We have CRM adoption, but a lot of work remains on data

76.9% of Italian companies have adopted a CRM, often cloud-based, often integrated with ERP, but more than half of those same companies enter data manually, have no data governance policy, and lose hours each week searching for and correcting wrong information. In other words: the tool is there, but the foundations are not, according to research by SDA Bocconi School of Management on 65 firms.

The result is, in its own way, revealing. It dispels any suspicion of technological backwardness, as cloud adoption rates exceed European averages across many business categories, but it shows that surface-level digitalization can coexist with deep operational fragility. And when this fragility overlaps with enthusiasm for artificial intelligence, with 30.8% of the sample already using Generative AI tools, the picture becomes potentially dangerous. Running algorithms on degraded data does not produce insights, but false certainties.

Data quality and governance

In recent years, academic and managerial literature on CRM has followed a path of progressive evolution: from sales force automation in the 1990s, to customer lifecycle management in the 2000s, to the more recent ambitions of Customer Intelligence, that is, the ability to transform relational data into forecasts, decisions, and competitive advantage. Along this path, the critical variable has never been the adoption of technology per se, but the quality and governance of the data that the technology is meant to process.

The 2026 SDA Bocconi research aims to answer four questions:

• What is the actual level of CRM adoption and integration in Italian companies?
• What is the quality of the data feeding these systems, and who is responsible for it?
• To what extent are companies actually ready to adopt artificial intelligence solutions?
• Are there systematic gaps by size, sector, geographic area, or organizational role that help explain any disparities?

The research involved respondents from companies distributed across three macro-geographic areas (Northwest, Northeast, Center-South), representing 11 sectors and two size brackets (below and above €50 million in revenue). The sample includes sales and marketing directors (32.8%), owners and CEOs (9.4%), IT managers (3.1%), and other senior roles. The structured questionnaire, divided into 7 sections and 15 questions, collected data on CRM adoption and integration, intensity of use, data quality and governance, AI readiness, perceived usability, and hidden operating costs.

Limited customer intelligence and tens of thousands of euros in hidden costs

The most relevant results can be summarized along four topics.

Adoption and use. 76.9% of companies have a CRM, with a clear prevalence of cloud solutions (86% among adopters). Integration with ERP exceeds that with Marketing Automation (64% vs. 58%), a sign that CRM is still primarily perceived as an operational management tool rather than a customer intelligence platform. Usage is concentrated in the conversion phase (39.3%), while post-sales (17.9%) and upsell/cross-sell (8.9%) remain largely underutilized.

Data quality and governance. 47% of staff enter data manually, 37.5% of companies have no data governance policy, and 65.5% of staff lose at least one hour per week searching for and correcting data. The estimated hidden cost, for a sales team of five people at €35/hour, ranges between €9,000 and €54,600 annually. And this is a lower bound, as it does not include the cost of wrong decisions or missed sales activities.

AI readiness. Only 16.7% of the sample considers itself truly ready for AI. 57.4% believe their data is inadequate. The three main obstacles—costs, lack of skills, and legacy systems—carry equal weight, confirming that this is not a localized issue but a systemic one. 30.8% are already using Generative AI on inadequate data foundations: what researchers define as “the illusion of AI without foundations.”

The trust gap. One of the most significant organizational findings concerns the misalignment between CEOs and operational managers. CEOs report data as 100% good or excellent; operational staff report 45.2% manual data entry. 50% of CEOs claim to already use autonomous agents, compared to 12.9% of sales and marketing directors. Investment decisions in AI are therefore made by those with the least accurate perception of the real problem.

Segment analyses confirm and deepen this picture. Tech, Finance, and Healthcare lead with 37.5% AI readiness; Industry & Manufacturing stands at 0% AI readiness despite 80% cloud CRM adoption; Commerce & Services shows 64.3% manual data entry and near-zero AI readiness (7.1%). Large companies (>€50M) adopt more predictive AI (33.3% vs. 17.4%), but show 73.3% manual data entry—almost double that of SMEs.

The four conclusions of the research outline a priority agenda for corporate leaders and policymakers.

Who is ready to use autonomous agents?

The first conclusion is that technology adoption alone is not sufficient. The gap between data quality and CRM system adoption is greatest in larger companies: scale and surface-level digitalization do not automatically produce data maturity. Managers must stop measuring digital progress by the number of tools adopted and start measuring it by the quality of the data those tools produce.

The second conclusion concerns the structural Trust Gap. Misalignment between top management and the organization regarding data quality perception is one of the most underestimated managerial risks. AI investment decisions are made based on an upwardly biased view of reality. The solution is organizational and cultural: open more transparent vertical communication channels, involve operational managers in AI decision-making processes, and build shared KPIs on data quality that cut across hierarchical levels.

The third conclusion concerns sectoral differentiation. AI is advancing at radically different speeds across industries. Tech, Finance, and Healthcare confirm their leadership; Industry & Manufacturing is stalled; Commerce & Services risks falling behind despite declared digital ambitions. For policymakers, this suggests the need for differentiated interventions: not horizontal measures, but vertical transformation programs that address the structural specificities of each sector—legacy systems in manufacturing, fragmented touchpoints in retail, and a widespread skills gap across the board.

The fourth conclusion is the most forward-looking. Autonomous agents—the next level of AI applied to business—reward those who have invested in data governance. Companies with high AI readiness are concentrated among large firms (56.7%) and in the Tech/Fin/HC sectors (55.6%). The competitive bifurcation between those who have built the foundations and those who have not is set to widen rapidly over the next 12–24 months. Those who fail to act on data governance now are not simply delaying a technological investment: they are progressively and potentially irreversibly losing competitive position.

The results of this research provide an operational compass for those participating in advanced managerial training programs in marketing, CRM, and Customer Intelligence. Understanding that the real constraint is not access to technology but data quality and that this quality requires governance, processes, and organizational culture is the starting point for a mature approach to customer intelligence.

The CRM Analytics e AI program (in Italian) addresses exactly this path: from understanding the fundamentals of CRM as a knowledge platform, to the strategic management of data, to the implications of artificial intelligence for marketing decisions. The findings of this research offer the real-world context in which those skills must be deployed.