How Viking Growth uses AI to identify relevant signals and prioritize the most attractive companies at the right time.
Most sales organizations have an extensive understanding of which companies fit their ideal customer profile. The challenge is rarely identifying potential prospects but rather determining which companies should be prioritized and when the timing is right to reach out.
Many organizations still rely on standardized workflows and predefined follow-up intervals. While these processes provide structure and consistency, they offer limited insight into whether a prospect is receptive to a conversation at a given moment.
At the same time, companies are constantly evolving. They hire new employees, launch new initiatives, expand into new markets, undergo organizational changes, or raise new funding. These events can serve as early indicators of emerging needs.
Vilde Kjos, Investment Associate Viking Growth
The challenge is that this information is often fragmented and scattered across news sources, company websites, job postings, industry databases, and public records. Even when signals are available, systematically monitoring them at scale is difficult.
This challenge became the starting point for an internal project at Viking Growth. The goal was to explore how artificial intelligence could be used to identify and combine signals that indicate when a company should be prioritized for further engagement.
Earlier this year, Viking Growth reviewed its investment process to identify bottlenecks and activities that were taking up unnecessary time and resources. Through interviews with the investment team,one challenge surfaced repeatedly: systematically following companies over time.
Viking Growth tracks several thousand software companies across the Nordics. Many of these represent potential investment opportunities, but only a small number are suitable at any given point in time.
The challenge closely mirrors traditional sales. Both investors and sales teams operate within a large universe of relevant companies, yet have limited visibility into which ones to prioritize right now.
According to Investment Associate Vilde Kjos, outreach efforts were often based on when a company had last been contacted, rather than on concrete signals indicating that the timing was right.
“We lacked strong indicators for when to engage with companies. As a result, follow-up efforts were often driven by the timing of our last interaction rather than by signals suggesting that a company was ready for a conversation.”
Many sales organizations measure and optimize activity through metrics such as emails sent, calls made, touchpoints completed, and follow-up sequences executed. While these metrics are important for maintaining momentum and discipline, they provide limited insight into whether the timing of outreach is optimal.
A company may be an ideal customer without being ready to enter a buying process. Another company may be a weaker strategic fit but have a more immediate need for the solution. If both are treated the same way in the CRM system, it becomes difficult to allocate time and resources where the probability of response and progress is highest.
Vilde points out that the challenge is not understanding which signals to look for but rather having the capacity to monitor them consistently.
“We know which signals we want to look for. The problem is that we don’t have the capacity to manually check every company and every signal. Individually, the signals may not be very strong, but when you combine them, they can provide a good indication of what stage a company is in.”
This realization became the starting point for exploring how AI could be used to identify and combine these signals at scale.
The initiative did not begin as a large-scale technology project, but as a practical experiment aimed at solving a specific operational challenge. Rather than asking how AI could be used, the team focused on a task that consumed significant resources and where better information could lead to better decisions.
“First, I needed to figure out whether it was even possible. I started by describing the problem and what I wanted to achieve, and then worked forward from there,” says Vilde.
The first versions did not deliver the desired results. The model was given access to too much information, attempted to analyze too many companies simultaneously, and spent significant resources processing data that was already easily available.
Over time, the workflow was simplified. The volume of data was reduced, companies were analyzed in smaller groups, and the model received clearer instructions on what tasks to perform and how to perform them.
The outcome was not a model that replaces people. It was a tool that enables people to use their time and expertise more efficiently.
The experience illustrates a broader shift across both sales and investment. Rather than relying on fixed follow-up schedules and historical activity, organizations can increasingly prioritize based on signals indicating need, readiness, and timing.
AI cannot replace human judgment, but it can provide a stronger foundation for deciding where to focus attention and when to engage.
Kongens gate 8, 7011 Trondheim
© Viking Growth 2025
.avif)
