The Evolution of BigQuery Billing: Chasing the Perfect Balance

2 min read

Apr 30, 2026 5:04:23 PM

3:24

As a long time Google Cloud practitioner and a frequent NEXT flyer, there was one announcement that really caught my attention this year that excites me. I have been working with and selling BigQuery for close to a decade, and one thing I always had practice was walking the thin webline like Spiderman, balancing “Great power with great responsibility.” While fast results are expected by consumers and easily within our reach, we must ask if the infrastructure scaling costs truly justify the outcome. Join me on my walk through BigQuery memory lane:

The Wild West: On-Demand Pricing and Analysis

We began with a simple but potentially dangerous model: On-Demand Pricing (Analysis Pricing). It was a blank check for speed, you scanned the bytes, you got the answer, and your query ran at full tilt. The power was mesmerizing, but for highly variable workloads, it could lead to a very angry call from the finance department. The check always came due.

Taming the Beast: The Era of Reserved Slots

To tame the beast, Google introduced Reserved Slots (Commitments). This was our anchor: pre-purchasing 1-year or 3-year slot commitments for a static, fixed price. It offered safety, price predictability, but at another cost. We traded performance for safety, as fixed capacity could artificially throttle speed during peak demand or leave us paying for unused capacity during slow periods.

Bridging the Gap: The Rise of Autoscaling Slots

The next evolutionary step, Autoscaling Slots, tried to bridge the gap. We secured baseline slots and gave ourselves the ability to burst and autoscale up to a reservation maximum. It was better, but still rigid. Slots were billed in chunky 100-slot increments with a punitive 60-second minimum. For quick, high-frequency queries, this often meant paying for capacity we barely touched, making it frustratingly inefficient.

True Freedom: Introducing Fluid Scaling

Now, we have Fluid Scaling. This is the closest we’ve come to true freedom, offering the maximum flexibility in both cost and performance. By introducing true per-second billing, we’ve finally eliminated the 60-second minimum and the rigid increments, effectively destroying the historical trade-off between speed and efficiency (which feels like 2017). Costs are finally—and justly—tied to the exact resources needed and consumed.

Ultimately, this means a better alignment between cost and actual usage, with Google positioning the potential for up to 34% in cost savings with this new feature.

Maximizing Your BigQuery Investment with Fluid Scaling

Whether you are an existing BigQuery user looking to assess your costs, or are exploring the power of BigQuery for your AI workloads, Fluid scaling presents a significant opportunity. Pythian's experts can help you assess your current BigQuery spending and demonstrate how to leverage these cost saving methods, without negatively impacting performance. For those with AI and ML workloads, we can help you design a cost-effective architecture from Day 0 that takes full advantage of this new flexibility.