Loyalty businesses accumulate data in ways that most other organisations do not. Every member interaction, every transaction, every redemption, every partner touchpoint - recorded, timestamped, stored. By the time a data leader joins an established loyalty business, there is typically years of this data sitting across multiple systems, in various states of quality, connected to each other in ways that were logical at the time and are genuinely confusing in retrospect.
The first 60 days in this environment are not about building a data strategy. They are about understanding what actually exists before committing to what should be built next.
This is a practical framework for that assessment - what to map, what questions to ask, and where the problems that most affect commercial outcomes consistently turn up in loyalty data estates.
60% of AI projects unsupported by AI-ready data will be abandoned through 2026
Gartner, 2026
That statistic matters in a loyalty context because AI is increasingly central to what loyalty businesses are building next - personalisation engines, programme intelligence tools, predictive churn models. None of that work reliably on data that has not been assessed and prepared for that purpose. The first 60 days are the investment that makes everything after them faster.
Start With the Data Landscape, Not the Strategy
The instinct when joining a new organisation is to move quickly toward a strategy - a roadmap, a set of priorities, a clear direction. Resist it for the first three to four weeks. The strategy can only be as good as the understanding of what exists, and in loyalty businesses, what exists is almost always more complex than the onboarding documentation suggests.
The data landscape assessment covers three things: what data sources exist, where they live, and how they currently connect to each other. In most established loyalty businesses, the answer involves at minimum a loyalty platform, a CRM, a marketing automation system, a reporting or analytics environment, and some combination of partner data feeds, transactional logs, and historical programme outputs. Each of these has its own schema, its own update cadence, and its own definition of the same concepts.
"Fragmented tooling and operational complexity are the leading challenges for loyalty teams heading into 2026. Data that cannot be connected cannot be commercialised."
- Open Loyalty Trends Report, 2026
The mapping exercise does not need to be exhaustive at this stage. The goal is to understand the shape of the estate - how many systems, what they contain at a high level, and where the joins between them currently exist or are missing. This becomes the foundation on which every subsequent assessment is built.
The Five Specific Places Data Quality Problems Hide in Loyalty Estates
Data quality issues in loyalty businesses have patterns. The same problems appear consistently, regardless of the platform, the sector, or the programme maturity. Knowing where to look before you start the assessment saves significant time.
Member Identity Fragmentation
The most common and commercially significant data quality problem in loyalty is identity fragmentation - the same physical person existing as multiple records across multiple systems, with no reliable mechanism to resolve them into one. This happens because members join through different channels (in-store, online, app), provide slightly different personal details each time, and change their contact information without updating every system. The result is a member base that appears larger than it is, personalisation that cannot work at the individual level, and lifetime value calculations that are structurally wrong.
Semantic Inconsistency Across Systems
The loyalty platform and the CRM frequently disagree about what the same data means. A 'member' in the loyalty platform may have a different definition to a 'contact' in the CRM. 'Active' in the reporting dashboard may use a different lookback window to 'active' in the marketing automation tool. These semantic inconsistencies are invisible in day-to-day operations but produce misleading numbers at the programme level - and incorrect inputs to any AI model built on top of the data.
Partner data gaps
Loyalty programmes that involve reward partners - retailers, travel providers, brand partners - depend on transaction data flowing reliably from each partner into the programme data layer. In practice, partner data feeds are inconsistently structured, intermittently reliable, and frequently undocumented. A programme that shows 40 active reward partners may have reliable, clean data from 15 of them and partial or missing data from the rest. This gap is almost never visible from the outside and takes direct investigation to surface.
Historical data that pre-dates current platform architecture
Most established loyalty businesses have migrated platforms at least once. Each migration carries a risk of data loss, field remapping errors, and historical records that do not conform to the current schema. This legacy data layer frequently contains valuable longitudinal programme intelligence - how redemption patterns have changed over five years, which member cohorts have the highest lifetime value - that is not accessible through current reporting because the historical data cannot be reliably joined to current records.
Consent and compliance gaps
Data collected before current GDPR frameworks were in place, or without the specificity of consent that modern regulatory guidance requires, represents both a compliance risk and a practical data limitation. Members whose data was collected under older consent models may not be contactable for certain types of outreach, which affects any segmentation or activation built on their records. Auditing the consent layer is not optional in a data assessment - it determines which data is actually usable for the purposes most relevant to the business.
The Platform Assessment: Three Questions That Reveal Fitness for Purpose
Alongside the data quality assessment, the loyalty platform architecture needs to be evaluated against three specific questions - not a feature checklist, but genuine fitness-for-purpose questions that determine whether the current stack can deliver what the business needs next.
Can the platform support individual-level personalisation in real time, or is it constrained to batch-processed segment-level offers? The answer determines whether the AI-driven personalisation ambitions that most loyalty businesses now have are achievable on the current infrastructure or require a platform change.
Is programme logic separated from member experience - or are they tightly coupled in a way that makes changing either one difficult without affecting both? Tight coupling is the most common reason programme changes take longer and cost more than expected, and it is visible in the architecture if you look for it.
How is data currently flowing between the loyalty platform and the CRM? Is the connection real-time or batch? Is it bi-directional or one-way? Are there fields in each system that the other cannot see? The answers to these questions determine what is possible with personalisation, cross-sell, and service integration before any new investment is made.
Turning Assessment into Priorities
By the end of 60 days, the assessment should produce a prioritised short list - not a comprehensive data strategy, but a set of decisions about where to invest first to unlock the most commercial value from what already exists. The prioritisation framework is straightforward: rank each issue by two dimensions - the commercial impact of resolving it and the effort required. The issues in the high-impact, lower-effort quadrant are the first-phase investments. These are almost always in data quality and connectivity rather than platform replacement, because a well-connected, reasonably clean data estate produces more immediate commercial return than a new platform with the same underlying data problems. The issues that require significant platform change or new technology investment belong in the second phase - informed by the assessment, validated by the first-phase results, and sequenced so that each investment builds on the foundation of the previous one rather than competing with it.
The 60-Day Output
The goal of the first 60 days is not a finished strategy. It is a clear-eyed picture of what exists, an honest assessment of where the gaps are, and a prioritised short list of investments that the business can make with confidence because they are grounded in evidence rather than assumption.
That picture is also the foundation for every conversation about what to build next - with technology partners, with platform vendors, with the commercial and product teams who depend on the data estate to do their work. The organisations that skip this assessment and move directly to strategy typically end up revisiting the same problems 12 months later when the data turns out not to support the use cases they planned around.
The 60-day assessment is not the slow route. It is the route that makes everything after it faster.
Schedule a 1-on-1 Data Architecture Assessment with the VE3 Team.
