Looking back at the AI Strategy Lab, I realise that developing an AI-enabled business solution involves so much more than just coming up with an innovative idea.
Ideation
Our project started with an idea during our brainstorm session from Emma: an AI-powered visualisation tool that could help customers design and furnish their rooms. I immediately liked the concept and contributed to further developing it into what eventually became Smart Space. It was motivating to see that our whole team was enthusiastic about the idea and independently started looking for relevant articles.
However during one of the lectures, a similar idea was mentioned as something that had already been explored. This initially worried me, as I thought we might have to change our concept entirely, after we already started. Fortunately, that was not necessary. Later, we also discovered that other groups were working on similar ideas. Although this was another surprise and bringing up some feelings of being unoriginal, it actually strengthened our group-motivation. We collectively wanted to develop a better and more thoroughly supported solution.
An advantage was that our concept closely connected to the technologies discussed in our lectures. This gave me a good understanding of which theories and articles could support our proposal. But I soon discovered that recognising a promising technological opportunity is very different from demonstrating its actual business value.
Contribution
My main contributions to our final project were further developing the problem and opportunity analysis, the business model and customer analysis, and conducting the future market test.
I deliberately combined the business model and the analysis of IKEA and its customers because I expected these sections to overlap. This allowed me to investigate IKEA’s existing and potential customers, its value proposition and how now called Smart Space could create value for both customers and IKEA. It also made our workflow more efficient, as I could develop connected sections without having to wait for input from others before being able to continue, saving time for the overall process.
An important turning point for me occurred while researching the problem and opportunity. In our initial proposal, we mainly focussed on customer choice overload as the problem they experienced when furnishing a room. However, during my literature research I realised that our initial explanation was too narrow.
Through my research it was discovered that choice overload was only one part of the problem. Time efficiency, uncertainty about preferences and product set complexity were also important considerations. Consequently, we adjusted our problem definition to better reflect these underlying challenges.
This was an important lesson. Without questioning our initial assumptions, we might have developed a solution that addressed only part of the problem.
Discussing these findings also demonstrated the importance of remaining open to changes. I appreciated that critical feedback was never treated negatively, but rather as an opportunity to improve our shared project.
Unexpected learning
Another valuable experience was testing our prototype. Before this project, I underestimated how much a relatively simple prototype could teach us.
Each of our team members conducted four tests, resulting in sixteen tests. Although our prototype covered the customer journey, individual features were not yet fully developed.
The feedback revealed several limitations that we had not sufficiently considered. For example, customers could modify recommended products, but the visualisations did not clearly reflect the specific changes. Test participants wanted to see the alternatives more accurately. They also suggested additional features and greater possibilities for personalisation for the rooms.
These findings made me realise that it is not enough to offer personalised recommendations. Customers also need to understand what they are choosing. Features that initially seemed like smaller details could therefore influence the overall experience considerably.
At the same time, I was slightly disappointed by some prototype limitations. These reduced the realism of our tests. Furthermore, because one teammate developed the prototype in a specific environment, the rest of us could not directly contribute to technical adjustments.
I had initially expected that current technology would make developing a sufficiently realistic prototype relatively straightforward. Looking back, I underestimated the time and coordination required.
Although the prototype helped us identify important customer preferences, I also realised that its limitations could have influenced the feedback we received. This made me more aware of the importance of prototype quality when interpreting test results.
For future projects, I would choose a collaborative development environment where the whole team can contribute and adjust. This would allow the whole team to make improvements.
Collaboration
Besides the analytical work, collaboration played a big role in our project. The combination of this assignment and the other subjects required planning and fair division of responsibilities.
We divided tasks individually while ensuring everyone remained informed about the complete project. Google Docs was particularly valuable here. By commenting on each other’s formulated sections, we created our own ongoing peer-review process. These comments developed into short discussions, allowing us to challenge reasoning and improve arguments without too many meetings.
I also contributed to keeping meetings focussed and time efficient. Discussions about planning and task allocation could take longer than necessary. Here, I tried to clarify necessary decisions to be made and move towards concrete next steps, saving time.
Teamwork requires transparency, constructive criticism, fair task division, and a shared responsibility, which I believe we achieved as a team.
Eye opener
One of the most interesting parts of my individual contribution was exploring Smart Space in a future market shaped by agentic AI.
At first, a future in which AI agents perform activities on behalf of customers seemed quite distant. While developing this scenario, I realised that an agentic future was more closely connected to our current concept than I initially expected.
Using the distinction between “Agents for X” and “X for Agents” from the lectures, I explored both sides of how IKEA could use agents internally and how Smart Space could become accessible to customers’ own futuristic AI agents.
It helped me realise that smaller technological developments may help a company remain competitive, but anticipating a still seemingly unrealistic future is also important when transforming businesses.
My takeaways
This project made me realise how many aspects are involved when developing an AI solution. It is not only about identifying an opportunity, but also about thoroughly understanding the problem, the stakeholders involved, how value is created for both customers and the company, and what potential risks or limitations there are.
Extensive and positive feedback is essential to keep improving while working within a team. Helping each other and making sure everyone can adjust in the systems you’re working with.
Overall, I learned that an AI concept is only the start of an effective business strategy. Research, experimentation, feedback and collaboration are critical for turning an idea into a meaningful solution.
