Why integration alone cannot create a digital government

Governments around the world have invested heavily in interoperability platforms to enable data exchange across agencies.

The expectation is clear: once systems are connected, services will become seamless and efficient. But in practice, this promise often falls short.

The issue is not technology. In many cases, the platforms work exactly as designed.

The real problem is that governments are trying to integrate systems without fully understanding the data inside them.

Kyrgyzstan’s experience

Kyrgyzstan’s experience offers a useful example. The country implemented an interoperability platform based on Estonia’s X-Road, one of the most advanced solutions for secure data exchange.

Today, it supports large-scale interactions between government agencies and businesses.

However, as integration expanded and a high volume of data was exchanged, a critical gap emerged as the platform adoption did not translate to effective service delivery.

Based on my (Meder Toitonov) experience working with the Kyrgyzstan government as an independent consultant on interoperability and data systems, I find that the challenge isn’t limited to a single agency but has appeared across multiple government entities.

One example is service delivery workflow that relies on citizen data across multiple agencies.

While the interoperability platform allows agencies to request data from each other, there are differences in how key fields are defined - such as address formats, status codes, or identifiers – which require additional manual verification.

In practice, this means that even when data is technically available, frontline staff still need to reconcile discrepancies between systems.

As a result, services that were expected to be digital still involve delays, repeated data checks, or requests for citizens to resubmit information.

In practice, this leads to tangible issues for both governments and citizens, as citizens are asked to provide the same data multiple times, agencies interpret shared data differently, and services are delayed due to manual verification and reconciliation.

Several systemic challenges sit behind these problems, including the lack of unified data catalogue across government, classifications and reference data not harmonised, and that data ownership and sources of truth not clearly defined.

As a result, data flows between systems but it is not always consistently understood or reliably used.

This highlights an important distinction: system interoperability does not equal data interoperability.

What the gap looks like in practice

In real-world government operations, this gap leads to two recurring problems.

The first is an invisible data landscape. Governments often lack a clear, consolidated view of what data exists, where it resides, and how it is used. This makes it difficult to design services, govern data, or even assess data quality.

The second is semantic fragmentation. Even when data is available, it is structured differently across systems, uses inconsistent classifications, or relies on incompatible identifiers.

This creates friction in service delivery, increases manual work, and introduces errors.

These challenges are not unique to Kyrgyzstan. Across Central Asia, a similar pattern is emerging: interoperability infrastructure is developing faster than data governance practices.

Rethinking the approach: from integration-first to data-first

This calls for a shift in how digital government is approached.

Instead of focusing primarily on connecting systems (an “integration-first” model), governments need to adopt a “data-first” approach - one that starts with understanding and organising data before scaling integration.

In practice, this involves a set of concrete, implementable steps:

1. Start with a pilot data inventory in one or two high-impact domains (for example, population or business registers)
2. Assign clear data ownership to specific institutions responsible for data quality and maintenance
3. Harmonise key classifications, code lists, and identifiers used across systems
4. Establish a small set of reference datasets that can be reused across multiple services

These steps do not replace interoperability platforms - they make them work.

Where integration infrastructure already exists, a data-first approach significantly increases its impact. Data exchange becomes more meaningful, services become more reliable, and coordination across agencies improves.

What governments can do in the next 90 days

For governments looking to move from concept to action, the transition can start small.

A 90-day timeframe is often used in public sector reforms as a practical window to test and demonstrate quick wins without requiring large-scale institutional change.

A practical 90-day plan could include:

• Identifying  one to two critical datasets used across multiple agencies
• Mapping how these datasets are currently defined and exchanged
• Assigning interim data owners responsible for coordination
• Documenting key inconsistencies in classifications and identifiers
• Launching a pilot effort to harmonise one shared dataset

This kind of focused effort can quickly demonstrate value and build momentum for broader reform.

The experience of Central Asia shows that integration alone is not enough.Digital government does not begin with connecting systems.

It begins with understanding, structuring, and governing data.

-------------------------------------------------------------

Meder Toitonov is an IT consultant working on digital transformation in the public sector, with a focus on data governance and interoperability. His recent work includes national-level projects in Kyrgyz Republic related to metadata repositories and government data exchange systems.

Felix Shapiro is an IT expert in government systems architecture, performance improvement expert (HICD), and business analyst. Gold Member of the International Society for Performance Improvement (ISPI), Certified Scrum Master.