Modules
One engine, six working parts.
Consolidate the data, define the policy, produce the audience, hold out a control group, measure the lift and keep the evidence.
- 01
Signals
Purchase history, reorder timing, spend, engagement and consent, consolidated from every connected source into one customer profile.
- 02
Audiences
A current, campaign-eligible list that refreshes as data changes and passes into Klaviyo or the client's own automation.
- 03
Campaigns
Standing policies that define who qualifies, who is suppressed and how often the audience is re-evaluated.
- 04
Experiments
Randomized holdout allocation built into the policy, so a control group exists before the campaign runs, not after.
- 05
Results
Treated-versus-holdout comparison and estimated incremental revenue, separating caused revenue from normal purchasing.
- 06
Evidence
The conditions each customer satisfied, the reason others did not qualify, and retained source records behind every decision.
Signals
What the consolidated data looks like.
Commerce, marketing and subscription data in one place, so policies are written against real conditions. Each panel is a signal a policy can use.
Where is revenue being left on the table?
Revenue, customers and buying patterns, consolidated from commerce data.
Revenue by month
Net of refunds, cancelled orders excluded, last 24 months
Customer lifecycle
Where every customer is today
- Prospect46%
- New1%
- Active3%
- Active subscriber1%
- At risk3%
- Lapsed46%
Usual device
From forms and orders, last 12 months
- Mobile53%
- Desktop46%
- Tablet1%
New and returning buyers
People who ordered each month: first order that month, or back again
- Returning
- New
Who can be contacted, and how often?
Send volume, engagement and deliverability, read from your sending platform.
Emails sent per day
Last 30 days, by the kind of email
- Promotional
- Educational
- Transactional and win-back
Email engagement
Deepest engagement in the last 30 days
- Clicked4%
- Opened26%
- Not opening34%
- Not emailed36%
Send mix
Share of about 2M emails sent in 30 days
- Promotional70.5%
- Educational29.3%
- Transactional0.1%
- Winback0.0%
Deliverability
Per email sent, last 30 days. The marker is the limit.
- Spam complaints0.03%within limitAbout 500 complaints
- Bounces0.18%within limitAbout 3.5K bounces
- Unsubscribes0.21%within limitAbout 4K unsubscribes
Who must be suppressed?
Subscription status, cancellation reasons and failed charges: conditions a policy can exclude on.
Subscriptions
Subscribers by status
- Active19%
- Paused1%
- Cancelled80%
Why subscribers cancel
Reason given at the latest cancellation
- Other reason47%
- I already have more than I need25%
- This was created by accident10%
- I no longer use this product8%
- This is too expensive6%
- Passive churn (billing)4%
Failed subscription charges
Last error, customers with a failed charge
- Payment method not found66%
- Closed max retries reached8%
- Invalid payment method7%
- Insufficient funds6%
- Payment method declined5%
- Other7%
All figures shown are sample data.