“Since I started using Pecan, it took me about a month to see real, tangible results. When I started, I knew almost nothing about the theory or the practical development of predictive models.”
Juraj Rosa, CEO and Founder, Habitable
Meet Habitable
Habitable helps B2B leaders prioritize sales effort and uncover revenue opportunities using predictive analytics. One of its clients is a large European job platform serving tens of thousands of employers across Central Europe. The SME sales team consisted of eight reps responsible for nearly 40,000 accounts without a data-driven way to prioritize where to focus.
Before Pecan AI, Habitable relied on external analysts and data scientists for predictive work, which reduced visibility, slowed timelines, and made consistent quality difficult to maintain. Habitable needed a simple, in-house platform that would let them build predictive models themselves and fully control output quality and speed.
The Challenge
Before Pecan AI, Habitable faced several challenges:
- Reliance on external data analysts and data scientists
- Limited visibility into predictive workflows
- Lack of control over timelines, data quality, and delivery standards
As Juraj explains: “We had to always rely on external data analysts and data scientists, and so we didn’t have full and comprehensive control over our projects.”
At the same time, Habitable’s client, the European job platform, needed to understand the future revenue potential of more than 40,000 existing customers in order to:
- Identify roughly 5,000 high-value accounts and allocate them evenly across eight sales reps (about 600 each)
- Improve conversion rates by focusing resources on the most promising opportunities
- Guide sales planning with forward-looking insight instead of intuition
Habitable needed a platform that could work directly with raw transactional data, scale to tens of thousands of records, and eliminate dependency on external experts.
The Solution
Why Pecan
Pecan stood out for its simplicity and guided workflow. Juraj quickly saw how easy it was to navigate the platform and begin building models without prior predictive modeling experience.
“When I gave it a try, I immediately saw how easy it is to navigate the platform, how easy it is to learn the basics, and I was almost ready to get started from the get-go.”
Pecan offered:
- A clear workflow from predictive question to deployed model
- Built-in modeling capabilities
- A way to bring predictive work fully in-house
Building LTV predictions from raw data
Habitable partnered with Pecan to build a next-year lifetime value (LTV) model using raw order-level history, automatically engineered behavioral and time-based features, and company-level financials and macroeconomic indicators.
“We were able to upload the data or read the data, and then let Pecan do its magic. Thanks to this, we were able, almost in no time, to generate the lifetime value for all these customers.”
The model delivered predicted LTV for about 40,000 companies, a ranked list of the top 4,800 highest-potential accounts, and clear allocation of 600 high-value accounts per rep.
From beginner to builder in one month
Juraj learned Pecan from scratch and delivered a production-grade predictive project in about a month, supported by Pecan’s guided onboarding and weekly working sessions.
This project created a repeatable blueprint Habitable now uses across clients.
Customer Success as an extension of the team
Pecan’s Customer Success team guided Habitable through its first project, ensuring the data was prepared correctly, the modeling workflow was structured effectively, and the first delivery was both accurate and fast.
How Habitable Uses Pecan AI
Predictive LTV at scale
Predicting next-year revenue for tens of thousands of customers with continuously refreshed scores.
Smarter account prioritization
Assigning each rep a focused book of the 600 customers with the highest predicted potential, replacing intuition with data-driven focus.
Territory and resource planning
Identifying growth-ready accounts, spotting early signs of decline, and planning outreach accordingly.
A repeatable predictive framework
The LTV workflow now powers new markets and additional use cases such as cross-sell, upsell, and reduction-in-activity models.
“Being able to predict future behavior of your customers gives you superpowers. You are able not only to see into the future, but also take preemptive actions to improve the future.”
The Impact
40,000 customer LTV predictions delivered quickly
Habitable generated accurate predictions for tens of thousands of customers in a short timeframe.
4,800 high-potential accounts prioritized
Sales leaders now focus on the accounts with the highest future value, not just past spend.
1 month from blank slate to real results
Habitable went from complete beginner to delivering tangible client outcomes in about a month.
Full ownership without external data scientists
Habitable now controls the entire predictive workflow internally – data, quality, timelines, and delivery.
Predictive modeling as a core capability
Predictive insights now power Habitable’s sales automation strategy across clients and markets.
Key Takeaway
Pecan AI enabled Habitable to replace external data science dependency with full in-house predictive capability. With fast time-to-value, guided onboarding, and scalable workflows, Habitable now delivers predictive models for tens of thousands of customers, empowers clients to act on future opportunities, and extends these capabilities across new markets and use cases.
