A set of numbers has been making the rounds in the industry lately. According to data from CCID Consulting and China Merchants Securities, the domestic database software market reached 7.93 billion yuan in 2025, a 26.89% market share, and is expected to cross 10 billion yuan in 2026 with a share approaching 30%. The vendor ranking is equally clear. Dameng leads domestic vendors at 13.2%, Kingbase follows at 11.5%, and OceanBase sits third at 10.8%. Together the three already hold more than 35% of China’s relational database market. China Merchants Securities’ judgment is that once the share breaks 30%, domestic databases will accelerate into a phase of replacement at scale.
The 30% claim comes with theoretical packaging. Economies of scale, a larger user base speeds up both product iteration and ecosystem growth. An anchoring effect reversal, using a domestic database goes from being a risk to not being a risk. Perimeter systems get validated, then core systems take over. All three arguments sound reasonable. I dug up data from different research firms over the past few years and cross-checked them, and it turns out the story behind this tipping point is more interesting than what the reports say.
Start with a reversal that already happened. By IDC’s measure, the combined share of domestic vendors among the top five in China’s relational database market was 27.1% in 2018 and 55.4% in 2022. Over the same period, the top five international vendors fell from 57.3% to 27.3%. In other words, counted by vendor share, domestic databases passed the halfway mark back in 2022. So what is the 26.89% figure for 2025? The answer is statistical scope. One set of numbers measures each vendor’s share of the entire database market, while the other measures domestic database software as a share of the software market, one wide and one narrow. Both are correct, and they describe different sides of the same story. Reading these reports the way a DBA reads monitoring data, check the scope before trusting the conclusion, works just as well on industry data.
The more persuasive evidence sits in the industry breakdown. First New Voice’s 2024 assessment gave replacement rates by sector. Party and government offices stand at 80%, with stock replacement at the provincial and ministry level nearly finished and now moving down to cities and districts. Financial institutions have replaced roughly 40% of non-core systems. Energy is under 15%, and healthcare, manufacturing, and several other sectors are below 5%. These numbers are more direct than any tipping point theory. The anchoring reversal has already played out in party and government, where choosing domestic no longer takes courage. Finance has finished its perimeter and is standing at the door of the core. Energy, healthcare, and manufacturing are still at the starting line. The next phase of replacement at scale has already been mapped out, and it is those 15% and 5% figures.
Behind all this sits a timetable that cannot be ignored. Document No. 79, issued by the State-owned Assets Supervision and Administration Commission in September 2022, requires central and state-owned enterprises to complete the replacement of foundational software with domestic alternatives by the end of 2027, with databases in the full-replacement category. Count the calendar, and a little over a year remains until that deadline. So the 2026 push toward 30% is half organic market growth and half a procurement rhythm driven by a due date. The tipping point narrative dresses up a policy gear shift as spontaneous market behavior, and that packaging needs to come off. The market is real, the results are real, and a substantial part of the momentum comes from official documents. Both things are true at once.
For DBAs, there is a more practical way to read these numbers. The geography of job demand is already visible in them. Replacement in party-government and financial perimeter systems is winding down, so work in those areas will shift from building to maintenance, with plenty of competitors and squeezed rates. The sub-15% replacement rates in energy, healthcare, and manufacturing mean these industries will see concentrated project rollouts over the next two or three years, from selection and evaluation through migration to dual-track operation, and every step needs people. On the vendor side, Dameng, Kingbase, and OceanBase together hold over 35% of the market, so betting your skill development on these three carries the highest hit rate. One more reference point, OceanBase is in talks for its first external funding round of 2 to 3 billion yuan on annualized revenue of over 1.4 billion, so capital is also voting for the leaders with money.
Beyond the numbers, one observation of my own. The word tipping point frames an industry shift as a single moment, as if everything gets easier once you cross it. What the process actually looks like is a long-distance run. Party-government took nearly a decade, finance’s perimeter took five or six years, and every step was stacked up project by project. 30% will not be a magic number, and 2027 will not be the finish line. In the years ahead, core systems will only run more and more domestic databases, and when something breaks, the person standing in front of the machine will be the one reading this article today.
Sources
CCID Consulting, “2025-2026 China Platform Software Market Research Annual Report,” and related China Merchants Securities analysis First New Voice, “2025 China Financial Industry Database Domestic Replacement Capability Assessment Report” Eastmoney, “Domestic penetration of core foundational software remains low; a replacement peak may arrive in the short term” IDC historical market share data for China’s relational database market