A/B Testing Database Storage (Per-User) Calculator
Estimate the database storage needed per user for A/B test assignments, exposure events, and conversion events over a chosen retention period.
Overview
Estimate how much database storage your A/B testing program uses for each user. Enter the number of active experiments, daily exposure and conversion records, retention period, average record sizes, and a storage overhead allowance.
How it works
The calculator stores one assignment record per active experiment, then estimates retained event records from the number of experiments, events per day, and retention days. It adds assignment and event storage together, then increases the total by your overhead percentage. This gives a per-user estimate that can be multiplied by expected active or retained users for broader capacity planning.
How to use this calculator
- 1Enter the average number of active experiments assigned to each user.
- 2Add the typical daily exposure and conversion event volume for each experiment.
- 3Set the number of days that event-level records are retained.
- 4Enter average byte sizes for assignment and event records.
- 5Add an overhead allowance and review the estimated storage per user.
Example Calculation
Active experiments per user
5
Exposure events per experiment per day
1
Conversion events per experiment per day
0.1
Event retention period
90
Assignment record size
150
Event record size
300
Storage overhead
30%
Estimated storage per user
189.48 KB
This setup estimates about 188.96 KB of A/B testing storage per user, including a 30% allowance for database overhead.
Frequently asked questions
What is included in the per-user storage estimate?
It includes experiment assignment records plus retained exposure and conversion event records, with an optional overhead allowance.
Why does retention period affect A/B testing storage?
Longer retention keeps more event records for each user, so event-level storage increases in direct proportion to retained days.
Should I include database indexes in the overhead percentage?
Yes. The overhead field is intended to allow for indexes, row metadata, storage-engine overhead, and similar costs not included in raw record sizes.
Does this calculator include experiment configuration storage?
No. Shared data such as experiment definitions, variants, targeting rules, and reporting tables is not allocated to individual users here.
How can I estimate total A/B testing storage?
Multiply the per-user estimate by the relevant number of users, then add separate capacity for replicas, backups, archives, and other systems if applicable.
What event record size should I use?
Use a representative average from your database or event schema. Include stored identifiers, timestamps, experiment details, event properties, and serialization overhead where relevant.
Explore Related Calculators
Assumptions and warnings
Assumptions
- Each user has one stored assignment record for every active experiment.
- Exposure and conversion events are retained for the full retention period entered.
- The average event record size includes the fields stored with each event, such as IDs, timestamps, and properties.
- The overhead percentage is a planning allowance for indexes, row metadata, and similar database costs.
- Results are per user and do not include shared experiment configuration, dashboards, backups, or data warehouse copies.
Warnings
- This is a capacity-planning estimate; actual storage depends on schema design, compression, indexing, replication, and retention policies.
- Multiply the per-user result by an appropriate user count and include separate allowances for backups and replicas when planning infrastructure.