User Experience Insights
How SAP Analytics Cloud Enables an Agile Delivery Model
In this article I am sharing my team’s experience of the implementation of SAP Analytics Cloud and how it helped to improve Business-IT collaboration, making it agile.
First about myself – my name is Igor Kamenetskiy, I am Tech Lead of several planning products in adidas. It is in my responsibility to make strategic architecture decisions for my products, together with my team and Enterprise Architecture. And then – implement these decisions, achieving business value as a result.
Almost seven years ago we have automated Integrated Business Planning (IBP) process in adidas using SAP BW-IP in combination with Analysis for Office (please note that IBP here is has nothing to do with the SAP product Integrated Business Planning for Supply Chain, it is used in the meaning as defined by e.g. Oliver Wight).
Since then we had some projects to improve our solution and extend functionality. But we never changed the platform.
Now, following adidas Strategy-2025, the IBP process was re-thought and extended significantly and it required a complete IT solution re-build as well. After deep analysis, the comparison of different platforms, several POCs, talking to vendors and other clients, we decided to select SAP Analytics Cloud as our target platform.
After the go-live which is targeted for the end of the year, I plan to write another article about why we chose SAC and what results we could achieve with it. The focus of this blog post is on that project delivery process and how it was different from our previous experience.
Using SAP BW-IP as a planning engine requires heavy time investment in data modeling. Doesn’t matter if we want to run agile project, we must have a communication break in the beginning of it:
- First in a close collaboration with Business we define a target vision of future solution, ideally – with layouts and planning logic described.
- Then the IT team needs to sit alone and design the best fit data model for these requirements. And based on complexity and size of a project, this phase can take weeks or even months.
- Only after that, IT comes back to business and starts working together again, building and demonstrating layouts and front-end logic, e.g. in sprints. But at that time, the business teams (Product Owners) are restricted with what they can get from the data model developed earlier. Any significant logic changes that require data model rebuild can completely jeopardize project budget and timelines and usually are not acceptable.
This is not a true agile delivery, of course.
Here I want to especially highlight that our Business-IT collaboration is already as good as it can be: we act as one team, have one backlog and priorities list, etc. Hence the major factor blocking agile way of working was in the architecture setting, not in org. setup.
With an introduction of SAP Analytics Cloud, the situation changed drastically. Planning logic can be implemented now directly in the frontend tool. And this provides the unique possibility to “build-in” the logic in a layout in a simple and fast way.
So, this is how our implementation looked like now:
- In a beginning of a sprint team aligns on layouts ideas and logic using mockups in Excel and sketching tools (e.g., Figma).
- Immediately they start building layouts in SAP Analytics Cloud. If some logic is too complex and requires a separate longer implementation – it is “faked” and hardcoded. The main goal – as soon as possible get a tangible result that can be demonstrated to the end user.
- In the end of a very fast 2-weeks sprint, the team demonstrates developed stories in SAP Analytics Cloud to the product owner and key users. And what is very special – we immediately provide them with possibility to open these layouts themselves. Take their time, try the look-and-feel and the logic directly in a system.
This interactive collaboration, with using not just drawings, but real SAP Analytics Cloud stories boosted our communications dramatically! Now users were getting first hands experience and understanding of how planning will look like for them and could provide immediate feedback.
And we as implementation team could immediately react on this feedback and already in a next sprint rebuild the logic without significant spends on rebuilding backend data models!
With the help of such intensive interactions, we could find as minimum as two serious misunderstandings we had originally, that couldn’t be revealed from just sketches and user stories descriptions. Finding them at the early stage of a project made it possible to change the solution with minimum costs.
Another achievement from the new working mode was that with having real-life solution at hands, product owner and key users could realize that part of the planning process even had to be designed differently! No need to say how expensive it would have been if this realization had come in the end of a project only.
Our team is still on the journey of implementing its first SAP Analytics Cloud planning solution. But already now, before go-live, we achieved significant cost and (even more importantly) time savings because of a new architecture of SAP Analytics Cloud!
In case of interest in this topic, I plan to write another article after go-live, to share first user feedback, lessons learned and challenges we faced. Stay tuned and contact me if you want to chat about SAP Analytics Cloud for planning and share experiences.