You don’t need a crystal ball to foresee that 2025 will be another big year for GenAI. In just a few short years the technology has proliferated as it’s become increasingly useful and accurate. Users are growing comfortable with GenAI tools and business leaders are embracing the potential of the technology across a range of critical areas.
But what specifically should enterprise buyers be looking for in GenAI in 2025? And what should you keep in mind, especially when it comes to research and insights?
Read on to discover what we think will continue in 2025, our three predictions for 2025, and the key buyer considerations that won’t change.
What will continue in 2025
One thing we can count on: we will continue to see GenAI aid creativity and save time. For instance, GenAI’s ability to generate content and automate repetitive tasks will increasingly be used to streamline the research process.
There are many ways enterprise researchers are already taking advantage of the capabilities of this powerful technology (or should be):
Reducing the time it takes to sift through existing research. GenAI has made it so much quicker and easier to locate and synthesize findings. Whereas researchers used to have to spend untold hours (a) looking for research and (b) searching through it for relevant data or ideas, now GenAI can perform these tasks for research professionals in a fraction of the time—allowing more time for analysis and insights.
Ideating multiple answers to a question or solutions to a problem. GenAI tools will continue to be an enormous aid to the brainstorming process. Researchers can generate innumerable hypotheses or possible solutions to a problem, and then evaluate the likelihood of the success of an idea—say, a new product or service—based on previous research. The quality of the solutions will only improve, especially, as in the case of Stravito Assistant, if the GenAI draws on institutional knowledge and trusted 3rd party datasets.
Finding further questions to investigate. Increasingly can help narrow down research questions or find further relevant topics to investigate—it’s a great way of looking into unknowns. We see users increasingly turning to GenAI to prompt discovery—to find new ideas and opportunities, and, as above, compare them in terms of viability.
Turning insights into stories and formats that travel. Most of the time AI is leveraged to synthesize or summarize research, but it can also be effective in how research and insights are shared. Research teams can turn data into stories and formats that are not only more digestible, but more likely to be remembered, recalled, and applied. The ability to do this quickly and at scale will inevitably bring greater visibility of both research and the teams behind it.
Three predictions for 2025
All of the above is already possible, especially with a knowledge management system like Stravito and AI-powered Assistant. Here’s what else we expect.
Proactive knowledge management. GenAI tools are great for extracting insights from large libraries of unstructured data and using them to drive innovation. But in general this work relies on what’s already happened. We see huge potential for GenAI to proactively cross-reference research to identify trends and threats—and to alert decision-makers in advance. Look for GenAI to become adept at spotting new trends, or shifting market shares, or new competitors, or old competitors launching new projects. This could be one of the most exciting potential developments in 2025 and beyond.
Improved language capabilities. GenAI tools provide an easy way to translate texts, but the subtleties still need work. Currently, it’s not advised to release a translation without an actual human checking it for culturally insensitive phrasing or even outright mistranslations. But we expect to see vast improvements in translation capabilities as the businesses behind the tools continue to train Large Language Models. And as the capabilities improve, they will be a boon for large, multilingual enterprises—especially as it will greatly reduce inaccurate translations of research and insights and make existing research more accessible.
Rise of the niche GenAI agent. Some GenAI tools are better for generating text, some for imagery, whilst others serve as ‘agents’ that perform specific tasks like responding to customer service enquiries. As businesses seek to unlock efficiencies through automation and users find the limitations of generalist options, we expect to see a rise in the number of ‘agents’ designed for specific use cases. These will require specialized capabilities or a combination thereof so as the barriers to building these tools become less, we anticipate seeing rise in experimentation and application of GenAI more broadly.
Key buyer considerations that won’t change
We will see more users and more integration of GenAI tools into our everyday work lives. But bear in mind it’s still a new technology, and like all new technologies, there are caveats, especially for large multinational enterprises.
Expect more hype. Not all of the claims and promises of GenAI are relevant or applicable; enterprise buyers should be discerning in evaluating the products. There should always be a solid business case for using these tools and a concrete means to measure the impact.
Does it integrate with your existing tech stack? Microsoft already allows for Copilot to be used across its applications; we expect users will turn to it with even more frequency. You should take a careful look at integration with your current tech stacks and existing research partners if you expect to use GenAI across your entire knowledge libraries.
The risks haven’t gone away. There are still ethical concerns about GenAI tools, particularly in the areas of bias baked into LLMs and privacy. Security issues may even grow, as malefactors find ways to manipulate the responses of chat tools or exploit security lapses. Users should always be made aware of the risks, and enterprises should have a clear AI usage policy. Buyers should engage their legal and procurement teams early to ensure any tool is evaluated properly and does not delay any planned implementation.
Humans still matter. There seems to be a growing temptation to think of GenAI tools as somehow replacing humans. And while it is already clear that GenAI reduces the need for workers to perform certain repetitive tasks, people are still at the heart of it, especially when it comes to knowledge management and the research process. We still need people to write the prompts, evaluate the research, and assess the solutions for their viability. GenAI is designed to mimic human reasoning. But it’s not a replacement for it.
One way to think of the human element is change management. Enterprise buyers have to consider the best way to implement GenAI across a large organization—who the users will be, how to train them, where to implement it first, and how to communicate the advantages of new ways of working. In short, it’s the same challenges as before, simply with another new technology.
Conclusion
We see GenAI growing as a strategic tool for augmenting and streamlining many enterprise processes, and as such it can have a big impact on the bottom line.
When it comes to research and insights however, the goal should be to empower human users—that is, researchers and consumer insights teams—in creating a more comprehensive and well-rounded understanding, leading to better insights and better business outcomes.
AI-powered knowledge management solutions such as Stravito will increasingly help researchers improve their efficiency and insights. For a real-world example of how Stravito, and our AI-powered Assistant, is helping to drive both efficiency and insights discovery, have a look at our customer success story: How HEINEKEN is Amplifying Insights with Generative AI.
