity to satisfy investors and shareholders may relegate innovation to a lower priority.

Allocation challenges

Short-term cycles may favor allocating resources to incremental improvements and swift victories rather than dedicating them to more ambitious, long-term innovation projects. This allocation strategy could impede the development of groundbreaking technologies with the potential to reshape the industry.

Competitive urgency

The highly competitive nature of the SaaS landscape can compel companies to keep pace with or surpass competitors. This pressure may result in focusing on short-term feature additions or replicating competitors' offerings instead of exploring innovative, market-disrupting ideas.

Risk avoidance

Short-term investment cycles may foster a reluctance to take risks. Companies might hesitate to invest in unproven, innovative concepts carrying higher risks of failure, opting instead for safer, incremental improvements that are more likely to yield short-term results.

Customer expectations

Immediate solutions to customer pain points are often demanded. Short-term investment cycles can prompt SaaS providers to prioritize addressing these immediate needs over investing in longer-term innovations that may not offer immediate gratification.

Shareholder demands

Shareholders and investors can wield significant influence over SaaS companies, urging a focus on profit margins and short-term returns. This dynamic can create a misalignment between innovation roadmaps and the pursuit of groundbreaking technologies.

What’s hot and what’s next in SaaS innovation?

Panintelligence’s research shows that three areas dominate SaaS vendors’ technology investment and innovation plans:

Security and data privacy

Over 90% of the SaaS vendors in the report expressed ongoing concern about security and data privacy. Within the next six months, 85% of these vendors intend to bolster their security and data privacy credentials.

 With vendors' increased utilization of integrations and AI and growing regulatory scrutiny of these technologies, innovations will likely be required to meet progressively stringent data privacy, sovereignty, and security standards.

Artificial intelligence and machine learning

Artificial intelligence (AI) and machine learning (ML) are now mainstream in SaaS platforms, with vendors using these technologies to enhance their products and automate tasks.

Around 76% of vendors are now using, building, or testing AI in their products or back-office. More than half (56%) have made AI an immediate investment priority and plan to progress AI projects in the next six months.

Integration and APIs

About 74% of SaaS vendors have made this a priority innovation, and 57% plan to improve connectivity with third-party services and applications over the next six months.

SaaS vendors commonly broaden their product offerings through APIs, collaborating with partners that now often include specialists in AI.

A red and black bar graph showing SaaS innovation priorities.

Source: Panintelligence

The rise of AI in SaaS

With a substantial focus on AI in 2023, our research delved into the specific AI technologies and applications employed by SaaS organizations. 

Over the past year, there has been a noteworthy expansion in the adoption of AI technologies among SaaS vendors.

A whopping 75% of the SaaS vendors in our study are currently incorporating or developing AI/ML capabilities within their products or back-office functions. Additionally, 23% are exploring potential use cases. Only 2% of the consulted SaaS vendors reported having no intentions to integrate AI.

A red and black infographic highlighting the adoption of AI/ML among SaaS vendors.

Source: Panintelligence

What are the different types of AI?

  • Generative AI: Models that can create images, text, and other media by pattern matching from training data sets to create new data with similar characteristics. The engine requires prompts to produce the output.
  • Large language models (LLMs): Generative AI is powered by LLMs, which are very large ML models pre-trained on vast amounts of data. The LLMs are trained on trillions of words across many natural language tasks, often referred to as natural language processing.
  • Natural language processing (NLP): Models use syntactic and semantic analysis to break down human language into machine-readable chunks to be analyzed and processed.

Below are the most commonly used types of AI.

Machine learning (ML)

ML algorithms serve as the foundational framework for AI and represent the prevailing AI technology employed by contemporary SaaS vendors. These algorithms form the basis of numerous AI applications.

Currently, 43% of vendors have incorporated ML into their products, with an additional 15% integrating it into back-office operations. This integration aids them in deriving meaningful insights and identifying relationships within data.

For example, Adobe utilizes ML to discern optimal actions for maximizing sales, while Starling Bank employs it to bolster system security, detect fraud, and enhance behavioral analytics.

Generative AI

The hype around Generative AI models has turned into adoption this year. About 38% of the vendors we studied have rolled out Generative AI capable of generating text, images, or other media within their products, most of which launched in the last 12 months. 

DocuSign, for example, is using Generative AI to summarise critical components in agreements, Shopify has introduced a Generative AI that can analyze sales data and redesign websites, and Beamery has launched AI to generate tailored job descriptions and career recommendations.  Many more tools like these are being developed.

The rapid adoption of generative AI in SaaS is set to slow down as vendors realize software users are still not meeting all of their requirements despite this AI injection. G2 even expects some categories to drop AI features that don't provide a meaningful impact on businesses.

Natural language processing (NLP)

AI, which enables machines to understand and interact with human language, also featured strongly in SaaS plans. Around 21% of vendors have already introduced this to enhance their platforms.

They include Zoom, which uses this form of AI to extract and summarise essential information, such as next steps and highlights from meetings.

2024: The year of pragmatic AI

A study conducted by Workday, one of the largest SaaS organizations globally, reveals that 73% of business decision-makers feel compelled to boost adoption or investments in AI and ML. 

According to our research, nearly half (42%) of SaaS vendors are actively working on new AI product innovations slated for market release within the next 12 months. Below are some AI innovations focusing on pragmatic use cases.

Predictive analytics

Predictive analytics, which utilizes data models to anticipate future events, is gaining popularity. Certain SaaS vendors, such as Salesforce, have long been pioneers in employing predictive tools to enable users to dynamically respond to customer behavior.

A new wave of predictive analytics is emerging, exemplified by innovations like Paychex Retention Insights, which employs AI to identify potential employee resignations. 

Momentum in this realm of AI innovation is increasing, with 28% of SaaS vendors in our study currently testing predictive analytics and approximately half of that proportion already incorporating it. Additionally, 15% are exploring the application of Predictive Analytics in back-office operations.

Deep learning

Deep learning, an AI method that processes data in a way inspired by the human brain, is expected to move forward at a fast pace as we move into 2024.

The report indicates that 15% of SaaS vendors have already deployed deep learning technologies in their products. ElevenLabs, for instance, uses a proprietary deep learning model to turn writing into audio. CrowdStrike, which uses AI within its cybersecurity tools, talks of deep learning models achieving incredible performance in a variety of machine learning tasks.

With another 17% of SaaS vendors developing or testing new deep learning capabilities, the number of SaaS vendors using this technology could double next year. 

Although deep learning can be incredibly powerful, it could be impacted by new laws and regulations and by the difficulty of explaining its logic to regulators.

Causal AI

Causal AI will grow in prominence as a tool to help SaaS users understand the data accumulating in the platforms they use daily. It is also a way for SaaS vendors to address various risks they will encounter from their wider use of AI.

Causal AI goes beyond simple correlations to explore the causal relationships between different factors. It can provide new insight to help SaaS vendors and their customers with decision making and to identify and address issues such as potential bias within AI models.

15% of the SaaS vendors we studied have already introduced causal capabilities into their products or operations.

They include Adobe, which uses AI to identify the root causes of anomalies in customer data, and Palantir, which uses AI to perform causal analyses of failures in the oil and gas sector. About 6% of SaaS vendors are currently testing causal AI for product use and 8% for operational purposes.

We expect those numbers to increase sharply as SaaS vendors adjust to the need for explainable AI and policymakers move to legislate. 

Causal AI will be a valuable tool to help vendors answer questions from regulators and other stakeholders about decision making in their systems. What’s more, it can show how regulations influence outcomes, allowing policymakers to fine-tune regulatory frameworks for better results.

Overcoming the obstacles to AI integration

Our report uncovered the challenges that SaaS organizations perceived in the widespread adoption of AI. Here are the top five:

Regulatory and legal concerns

One of the hurdles hindering broader AI adoption in SaaS, as identified by more than half (52%) of the surveyed organizations, is regulatory compliance. 

They expressed concerns about ensuring that AI systems align with existing laws and regulations, with the current ambiguity surrounding future legal frameworks posing a significant barrier.

Security and privacy risks

Over a third (37%) of SaaS organizations perceive potential security vulnerabilities and privacy risks as obstacles to AI adoption. 

Concerns include the possibility of AI-generated code introducing undetected security risks and the risk of leaking trade secrets and sensitive data. A quarter (26%) consider this a major barrier.

Data quality and availability

The same proportion (37%) of organizations identify the lack of sufficient relevant and reliable data to inform AI models as a hindrance to adoption. 

About 19% view this as a major barrier, emphasizing that AI, reliant on high-quality data, is inherently flawed if built on inadequate foundations.

Potential reputational risk

A third (33%) of SaaS vendors highlight the potential for reputational harm and negative publicity as a deterrent to AI adoption. While most perceive this as a minor barrier, it remains a concern.

Transparency and explainability

The fifth barrier to AI adoption in SaaS is transparency, emphasizing organizations’ need to comprehend and articulate the logic behind a model's decision-making processes. Around 30% of organizations consider this a barrier, with many regarding it as minor. 

However, as new regulations are introduced and the industry increasingly employs deep learning and machine-generated 'black-box' models, the challenge of understanding such models may become more significant.

Neglecting data quality can bring dire consequences

The rise of AI in SaaS throughout 2023 has been striking. Despite the ongoing efforts of SaaS organizations to innovate and swiftly adapt to market shifts, the critical significance of data quality cannot be overstated. 

Neglecting data quality may inadvertently jeopardize the accuracy of AI models, potentially resulting in regulatory issues and unforeseen expenses for retrospective data cleansing.

As highlighted in this post, the perspectives shared by leading SaaS organizations globally offer a compelling exploration of the realm of SaaS innovation and the escalating significance of AI worldwide.

Learn more about natural language processing (NLP) and how it works.


Get this exclusive AI content editing guide.

By downloading this guide, you are also subscribing to the weekly G2 Tea newsletter to receive marketing news and trends. You can learn more about G2's privacy policy here.