The diagnostic proves it. The score runs it. The playbooks compound it.
Tacet is one engine sold as three products: the free Diagnostic on your own data, the Forward CLV subscription that lives inside Klaviyo, and the Playbooks the model personalizes to your customer base.
One engine, three products.
The Diagnostic proves the engine on your data. Forward CLV runs it daily. Playbooks compound it. Each product earns the next; nothing is sold on a promise the previous one didn't already prove.
Free
The Diagnostic
A forward-value audit of your customer base.
You give
One read-only connection, two weeks.
You get
Every customer scored, the written report, the model's published back-test, a walkthrough. Yours to keep either way.
Daily
Forward CLV
The score, live in your stack.
You give
A flat monthly fee, banded by customer count, month to month.
You get
Daily re-scores, named properties on every Klaviyo profile, new high-value customers flagged on entry, the monthly brief.
Per playbook
Playbooks
Measured interventions on the customer pools the model nominates.
You give
A per-playbook fee, experiment included.
You get
The flow in your Klaviyo, a pre-registered experiment, a written verdict. Nulls get retired.
Day 0
Connect read-only
One Shopify link. That is the whole diagnostic integration.
Day 3
Every customer scored
Forward value for the full book, net of returns and discounts.
Week 2
Report and walkthrough
The written diagnostic, plus the call that explains it.
From then on
Daily re-scores
Fresh scores on every Klaviyo profile as orders and returns land.
No engineering time on your side, at any point.
This is the diagnostic. Yours arrives on your data.
The free analysis returns a report like this: customer tiers, value concentration, named leak sources, and the model's own back-test on your history.
A forward CLV score for every customer
Not what a customer has spent, but what they will contribute, in dollars, net of returns and discounts. Scored the day they place their first order.
Where your value actually concentrates
Your book ranked by forward value, so you can see the thin top tier worth a multiple of the rest, and fund acquisition and retention toward it.
Where contribution leaks
The exact dollars lost to returns on high-AOV buyers, discount stacking on customers who'd pay full price, and win-back spend on customers who never come back.
The back-test, published
How well the model predicted a slice of your own history it never saw: aggregate revenue accuracy and value-ranking accuracy, before anything directs a dollar.
In your hands after the Diagnostic
- A written diagnostic report, yours to keep either way
- The model's back-test: two published accuracy numbers on your own history
- A walkthrough call with the modeling team behind your scoring
Forward value per customer, by tier
8% of customers
19% of customers
41% of customers
32% of customers
Where contribution leaks
Recommended reallocation
Move $310k from Tier 4 win-back into Tier 1 retention. Projected 12-mo net contribution +$1.2M to +$1.9M (95% CI).
Figures illustrative. Your diagnostic is computed on your customers, from your own order history.
The score, live in the stack you already run.
Forward CLV is a subscription that re-scores daily, not a dashboard. Your ESP already knows how to send; it doesn't know who is worth it. Tacet supplies that one missing input.
- Every customer re-scored daily as new orders, returns, and discounts land
- Named properties written to every Klaviyo profile, forward value and archetype, so your team can segment on them directly, like "Tacet · Top decile forward value"
- New high-value customers flagged the day they enter, not a year after the spend
- A monthly brief: score movement, new entrants, and one honest read on what changed
How it connects
Shopify
orders · returns · discounts · read-only
Tacet · Forward CLV
every customer scored daily, net of returns
Klaviyo
segments · profile properties · paid audiences
No new dashboard, no new build. In a recent controlled trial, the winning flow ran entirely inside the client’s existing account.
The month in numbers
The honest read
Value creation held steady. One watch item: new top-decile entrants slowed 8% month over month as acquisition mix drifted toward discount-led channels.
Segments refreshed in Klaviyo
Scores refresh daily; this brief lands monthly. Your flows and paid audiences pick the changes up automatically.
Figures illustrative. Your customers are scored daily; the brief is computed monthly.
The model nominates the playbook. The experiment renders the verdict.
Playbooks are not templates. Forward CLV identifies which pools of your customers can pay for an intervention, and every playbook ships with its own controlled experiment. The ones that read null get retired.
The flow. Built and live in your existing Klaviyo account
The experiment. Pre-registered against a holdout before launch
The verdict. A written read at the end: scale it, or retire it
Second Purchase
AVAILABLEHigh-value one-time buyers
Roughly four in five DTC customers buy exactly once, and a second purchase multiplies expected value five to nine times depending on category. This playbook spends only on the one-timers worth converting.
Get the free manualSubscription Conversion
AVAILABLERepeat buyers with subscription-shaped cadence
Converts your steadiest repeaters into committed recurring revenue, the most durable asset a DTC book can hold.
Sleeping Giants
AVAILABLEDormant high-value customers
Wakes the lapsed customers whose forward value justifies the spend, and lets the rest sleep.
High-Value Protection
AVAILABLEThe thin top tier
Defends the customers carrying most of your forward value before their churn shows up in a report.
Discount Weaning
AVAILABLEPromo-dependent repeaters
Recovers margin from customers trained to wait for a code, without losing them.
The library grows as the model surfaces new pools. Each playbook targets a non-overlapping set of customers, so the value stacks rather than competes.
Final read · week 8 of 8 · pre-registered
Cohort held constant · bootstrap intervals
95% CI +$0.60 to +$5.20 · vs control
The offer arm held a confidently positive lift across the full window. Recommendation: roll the flow to the full eligible pool. Next read in 90 days.
Prior variant · reminder without offer
$0.00 per customer · CI spans zero · all 8 reads
Every playbook ends in one of two words. You stop paying for the ones that read null.
How pricing works.
No quote arrives before the model has seen your book. But the shape of it is simple, and there are no surprises inside it.
The Diagnostic
How both sides find out whether there is enough value in your book to go further.
- One read-only connection
- The report, the back-test, the walkthrough
- Yours to keep, data deleted if we stop
Forward CLV
Banded by the size of your customer base.
- Month to month, no annual lock-in
- Daily re-scores, Klaviyo properties, and the monthly brief included
- The verdict renews the contract, not a term sheet
Playbooks
Experiment and verdict included in the fee.
- The flow, the experiment, the verdict, one price
- Nulls get retired: you stop paying for what doesn't work
- Sized to the pool at the walkthrough
Exact figures come with the diagnostic walkthrough, sized to your customer base.
Start with the Diagnostic.
The free diagnostic replaces every illustrative figure above with your customers' actual forward value, and you keep the report whether or not we go further.
Read-only access · No deck required · You keep the analysis either way