AI & BI in SMEs: A Five-Week Roadmap
How Deuna Integrated AI to Accelerate Decision-Making and Quality
Transforming an SME with Artificial Intelligence (AI) starts with people and data. The story of Deuna Tortilla, a small Madrid-based company specializing in gourmet potatoes, shows how a practical, human-centered approach can create measurable impact in just five weeks.
The Challenge: Seasonality and Limited Resources
Like many companies in the food sector, Deuna faced strong seasonality: higher summer sales along the coast and lower volumes in major cities. Managing these peaks required agility and precise data interpretation. As a small business, Deuna had a lean, multifunctional team where each person handled multiple roles—sales, marketing, and innovation. The challenge was to scale human capacity without increasing costs.
This was the context for launching the AI Kick Off program, led by WalkerTrust, lasting five weeks—three for implementation and two for guided adoption.
The Approach: Data First, Then AI
The first step was mapping all company processes end to end to identify key pain points. A visual exercise using post-its helped discuss critical workflows and prioritize them through a cost–benefit matrix. The sales process was identified as the primary candidate for automation.
1. Business Intelligence (BI) Phase
Before introducing AI, it was crucial to structure data and create an accessible decision dashboard. Simple, low-cost tools were used:
- Google Sheets to centralize client and sales information.
- Looker Studio to build interactive dashboards showing regional sales, margins, and customer trends.
The real breakthrough came from developing a customer segmentation matrix and integrating it with sales data, allowing the team to go beyond basic metrics and understand patterns such as sales growth by segment or promotional impact by region.
2. Generative AI Phase
With structured data in place, Generative AI was integrated into daily workflows. The team first experimented with Gemini but quickly switched to ChatGPT for its stronger compatibility with Google tools. The goal was to free up time and improve quality—without replacing people. Practical applications included:
- Preparing reports for partners (roughly from 60 to 10 minutes, estimated).
- Reading and replying to international emails (from 40 minutes to nearly instant).
- Drafting sales proposals with greater consistency and clarity.
Within two weeks, AI was fully embedded in daily operations, and the team reported tangible improvements in clarity, speed, and confidence.
Results: Productivity, Quality, and Motivation
Even without formal measurement, team perception was clear: repetitive tasks became much faster, written work improved in quality, and decisions were increasingly based on shared visual data.
According to Deuna’s commercial director:
“Being able to send a dashboard snapshot to a client showing their sales evolution is truly differentiating.”
These outcomes align with the Harvard/BCG (2023) study, which found that professionals using generative AI completed tasks approximately 25% faster and with 40% higher quality, in work comparable to this implementation. They also reflect insights from the NBER (2025) paper — How People Use ChatGPT — which highlights that AI’s main economic value lies in augmenting human judgment and supporting complex, knowledge-intensive tasks.
Beyond efficiency, Deuna achieved a +200% increase in sales volume in 2025, during one of the company’s most demanding seasons. The team highlights that the integration of Business Intelligence and AI capabilities played a key role in enabling this performance, helping them to identify patterns for key clients, communicate faster, and more effectively.
Lessons Learned: Less Theory, More Practice
The main takeaway was clear: less theory, more practice. Entrepreneurs learn by doing—and AI, by its very nature, rewards experimentation.
Deuna discovered that success comes from testing, failing fast, and continuously improving.
Best Practices Identified:
- Start by mapping end-to-end processes to identify and prioritize all key pain points—this alignment ensures ownership and commitment from the entire team throughout the project.
- Define a clear Problem Statement anchored in measurable business value.
- Establish a clear performance baseline and measure impact using simple, transparent metrics. This step is critical to demonstrate ROI, particularly in complex implementations with higher investments and more variables.
- Visualize and structure data (e.g., through Looker Studio dashboards and snapshots) before introducing AI or automation.
- Foster a culture of experimentation, learning, and continuous improvement.
Next Steps: Turning AI into a Revenue Engine
The next phase is even more ambitious: turning AI into a direct source of revenue. Deuna is developing automations with n8n to identify business opportunities and anticipate demand peaks linked to local events—like festivals and regional fairs.
With these foundations, the company aims to make AI not just a driver of efficiency but a true engine of growth.
Conclusion: The Future Is Collaborative
Deuna’s experience shows that when applied wisely, AI doesn’t replace people—it amplifies human talent.
In the SME context, it acts as a strategic ally that expands execution capacity, accelerates growth, and reinforces collective purpose.
The future doesn’t belong to machines, but to the people who know how to work with them.