Alexander Noll

How Learning Loops Became a Real Bottleneck in the AI Era

The core message of the session is that, for a scaling software company in the AI era, the bottleneck is increasingly not the ability to build software but the ability to learn quickly enough to decide what is worth building, doubling down on, or killing. After moderator Sharra introduced the masterclass format, Jannis from PwC framed the discussion as a founder-focused deep dive, noting his work with ventures from early startup stages through larger companies. Alexander, co-founder and CPO of plancraft, then positioned the talk around plancraft’s journey from startup to scale-up and the mistakes the founding team made along the way, with the goal of giving other founders practical awareness they could use immediately. Alex explained that plancraft helps trades businesses run their operations and make more money, covering the flow from quotations to invoicing and the work in between. He described the company as post-Series B, with €38 million raised the previous year, and said it had expanded beyond Germany into the Netherlands and was moving into Italy, Spain, Poland, and other European markets. Against that backdrop, he argued that learning has become the most important company-level capability. AI has accelerated the speed at which competitors and product capabilities can appear; he cited call-assistant products for trades businesses, where plancraft suddenly saw around ten companies in Germany building similar offerings. The first major operating lesson was to create sharper learning loops around strategic decisions: when to start something new, when to double down, and especially when to stop. Alex emphasized that plancraft made many mistakes by pursuing too many theoretically useful ideas at once and by lacking hard kill criteria. He advised founders to identify two or three things to kill in the coming weeks and to focus only on initiatives that truly move the needle. He connected this to AI agents by arguing that faster building makes judgment even more important: teams must know what is worth building, and engineers or designers hired primarily for craft may need to redefine their value as agents change day-to-day creation work. The later portion of the available transcript focused on organizational alignment. Alex warned that product, sales, customer success, and broader go-to-market teams often fight different battles unless leadership clearly defines non-negotiables. For plancraft, preparing for a future Series C created midterm goals that product and GTM needed to pursue together. He also suggested flipping the organization chart sideways to examine which teams are closest to customers, using that perspective to bring engineers nearer to customer reality. The provided transcript ends as he begins discussing high agency as a core cultural requirement: people should not wait passively for leadership decisions but proactively identify problems and move the company forward.