That distinction matters when businesses evaluate connected customer data. A company can connect many systems and still have weak identity matching or outdated records. Another company may connect fewer sources while giving sales or service teams much more dependable customer context. The useful benchmark is therefore tied to what connected data changes in daily decisions and how accurately the platform represents the customer.
The 42% benchmark measures adoption rather than implementation quality
The Adobe 2026 AI and Digital Trends research gives companies a useful comparison point because its sample spans industries and business sizes. Adobe found that 49% of surveyed organizations said their ability to advance AI initiatives was limited by their current level of data unification and structure. Among martech leaders, 74% cited data integration and quality as a top barrier to implementing agentic AI.
Those percentages measure reported conditions across a global sample. They don’t establish how many systems a specific company should connect or what identity-match rate it should reach. Industry rules can change the acceptable data model, while company size affects the number of source systems involved. Timing matters too because an organization early in a migration shouldn’t be compared directly with one that has spent years cleaning customer records.
A Salesforce Data Cloud Implementation should therefore begin with an internal baseline. Teams need to know how many customer systems currently operate separately and which business decisions suffer because of that separation. The benchmark becomes more useful when it’s attached to a measurable problem rather than treated as an industry score.
Connected data changes the customer from several records into one reference profile
Salesforce uses identity resolution to connect source profiles that appear to belong to the same person or account. Its current documentation explains that unified profiles provide access to source records associated with that customer. Salesforce also makes an important distinction: a unified profile isn’t a golden record and doesn’t function as a traditional master data management system. Salesforce’s identity resolution guidance explains that match rules determine which records should be linked.
Salesforce also renamed Data Cloud to Data 360 on October 14, 2025, although current documentation can still contain Data Cloud references during the transition. The naming change doesn’t alter the underlying evaluation problem. Teams still need to determine whether the connected profile represents the customer accurately enough for the intended use case.
That is where Salesforce Data Cloud becomes more than a connection layer. The platform can bring source identities into a shared profile that sales, service, marketing, or AI applications can reference. The useful measure is how much uncertainty remains after those records are connected.
Identity match quality matters more than the number of connected sources
A high source count can look impressive while hiding poor matching. If email addresses differ between systems or household members share contact details, identity rules can link records incorrectly. Conservative rules can create the opposite problem by leaving several profiles for the same person.
An internal benchmark should track the percentage of source records that resolve into expected profiles and the number of suspected false matches. Teams should also compare profile freshness against the business process using the data. A service use case may need current account status much faster than a quarterly reporting use case.
This is why Salesforce Data Cloud Consulting should evaluate identity rules before judging success by record volume. HyphenX states that its Data Cloud work includes data streams, identity resolution, unified profiles, segmentation, activation, governance, and monitoring. Its page describes about 10 to 16 weeks as a typical implementation range when several systems are being connected, which should be treated as a planning estimate rather than a universal delivery benchmark.
Customer expectations explain why connected profiles matter
The business case for connected data becomes clearer when customer expectations are added to the comparison. Salesforce’s current State of the AI Connected Customer research reports that 73% of customers say companies treat them like an individual rather than a number, up from 39% in 2023. The same research reports that 71% of customers feel increasingly protective of their personal information.
Those figures create 2 separate tests for customer-data programs. Connected profiles need to make interactions more relevant, while access and use of personal information must remain controlled. A higher profile-match rate has limited value if the organization can’t explain where data came from or who should be allowed to use it.
For this reason, companies comparing themselves with a market percentage should also measure consent coverage and data-access controls. Geography can materially change those requirements because privacy obligations vary by jurisdiction. Industry rules in healthcare or financial services can create additional limits that a general global survey won’t capture.
Personalization benchmarks can expose gaps between internal confidence and customer experience
Internal performance measures can overstate what customers actually experience. Twilio’s 2025 customer engagement trends research cites a gap in which 84% of businesses said they provided good or excellent personalized engagement, while 54% of consumers agreed. Twilio links that difference to issues that include siloed systems and inconsistent data quality.
That gap is more useful as a warning than as a score. It suggests companies should compare operational measures with customer outcomes instead of assuming that connected technology automatically produces a connected experience. Retail organizations may evaluate offer relevance, while a B2B company may care more about whether account teams share current contact and activity information.
A Salesforce Data Cloud Consultant can help define those internal comparisons before configuration choices are made. The starting question should identify which customer decisions currently rely on incomplete records and what measurable change would show that the connection is working.
The strongest benchmark starts with the company’s own baseline
External research helps establish context, but internal before-and-after measures provide a clearer test of progress. A company can record duplicate-profile rates before identity resolution and compare them after rules are applied. It can also track how quickly new source data becomes available for the use case that depen
