All blog posts

Tiqum: replacing imported databases slowed down every fifth online store

A technical audit of 100 retail sites revealed hidden risks: incorrect data migration leads to slower page loads and a rise in cyberattacks

Tiqum: replacing imported databases slowed down every fifth online store
In 2025 Russian retail ran into “growing pains” during the forced switch to domestic database management systems. In a technical audit of 100 large and medium-sized e-commerce sites, analysts at the digital integrator Tiqum found that for 20% of retailers the response time of critically important pages (personal account, cart, checkout) worsened by 15–30% after migrating from foreign software.

The main problem of migration is the attempt at a “mirror” transfer of complex business logic from systems like Oracle or MS SQL to PostgreSQL-based solutions without redesigning the architecture. Unoptimized queries put excessive load on servers, which directly affects the user experience.

According to Tiqum's research, every extra second of waiting in 2025 costs retail up to 8.4% of conversion. Given that many users are ready to leave a site if a page takes more than 10 seconds to load, a technical slowdown becomes a direct threat to revenue.

Alongside the drop in performance, the audit revealed a growth in vulnerabilities. In the first half of 2025 the number of cyberattacks on online stores in Russia almost tripled, reaching 87.4 million incidents. The attackers' main target is customer databases.

Tiqum's experts note that a “quick” import substitution often neglects a deep security audit:

  • API vulnerabilities: 35% of the sites had errors in integration gateways that allow unauthorized access to personal accounts.
  • Authorization errors: attacks through employees' personal devices grew by 30% in 2025, which makes the unprotected admin panels of online stores an easy target.

“Migrating to a domestic stack is not just a change of licenses but a complete revision of the IT landscape,” Tiqum stresses.

To stay competitive, retailers are advised to audit their system architecture and code regularly and to introduce automated load testing before any large-scale updates.

Optimizing the data structure and introducing modern compression methods restores the usual speed and cuts the volume of stored data by up to 50%, which lowers the cost of maintaining the infrastructure.

More news

Telegram channel “Founder for Life”

@gizu_ru

Thoughts on technology, business and the future, before they reach the blog. The channel is in Russian

4,000+ subscribers Subscribe