Ipsos Data Council Report
AI and automation Balancing excitement and fear to maximise potential
Ipsos Data Council Report
AI and automation Balancing excitement and fear to maximise potential
Ipsos Data Council Report
AI and automation Balancing excitement and fear to maximise potential
Context
While the potential for AI and automation is clear, many organisations have yet to effectively implement AI solutions at scale. One of the main reasons is that it’s incredibly complex - requiring not only the right technology but also the right people and processes. With many organisations still grappling with a skills gap, there simply aren’t enough people with the technical expertise to build, manage, and optimise AI systems.
On top of that, there’s the challenge of data quality. AI is only as good as the data it learns from, and many organisations are still struggling with poor data management, inconsistent data sets, and a lack of data governance. Without clean, reliable data, AI systems can’t function at their best. So, while the tools may be available, organisations are finding that the foundational elements – skills, strategy and data – need significant attention before AI can truly deliver its promise.
Scroll down to read how Ipsos Data Labs have solved this type of challenge for clients before. View Impact Story Developing an AI-enabled chatbot to improve team productivity
Key challenges from the Data Council
Council members shared that AI is on the radar for most organisations, but its implementation remains in the early stages - only a handful have clearly defined use cases or have successfully scaled AI initiatives across their operations. There is a general sense of excitement about what AI could achieve, but it’s tempered by caution. The complexity of integration, combined with uncertainty about the true return on investment (ROI), has made organisations hesitant to dive in too quickly.
In fact, while nearly all (9 in 10) interviewees confirmed that their organisation has made investments in AI and automation, the level of adoption and implementation varies widely. One challenge we heard is that many employees aren’t entirely comfortable with the adoption of AI tools with concerns ranging from the technology’s perceived complexity to fears about job displacement:
“I think there is an underlying fear in everybody around AI coming in and stealing their jobs.” Strategy, Operations & Supply Chain Director
Half of interviewees admitted that their organisation still isn’t doing enough with AI, with 4 in 10 believing the benefits of AI aren’t over-exaggerated. So, while there are some who are currently sceptical or cautious:
“We have an over-excitement of AI” Senior Director Global Business & Sales Operations, multinational technology company and “there are people that also don’t want to be first movers, so they are a bit more cautious about the AI strategy.” Senior Digital Product Manager
There are others keen to embrace AI’s transformative power:
“There are a significant minority of enthusiasts that are keen to test and trial AI tools.” Senior Digital Product Manager
This divided sentiment within organisations underscores the challenge of getting everyone on the same page when it comes to AI adoption.
Another key issue highlighted was trust:
Source: Ipsos Data Council,Oct 2024-Jan 2025 Base: 28 Ipsos Data Council members
“There’s a natural caution around what data gets shared (with AI) and what data we use and extract.” Direct to Consumer Lead, global domestic appliance company
Ipsos Data Labs learnings
AI has significant potential, but like any new big initiative, its successful implementation requires a clear, strategic approach. While generative AI, particularly using large language models (LLMs) for text analysis, is dominating the current AI hype, it’s important not to overlook the less glamorous, but equally important, areas of AI such as quantitative AI for numerical data analysis. This side of AI hasn’t yet reached the same level of maturity as its generative counterparts, but this shouldn’t deter organisations.
Instead, it’s a reminder that preparing robust data systems now can pave the way for future advancements and ensure readiness for what can be automated further down the line. For AI to work, organisations need to clean and structure their data properly. Without this foundational step, the promise of efficiency, automation, and insight will remain out of reach.
AI integration is complex, demanding a cautious, rational approach. Start small, with targeted, manageable projects to allow you to demonstrate the value of AI and its feasibility before scaling up.
Staying informed about the latest advancements in AI and being open to adapting your strategy is crucial to ensuring long-term value. This is also key given that as governments and regulatory bodies catch up with technological progress, evolving regulations could significantly impact AI adoption and implementation.
For AI to work, organisations need to clean and structure their data properly. Without this foundational step, the promise of efficiency, automation, and insight will remain out of reach.
Source: Ipsos Data Council,Oct 2024-Jan 2025 Base: 28 Ipsos Data Council members
Key takeaways
Prototype your approach
Focus on achievable goals and avoid getting swept up in the hype of full-scale AI transformation. Starting with smaller, targeted projects will help demonstrate value before committing to more significant initiatives.
Don’t rush in
AI needs to be handled with care, patience and informed foresight. The complexity of integrating AI cannot be underestimated, and organisations should take a measured, thoughtful approach that aligns with the culture and strategic goals of the organisation.
Flex your use cases
Continuously assess the relevance of your AI use cases in line with your business goals, staying responsive to changes in both the technology and regulatory environments.
Thought starters
What specific AI use cases could your organisation implement in the short-term to demonstrate tangible value, and how can you measure their success?
Which of your competitors or industry disruptors is doing a really good job of this?
Discover how Ipsos Data Labs have solved this type of challenge before
Impact story Developing an AI-enabled chatbot to improve team productivity
The challenge
Our client, a multinational consumer technology corporation, was looking for a solution which increased the productivity of the business sales operations (BSO) unit. The BSO team were faced with countless queries from field teams, account managers and other internal stakeholders on a wide range of topics - from compliance queries to invoice issues and they needed to know a little bit about a lot of things. It was therefore essential to find a solution which would help the BSO team answer queries more efficiently whilst also empowering other stakeholders to self-serve the information.
Our solution
The Ipsos team conducted an initial exploration to establish common query themes and regularly repeated questions. The team then collated this data into a central repository of knowledge, documenting queries, answers and processes which would ultimately feed into ‘George The Chatbot’. Hosted on Microsoft Teams, and leveraging technologies such as Power Automate, Power App Flows and Copilot Studio, the chatbot allowed stakeholders to self-serve the information they required, bypassing the BSO team and therefore freeing up their time to focus on other more valuable tasks. To ensure the chatbot was a forward-looking solution, the Ipsos team built it to be AI-enabled with a plan to implement large language models (LLM) or Azure AI studio in the future. As ‘George’ gets fed information from a live repository, it’s an ever-evolving solution which will continue to advance and gain increased internal adoption.
