Klaviyo tells you which variation got more opens. SplitDial tells you whether to believe it, and shows its work, computed from raw events rather than aggregate reports.
Paste a read-only Klaviyo API key. SplitDial finds every A/B test you have ever run, campaign and flow. It never writes to your account except single actions you click yourself.
Instead of trusting Klaviyo's aggregate report, SplitDial reconstructs every test from the underlying events, then removes Apple's privacy auto-opens and bot clicks before anything is scored.
Each test resolves to plain language: a real winner, no real difference, not enough data yet, or an invalid test. The numbers are reconciled against Klaviyo's own report so you can trust them.
Plan tests that can actually resolve at your send volume, catch tests that collide or get cut short, and build a record of what has won before, per brand and pooled across brands.
An inconclusive verdict is the product working. It is the difference between knowing and guessing.
A read-only key is all a sync needs. SplitDial never changes anything in Klaviyo except a single action you explicitly click, and it says so before you click.
API keys are encrypted at rest (AES-256-GCM envelope encryption). They are never logged, and never returned to the browser.
Tenancy is enforced in the database itself with row-level security, so one workspace can never read another's data, even in the event of a bug.
Each sync checks SplitDial's recipient and revenue totals against Klaviyo's own report, and shows you the result, so the verdict rests on numbers you can verify.
SplitDial works from event counts and test structure. It does not store your subscribers' names, emails, or profiles.
The product is put through structured security reviews (over 2,000 individual checks across two passes), with optional two-factor authentication for your own login. A summary is available on request.
See the real tests hub with a fixed sample account, then connect your own when ready.