SolutionsA media group
Sell for lifetime value, not this month’s conversions.
The board now asks what a reader is worth, while the wall and the promo still chase this month’s conversions. EdgeKit puts the score from your own lifetime value model where the offer is decided.
Same promo, two readers
- Arrived from
- Search
- Reads, last 30 days
- 2
- Bought at the introductory price
- Left at the first full price
- Your score
- value_band · Low
- The offer a rule shows
- Monthly, introductory price
- Arrived from
- Newsletter
- Reads, last 30 days
- 18
- Bought at the introductory price
- Renewed at full price
- Your score
- value_band · High
- The offer a rule shows
- Annual, no discount
The situation
What is going wrong
Everyone agrees lifetime value is the number, and none of the decisions that set it can read it.
The promo wins the month
The deepest discount brings the most starts, so it wins every review held in the month it ran. The readers it brought leave at the first full-price renewal, months later, in somebody else’s report.
What a reader is worth is a spreadsheet
Lifetime value is worked out by finance, for the whole base, a year behind. The wall and the renewal never see it, so at the moment each is decided every reader is worth the same.
The signals live in other tools
How often a reader comes back, how they joined and what they hold each say whether they will stay, and each sits in a different system from the invoice, under a different id.
What changes
When the offer reads what a reader is worth
EdgeKit Paywall
Your value score is a fact a rule reads
Send the score your own model computes, a predicted lifetime value or a churn risk, and a rule reads it on every request from a signed-in reader, beside how they joined, what they hold and how often they read. The reader your model ranks high meets the annual offer and the occasional one the monthly, with no ticket. EdgeKit computes no lifetime value, and a reader who has not signed in is decided without your score.
- Your own score reaches EdgeKit through one endpoint, sent once for each title, and a rule reads it today
- How often a reader reads is counted for signed-in readers, once we switch it on for your titles
- The rule is live 60 seconds after you publish, and the version before is one publish away
- Metered reads, 30 days
- 4
- Membership
- Registered
- Signed up from
- Newsletter
- New device
- No
- RFV
- 5 · 4 · 3
- Your score
- 0.82
AI Paywall
- Likely to subscribe
- High
- Managed
- Churn risk
- —
- Managed
- Price sensitivity
- Low
- Managed
EdgeKit Analytics
How long each reader stayed, beside what they read
A signed-in reader’s reads reach your warehouse carrying their reader id, beside when and how they joined each title, what they hold, and when that access started and ends. Who still holds access, and for how long, by sign-up month or by how they joined, is a query your analysts write, and your model trains on the same rows and sends its score back under the same id. What a reader paid is not in that export; revenue per reader still comes from your processor’s reports.
- Every read on your domain, loaded into a warehouse table you query
- When and how each reader joined, and when their access started and ends, beside the reads
- Payments are not exported; your processor still reports revenue
select * from edgekit.events limit 5
5 rows · 2 titles · one table
EdgeKit Subscriptions
What they bought stays with what they did next
Through EdgeKit checkout, the offer a reader bought, its introductory price and its term sit on their record, and every renewal and pause after it is a dated row naming the provider it came from. What each invoice charged, after discounts, is held in the same database, so whether a promo reader paid their first full-price renewal sits on one record. That history stays in your database: a rule does not read which offer was bought, and none of it reaches your warehouse.
- The offer bought, its introductory price and its term, on the reader’s record
- Every renewal and pause is a dated row naming its source, beside what each invoice charged
- A rule does not read which offer was bought, and billing never reaches your warehouse
- Billing
- Your provider
- Sold as
- Individual
- Access to
- 3 Nov 2026
- StartedEdgeKit checkoutBought Digital Monthly at its introductory price on an article page
- ExtendedRecurlyFirst full-price renewal paid · access to 3 Sep 2026
- ExtendedRecurlyRenewal paid · access to 3 Oct 2026
- ExtendedRecurlyRenewal paid · access to 3 Nov 2026
Written from the provider's own events
EdgeKit Profile
One person across every title, in every query
A reader of three of your titles is one record, so every query your analysts write sees one person across the group, and each title’s rules read that record. How and when they joined each title, and the answers they gave at sign-up, sit on it from the first sign-in, before the first payment. Your score and how often they read are kept per title, so a score for the whole group is sent once for each title.
- One record per person across every title you publish
- How and when they joined each title, from the first sign-in
- Your score and how often they read are kept title by title
Registered 4 Mar 2024 · Email verified · Signs in with Google
Access · computed by EdgeKit
- Access
- Paid
- Products
- Markets Pro
- Titles they can read
- 2 of 3
- Access ends in
- 41 days
- Lapsed
- Never
On this title
- Joined
- 19 Sep 2025
- Joined through
- Newsletter
- Newsletters
- Morning Brief · Markets Close
Answered at sign-up
- Industry
- Financial services
- Job level
- Director
- Interests
- Markets · Policy
Your fields
- account_type
- Individual
- seniority
- Senior
How it goes live
Four steps, signed-in readers first
White-glove onboarding: we set it up with you, start to finish.
- 01
Bring the number finance already uses
Bring your definition of lifetime value, or the score your data team already computes, and whoever owns the offers. We go line by line through which system holds each signal today: the reads, the sign-ups, the renewals and the invoice.
- 02
One tag, one CMS integration and your score
We do the install and your developers review it. Your model sends each reader’s score to one endpoint on your schedule, under the reader id your warehouse already holds, once for each title. With no model of your own, EdgeKit AI Paywall is managed today: our data team builds likelihood to subscribe, churn risk and price sensitivity, and none of them is a lifetime value.
- 03
The first rule, for signed-in readers
Which signed-in reader meets the annual offer and which the monthly, by your score, written by your revenue team and tried on any reader before it ships. EdgeKit runs beside your current wall on one section first.
- 04
Read who stayed, then move the next title
Each decision returns the version the reader saw, and your page sends it with the reads. A renewal later, read who still holds access per version and per sign-up month in your warehouse, then move the next title.