AI seemed to be the solution for every problem in the midst of the excitement cycle, but a recent report from the Capgemini Research Institute shows C-suite executives have moved beyond the hype to a more realistic and pragmatic approach on how to move forward.
Organizations and their leaders are becoming more deliberate about governance, skills, accountability, and human–AI chemistry.
There is greater emphasis on return on investment, as early adoption is delivering value. More than half of CxOs report reduced time and cost to make decisions, and improvements in creativity and foresight. But AI remains largely limited to input, as just one percent of CxOs believe AI could autonomously make certain strategic decisions in the next one to three years.
Venture capitalist mindset
Speed-to-value is becoming imperative. To move proofs of concept out of the experimental cycle and into productive use, the focus needs to be on delivering results while maintaining trust.
The key is to have a succinct set of defined success criteria, so KPIs can show incremental performance improvements. This requires a venture capitalist mindset. Executives need to reinvest the savings so the value from PoCs can be used to expand and scale out into larger, more production-ready solutions.
For example, Capgemini helped an automotive company answer some key marketing questions: How can we increase levels of loyalty or retention, and how do we speed up the sales conversion cycle through a better understanding of customers and hyper-personalization in our direct marketing?
The small PoC pilot was competitive with a variety of agencies and technology providers. It created more agile ways of working, using the advanced analytical capabilities of the SAP platform against the defined use case. Once that value was proven and accepted, the chief marketing officer reinvested the money and scaled it out.
Proven use cases
With more than 300 AI and agentic use cases already defined, Capgemini has a solid industrialized library to work from. This facilitates demos and scenarios that can lead to imaginative ideas about how AI can impact the business.
The most important question for organizations to answer is: what drives our customers from a value perspective? Use cases also need to include investment, ROI, and reinvestment numbers. Since AI is a business investment, parameters need to be set for every dollar invested. The companies that do so will secure the funds required to invest in future use cases.
Moving beyond data scientists
In the SAP universe, context matters when it comes to data. But context doesn’t come from table-level data – it comes from the application level. It is a very hard investment to move SAP data from a pure ERP system into another platform. The effort to remodel and rebuild all the contextual information is tremendous. It takes years to rebuild properly.
Now, SAP with Business Data Cloud and the new connectors to Snowflake, Databricks, and others bring significant benefits to enterprises, allowing data to be removed with the context intact so it can be used for predictions or outside elements in Snowflake and Databricks. Organizations can also use the outcome of a prediction model and bring it back into SAP processes to actively trigger a dual agent to create a maintenance order or a new supplier registration.
This opens multiple use cases that do not depend on taking data linearly into another data warehouse’s data fabric for analysis. Now, data-driven augmentation triggers a process on the audit-to-cash side or on the hire-to-retire side. In the case of an end-to-end loop, the free flow of data, the utilization of data, and the context are really important. There is a real benefit to having the SAP data products as well as building in new and unique ones.
The entire process relies on having business- and AI-ready data that is orchestrated, harmonized, and quality-assured over multiple systems. With less time spent on data modeling and analysis, enterprises can speed the process and increase ROI – resulting in a greater focus on driving business outcomes.
It is always beneficial to work with data as close as possible to its source. Transferring data between platforms introduces complexity, points of failure, and inaccuracies, ultimately eroding trust.
Focus on value creation
Building a solid architectural strategy that focuses on incremental value is the north star for AI-enabled organizations. Information needs to flow and be accessible. If a data lake is just a swamp of data sitting somewhere, it has little to no value. The one constant across enterprises that win in the AI era will be those that first understand this value and build a data-first culture.
Companies operate in a multi-cloud, multi-hyperscaler, and multi-solution world. This requires understanding the openness of the platform and architectures that keep getting more complex. The ways of working need to pivot toward new scenarios.
The answer harnesses the strengths of existing technology platforms to allow people to focus on the value creation aspects of their roles.
Join our virtual session at Sapphire Orlando 2026 to continue the discussion on how to turn use cases into business impact.
