“Within a few months, we saw clear benefits from understanding which campaigns worked through A/B testing. Our investment in Pecan paid off just from these early experiments. With Pecan’s tools, even analysts with SQL skills can build predictive models that genuinely drive business improvements.”
Adam Smith, Director of Strategic Analytics
Meet Little Spoon
Little Spoon is a fast-growing subscription service delivering fresh baby food and kids’ meals directly to parents’ homes. With more than 80 million meals shipped, the company focuses on providing high-quality, nutritious food with transparency and convenience.
Inside the company, the Strategic Analytics team supports decisions across marketing, growth, and customer experience. As Little Spoon scaled, the team needed accessible predictive analytics to understand lifetime value, evaluate acquisition channels, and anticipate ordering behavior – all without hiring a data scientist.
The Challenge
Before adopting Pecan AI, Little Spoon relied on manual, high-level LTV estimates that lacked precision. This made it difficult to:
- Understand predictive LTV for every customer
- Evaluate channel and campaign performance to improve ROAS and MROI
- Predict weekly orders and identify upsell opportunities
As the business grew, accurate predictive LTV and order likelihood became essential for efficient spend and sustainable growth.
The Solution
Why Pecan
Little Spoon chose Pecan AI because it offered powerful predictive capabilities at a cost and skill level the team could manage. Analysts could build predictive models using only SQL knowledge, with Pecan’s guidance to get started quickly.
“We needed a solution within our budget that still offered the benefits of predictive analytics, and Pecan AI fit that bill. They offer a great, easy-to-use tool at a cost that is manageable compared to other options.”
Pecan partnered closely to help the team:
- Prepare and structure data effectively
- Build the first predictive LTV model
- Interpret model outputs with confidence
- Transition into self-service predictive modeling
Fast onboarding and early value
With Pecan’s support, Little Spoon generated insights quickly. The team created its first predictive LTV and order-likelihood models in a matter of weeks, not months.
“Pecan allows you to get results quickly, within weeks, not months, and their customer support team will guide you every step of the way.”
How Little Spoon Uses Pecan AI
Predictive LTV for smarter acquisition decisions
Little Spoon now uses predictive LTV to understand:
- The expected value of each new customer
- LTV variation across channels, cohorts and customer attributes
- Which campaigns to scale or discontinue
- Where profitable growth opportunities are strongest
These insights allowed the team to eliminate underperforming channels, invest in effective ones, and improve ROAS and MROI.
Predicting weekly orders for stronger retention
One of Little Spoon’s first models predicted weekly order likelihood – critical for any subscription business. The model revealed:
- Customers likely to place an upcoming order
- Customers at risk of skipping or lapsing
- Individual-level behavior patterns
This helped the team improve communications, lifecycle interventions, and retention tactics immediately.
Expanding into upsell and next-purchase modeling
After early success, the team began building new predictive models independently, including:
- Upsell propensity
- Next-purchase and product-interest models
- Additional behavior-based forecasts
These models support personalization and unlock new revenue opportunities.
Predictive analytics embraced across the company
One of the most meaningful shifts has been cultural. Predictive analytics is now embraced across teams, with stakeholders actively approaching the analytics team with ideas for new predictive models.
“Instead of us pitching ideas to stakeholders, they’re approaching us with ideas on how predictive models can enhance their areas of the business.”
Accessible to analysts with SQL skills
Pecan makes predictive modeling achievable for teams without data science backgrounds.
“Analysts with SQL skills can build predictive models. This ease of use is a key takeaway – no advanced statistical knowledge is required to build impactful models that genuinely drive business improvements.”
Pecan’s support team continues to help the analytics team refine and expand their predictive capabilities.
The Impact
Improved ROAS and MROI
Predictive LTV gave Little Spoon the clarity to scale campaigns that work and remove those that did not. Marketing effectiveness became measurable and actionable.
Stronger retention and recurring revenue
The order-likelihood model helped identify at-risk customers earlier and boosted re-order consistency in the subscription flow.
Fast payback and measurable ROI
Little Spoon saw their investment in Pecan pay off quickly through early A/B testing and improved marketing efficiency.
A self-sufficient analytics team
The team now builds predictive LTV, order propensity, upsell and product-interest models on its own. No data science headcount required.
Predictive analytics driving decisions across the business
Teams across marketing, product and operations now look to predictive models to guide improvements and inform planning.
Key Takeaway
Pecan AI enabled Little Spoon to adopt predictive analytics quickly and affordably. The company now uses predictive LTV and order-likelihood models to improve ROAS, strengthen retention and grow profitably. With intuitive workflows and exceptional support, Pecan empowers SQL analysts to build predictive models that deliver meaningful business value in weeks.
