Artificial intelligence news.
Straight from the original announcements.

Fivetran finds data rework is driving an AI tax for companies

November 12, 2026, 20:24By the AI News Bites Editorial TeamSource: Fivetran
Fivetran finds data rework is driving an AI tax for companies
Image: Fivetran

Companies are losing a significant share of their AI spending to avoidable data work, according to new Fivetran research. Organizations that devote at least 25% of AI budgets to this work are about 5x as likely to report major project delays.

Key points
  • 39% of respondents spent at least a quarter of AI/ML budgets on avoidable work.
  • Organizations with a high AI tax were about 5x as likely to report major delays.
  • Data quality and integration accounted for unexpected work, delays, or costs for 66% of respondents.
  • Only 37% of organizations had multiple AI initiatives operating at scale.
  • 31% would direct additional AI investment to the data foundation.

Data problems are consuming AI budgets

Fivetran surveyed 500 senior IT, data, and AI/ML decision makers about their organizations’ AI spending and projects. The company calls the share of AI/ML spending consumed by avoidable rework the “AI tax.” This includes fixing data quality issues, rebuilding integrations, reprocessing data, and retraining models.

According to Fivetran, 39% of respondents said at least a quarter of their AI/ML spending went to avoidable work during the past 12 months. Only 16% said the AI tax stayed below 10%. Spending on AI is already substantial: more than 80% of organizations surveyed spent at least $1 million on AI/ML during the past year, while 54% spent $5 million or more.

Despite that investment, only 37% of organizations had multiple AI initiatives live in production and operating at scale. Fivetran said 66% of respondents traced unexpected work, delays, or costs to data quality and data integration.

A high AI tax is linked to delays and weaker returns

The research connects poor data foundations with stalled projects and longer delays. Among organizations with a low AI tax, 17% said at least a quarter of their AI initiatives had been delayed or stalled by data foundation gaps. That share rose to 50% for organizations with a moderate AI tax and 86% for those with a high tax.

Seventy-six percent of organizations in the high tax group said delays typically lasted 3 weeks or more, compared with 46% in the low tax group. Reported returns also differed. Sixty-two percent of low tax organizations said AI ROI exceeded expectations, compared with 40% of organizations in the very high tax group.

Poor data also affected confidence in AI systems. Fivetran said 57% of respondents reported inaccurate outputs or predictions, and the same share had to retrain or recalibrate models. Across every region surveyed, 51% to 54% said poor data caused stakeholders to lose trust in AI outputs.

Standardized integration could reduce repeated work

How organizations connect data was one of the clearest differences in the findings. Among companies relying mainly on custom, project specific integrations, 53% said at least a quarter of their AI budgets went to avoidable work. The figure was 40% for organizations using standardized integration platforms and 35% for those using a standardized platform while keeping some custom development for specialized needs.

Fivetran said the findings do not mean companies should eliminate custom development. Instead, they point to using shared integration infrastructure for common work and reserving custom solutions for exceptions. The company also recommends tracking data preparation, integration, reprocessing, and maintenance costs across projects.

Ownership is another factor. Forty-two percent of respondents cited unclear ownership or cross-team misalignment as a reason initiatives stall or underperform. Organizations reporting strong alignment were more than twice as likely to say ROI exceeded expectations as those reporting misalignment.

If AI/ML budgets increase over the next 12 months, 31% of respondents said they would direct additional investment to the data foundation, more than any other category. Fivetran said 41% expect eliminating avoidable data work to improve AI time to value by more than 25%, while nearly 9 in 10 expect an improvement of more than 10%.

Related news

Huawei unveils AI Data Platform for European businesses
Chips and infrastructureNovember 16, 2026, 09:56

Huawei unveils AI Data Platform for European businesses

Huawei has introduced a new AI Data Platform in Florence, positioning it as infrastructure for improving data management and the performance of artificial intelligence…

SAS and Posit Partner on Governed AI for Life Sciences
BusinessNovember 16, 2026, 09:00

SAS and Posit Partner on Governed AI for Life Sciences

SAS and Posit PBC have formed a partnership to give life sciences organizations more flexibility in using SAS and open-source technologies, initially for clinical…

Fivetran study finds coding model price is a poor value guide
ResearchNovember 9, 2026, 17:02

Fivetran study finds coding model price is a poor value guide

A blinded Fivetran study of 39 engineers found that paying substantially more for an OpenAI coding model produced only a modest preference advantage in everyday software…