“Not chatbots! We can all use chatbots, but tell us about real applications!”
I recently spoke with executives at a workshop about new trends in analytics, and generative AI (GenAI) sat at the top of our list, rightly so. However, participants told us that they wanted to understand how GenAI could be used to improve outcomes in their business— moving beyond the chatbots, which most BI vendors are tacking onto their products as an afterthought.
So, we had a conversation about real GenAI applications. Whether for sales, marketing, or finance, we looked at the key differentiators for AI leadership that impact operational applications and the software that runs business processes. It is more apparent than ever that the market is ready for more innovative, more insightful (and accurate) AI for business software.
So, suppose your business is building–or considering building–analytics in-house. In that case, it’s important to understand why and how embedding GenAI analytics/technology with best-in-class accuracy can give your organization an advantage in the race to impact.
While many business intelligence platforms allow people to embed reports, visualizations, and basic analytics, most AI add-on capabilities use LLMs alone. Those approaches yield answers of limited accuracy - perhaps only 60-70% - and mostly do not deliver on the promise of self-service, personalization, or insights for all users. Nevertheless, GenAI is powering a new generation of data-driven decision-making, and it’s more important than ever to understand what is actual versus what is just hype.
The challenge of rising expectations
With the acceleration of AI developments, user expectations have dramatically risen, and business users will not be satisfied with basic dashboards or static reports. They now expect:
Consumer-Grade AI and Analytics: AI-powered tools that are as intuitive and easy to use as the consumer apps they engage with in their personal lives. Why should their business experience be less impressive? Users want to interact with business data naturally and confidently without special skills and minimal training.
Enhanced Productivity: GenAI can, of course, be fun to use, but business users should also be able to streamline their workflows with AI features to help them work more effectively and efficiently.
Responsibility and Assurance: We have all heard of potential issues with AI, and we’ll talk about them in more detail later. Data privacy and ethical considerations are front and center. As a starting point, users know that their analytics tools should be not only powerful, but also responsible. Security, privacy, and ethics must be integral to the platform, not treated as afterthoughts.
These new expectations are undoubtedly challenging, but they also present us with an opportunity. If you can get it right, with an underlying technology that is robust, secure, and scalable, you can deliver an analytics experience that exceeds user expectations, driving deeper engagement with your application and enabling more impactful decision-making.
The irreplaceable role of human insight
For all that GenAI can achieve, there remain essential qualities that only we, as humans, can bring to the game. AI may process mind-boggling amounts of data for analysis, but only we can understand the full business context with the judgment required for truly strategic decision-making.
AI cannot fully understand the nuances of a business context–only humans can do that. Understanding how choices affect customers, employees, investors, and public perception – that is a uniquely human capability. Therefore, your GenAI-analytics platform must provide a path for humans to relay broader implications and apply industry-specific and cultural knowledge to help AI deliver accurate answers.
The same is true for ethics. Sure, we can train AI to follow some preset guidelines, and, as we shall see, we can build in controls for privacy, security, and bias management. But the subtleties of ethical judgment, particularly in complex situations, require our human touch. Conflicts of interest, balancing short-term gains with long-term sustainability, or achieving a sense of fairness may require insights we cannot find in the data alone.
Much of our human sense comes down to empathy. You know that in business, relationships matter when managing a team, interacting with customers, or negotiating with partners. AI can’t build the empathy or trust to sustain those relationships.
Finally, while AI can provide insights based on historical data and predictive analytics, strategic vision—the ability to navigate the long-term goals of a business–remains a human domain. Strategic vision involves not just analysis but imagination, ambition, creativity, and resilience: not qualities you will find in an AI model, however sophisticated. Your platform must have foundational ways to capture the irreplaceable roles of humans.
💡Clarify best practices for GenAI analytics at ThoughtSpot’s virtual event—join live on Oct. 30 or watch on demand.
Action-oriented analytics
I say this to every audience, but it’s worth repeating: Data without analysis is a wasted asset, but analysis without action is a wasted effort. Embedding analytics and insight into your business workflow aims to support immediate, informed decisions that improve your business.
Visualizing the basics of a business is very useful, but the ability to get immediate answers to complex queries is invaluable. Embedded analytics powered by GenAI enables users to ask questions in natural language and, in turn, receive near-instant insights in their own terms without learning a new query language. This capability alone can transform how your users interact with data.
However, the true value of embedded intelligence becomes apparent when users can do this without having to break or interrupt their workflow to open a separate analytic tool outside of their operational application. If they have to do that, then at that moment, you have lost their focus, and the value of your software to them diminishes.
Embedding analytics directly into business applications ensures that users can access insights within your application and act on them without switching contexts. You now own their entire operational workflow within the scope of your application.
Act-On’s rapid deployment of ThoughtSpot
The story of Act-On’s adoption of ThoughtSpot’s embedded analytics is a compelling example of how a forward-thinking company can dramatically transform its data strategy and its user’s workflow in this way.
Act-On, a leading cloud-based marketing automation platform based in Portland, Oregon, faced many of the challenges I described earlier. Early in 2023, their platform’s reporting tools were limited and inflexible, offering no advanced analytics or AI features. As a result, power users had to resort to exporting data to Excel to do further work. And that is never a good story for security and compliance.
An overhaul was needed from the ground up. So, in the first half of 2023, Act-On made a strategic choice to re-architect their data storage (using Snowflake) and to integrate ThoughtSpot Embedded for its cutting-edge AI capabilities, ease of use, and ability to integrate with their platform.
ThoughtSpot’s natural language search and intuitive interface enabled users across all levels of technical expertise to work directly with complex data to analyze marketing funnels, campaigns, and marketing spend.
The speed at which Act-On deployed ThoughtSpot’s embedded analytics is remarkable. Within 30 days, they were in beta with a select group of users to test and provide feedback. The feedback was overwhelmingly positive. Building on that success, Act-On expanded and developed the capability further and released it to general availability within 90 days.
The impact was immediate and compelling. Act-On not only delivered powerful new features but also introduced a three-tier pricing structure, which included upsell opportunities for advanced capabilities. What started as new features became a new revenue stream.
Act-On’s customers were no longer limited to static reports or external tools. Instead, they could interact with their data by using natural language for insights that were previously out of reach. They now had more than just analytics; they had a leading-edge, scalable, secure, and responsible AI system that customers could trust for ethical data management as much as for insight.
Why ThoughtSpot?
By embedding ThoughtSpot, Act-On was not just integrating a powerful tool—they were effectively onboarding an entire team of experts. ThoughtSpot’s ongoing research, development, and testing meant that Act-On could rely on a continuously improving platform backed by deep expertise in AI and analytics.
This partnership allowed Act-On to focus on its core business while benefiting from ThoughtSpot’s innovation and commitment to excellence. When organizations choose to integrate ThoughtSpot into their systems, they achieve far more than just adding a new tool—they bring a wealth of expertise, continuous innovation, and a comprehensive approach to embedded analytics into their development process.
ThoughtSpot’s platform has evolved from years of dedicated research, rigorous development, and extensive testing. So, companies like Act-On are also tapping into the collective knowledge and expertise of a team continuously pushing the boundaries of what’s possible in data analytics.
Doing it right
It’s easy to be excited about AI's possibilities—I hope you are. However, security, privacy, and ethical considerations are also critical. Indeed, the challenges of AI weigh so heavily on some organizations that they hesitate to make the best use of this latest, powerful capability.
ThoughtSpot’s platform is built with certain principles at its core, ensuring that every deployment meets the highest standards of trust and responsibility. For example, ThoughtSpot never shares your data with the GenAI model; it only shares metadata. This standard alone meets many of the concerns of security and privacy teams. In addition to these built-in safeguards that protect sensitive data, ThoughtSpot’s interactions with AI models ensure that insights are fair and unbiased, integrating responsible AI practices into its platform.
Another key reason to bring ThoughtSpot on board is simple but compelling. It’s difficult and expensive to hire AI talent today. Many developers can connect to a simple model and its APIs, but very few have experience in all these challenges we have discussed. ThoughtSpot brings all that experience into your development process.
Conclusion
As you can see, not all embedded solutions are created equal, and they differ even more when you bring GenAI into the picture.
The classic question of build or buy is still a lively conversation in embedded analytics. But just because you can build it doesn’t mean you should. To deliver an impactful embedded analytics experience, you have to offer more than the bare minimum.
