Catalyst: Here’s what you missed in London, Berlin, and Paris

Were you unable to attend Pigment’s events in London, Berlin, or Paris? No worries, we’ve got you covered. 

As well as presentations on Pigment’s vision and platform, we heard from a variety of business planning leaders - who shared stories and advice based on their successful careers. 

Here’s what you missed…

Berlin

Customer panel

Catalyst Berlin also featured a panel of Pigment customers, sharing their thoughts on change management, AI, and the results they’ve achieved with Pigment. We were joined by Arno R. Schleussner, VP of Finance at N26, Oliver Homann, Head of Controlling at Europcar, and Alexander Konrad, Vice President, Global Finance Technology & Data at BMG.

Change management and adoption strategy

‘Forcing’ adoption of a new tool can be a daunting move, but it can really pay off. The panel spoke about using key stakeholders, like the CTO, to champion the messages internally. Arno spoke about how N26 business partners from HR work directly in Pigment, rather than just using it to read reports.

Redesigning processes to leverage new technology effectively, rather than replicating old processes is critical. Small details like user experience, interface, and even color coding can significantly impact adoption.

And communication as key. Providing a system that makes it easy to find and produce information, with an accessible user interface, helps users see the change as less complicated. BMG used a creative, music-themed enablement video to explain their financial technology team's work and how Pigment helps them plan dynamically.

‍AI and the future

Arno from N26 is excited about AI taking over manual report compilation and commentary, allowing skilled people to focus on identifying drivers, discussing them, and developing actionable recommendations. AI could enable thousands of scenarios in parallel, shifting FP&A to focus on identifying relevant drivers and path dependencies.

But the panel agreed a strong data foundation is crucial before implementing AI. Alex from BMG flagged that for them, Pigment serves as this foundation, holding all their data and models, which can then be improved and used as reference datasets for AI. He sees potential for highly automated P&L and cash flow forecasts, machine learning approaches, and using AI analysts.

BARC: Why usability is key for resilience in planning

Robert Tischler, Senior Analyst at BARC gave an interesting perspective on why tool usability is actually one of the most important parts of watertight planning.

Key takeaways from his session:

Build in the business

Finance professionals, as problem owners, should be equipped to adapt software quickly without losing trust in the data. This involves self-service and decentralization of certain planning aspects.

Look for integrated solutions

Prioritize software that is easy to use and consistent across contexts to reduce onboarding time and increase the number of skilled users. Look for integrated planning and analytics solutions where planning, reporting, and dashboards are in one place.

Learn from your vendor

When acquiring new software, evaluate how models are built from scratch by vendors to understand every step. Start with experienced consultants who can guide future power users, transferring knowledge through on-the-job coaching rather than just classroom training.

Champion power users
Identify who your power users will be and give them a voice in software selection. Empower them to build, implement, and change models. This hands-on approach is more engaging and effective.

London

In London, we heard from a panel of two: Mikail, Director of Finance Platform at Bolt, and Erika, Finance System Lead at Vinted.

Staying agile

Bolt, as a scale-up, continuously refines its planning approaches. Their backbone is annual planning, followed by more strategic and operational levels. Strategically, they conduct five-year planning annually, modeling new business lines, expansions, cost optimization, and incentive optimization. Operationally, they perform quarterly forecasts, with business verticals and central teams adjusting to market realities every quarter.

Vinted has a similar need to stay reactive, mostly down to external factors such as the digital service tax, the uncertain geopolitical situation, and emerging technological shifts like AI. In response, Vinted is moving towards truly continuous and live planning, which involves faster data closing to enable quick decisions and the ability to input data and plans on a near-live basis. They also make liberal use of Pigment’s scenario planning capabilities, which enable them to prepare for a range of outcomes - not just one.

‍Tackling data and collaboration problems

Erica spoke about the importance of a single source of truth for analytics, actuals, and reporting, to reduce friction between teams. She explained that historically, with Google Sheets, it was nearly impossible to provide individual reports to 200-plus input providers for cost planning without many human agents. Pigment's access control allows them to provide visibility without compromising confidentiality, allowing finance business partners to focus on more important tasks.

Mikhail added that Bolt has over a thousand metrics in their Pigment metric tree, covering standard margin, OPEX, and headcount. Different layers of access management allow business verticals, cost center owners, and HR/people teams to work with their specific data while also seeing a holistic picture. 

He also pointed out that Pigment provides a platform for future consistency in metric definitions, which can be a point of contention between different teams (e.g., how "headcount" is defined).

AI strategy

Bolt maintains an AI working group composed of champions from various teams (customer support, engineering, operations) who investigate technologies and use cases. They hold bi-weekly meetings to share progress and findings. Bolt uses AI in engineering for code generation, checking, and testing, and in customer support for note-taking and information search.

Looking ahead, Mikhail envisions new professions and departments to train AI, adjust models, and prompt for Bolt's specific needs, as generic models are not sufficient. He also anticipates a need for "AI recruiters" to choose which AI components to use across the organization.

Erica quoted a former IBM CEO, who stated that while 10% of jobs might disappear, 100% will change. Just as Excel created new roles for Excel experts, AI will necessitate a shift in how people think about their roles.

Paris

In Paris, we were joined by Sophie Höhn, Global IBP/Demand & Performance Director at Danone.

Sitting between functions

If you’re a manufacturing head, your instinct might be to maximize my plant output — but that’s not always the best decision for the company overall. The real challenge in Sophie’s role is to weigh up the three KPIs of cash, cost, and revenue to make the best decisions for the business.

Imagine you have a sudden sales boost. That’s great, but capacity constraints might mean you can only deliver to half your countries. So how do you make allocation decisions? To make the right decision for the company, you need the financial valuation — margin per country, gaps vs. targets, and so on. All these decisions, while operational, must be assessed financially and planned collaboratively with finance to reach the best decision.

One of Sophie’s responsibilities is long-term forecasting (3-5 years) to determine plans for future factories and production lines. To calculate the expected volumes per plantDanone developed a Pigment application to translate macro-level topline and financial data into operational sourcing data — where future capacity issues may arise and where to invest.

AI adoption

Sophie makes liberal use of machine learning in her role, which involves consolidating sales forecasts and ensuring their accuracy. To improve forecast reliability in Europe, Danone has long used ML models.

But what really fascinates me is adapting these models to less mature markets. For example, in Indonesia, 60% of sales come not from modern retail but from small local shops — informal and unstructured. The data isn’t organized or even consistently collected. But there are models that can automatically clean and project forecasts using limited or imperfect data.

When Sophie managed the water business in Indonesia, external factors like weather had a huge impact. So she integrated external datasets like weather data into ML models to improve accuracy. It also allows local forecasters to focus their attention on the 20% of products where the model fails — since 80% of the forecast is now automated and reliable.

What next?

The entire Paris event is available to view on-demand here.

To sign up for Pigment’s virtual event, Prism, click here.