EdgeKit Personalization
A front page for every reader, steered by your editors
Ranks the stories each reader is most likely to read next, on your homepage, under every article and in the newsletter. Built so one read leads to another, and readers come back and subscribe.
Signed in · reads Markets, Policy
The LedgerMarketsPolicyCulture
For youRanked for this reader
Anonymous · reads Culture
The LedgerMarketsPolicyCulture
For youRanked for this reader
The gap
One homepage, edited for everybody, read by one person at a time
Homepage
TodayThe same front page for the reader who follows Markets and the one who came for Culture.
With PersonalizationYour editors keep the lead. The boxes around it are ranked for whoever is reading, so the next story is one they want.
Under the article
TodayA read-next box that knows the article and nothing about the reader.
With PersonalizationRead next starts from the story on screen and ranks for the reader on it, so a visit does not end at the last paragraph.
Newsletter
TodayOne edition, the same stories, for every subscriber on the list.
With PersonalizationEach recipient’s For you slot fills at send, so every edition gives them a reason to click back to your website.
What it does
Recommendations that know the reader, and answer to the newsroom
Every placement
Each box on the page has its own logic
For you on the homepage. More like this under the article. Trending in a section. Editors’ picks. Each placement has its own logic and its own number of stories, and every one updates as the reader reads.
- Read next takes the article on screen as its starting point.
- Pages rendered on your servers and pages built in the browser both ask for it, from your own domain.
- If the engine does not answer in time, your website shows its own fallback.
Editorial control
The model ranks. Your editors decide what is eligible.
Every placement carries filters and boosters your editors set: pin a story, boost a section, favour fresh reporting. The newsroom’s judgement comes first, and the ranking works inside it.
- Your team sets each placement’s rules in the engine’s own admin, and changes them whenever the newsroom needs to.
- Only published stories reach the engine. Drafts and scheduled stories never leave, and a deleted story is never recommended again.
- It chooses which stories to show, never who may read them. A recommended paid story meets your paywall like any other read.
Rules
- AlwaysPublished to The Ledger
- AlwaysPublished stories only
- FilterPublished in the last 7 days
- FilterMarkets and Policy only
- BoostMarkets
- BoostFresher first
- PinTariff ruling: what changes on Monday → 1st
This reader sees
- PolicyTariff ruling: what changes on MondayPinned
- MarketsBond yields jump as the ruling lands
- MarketsPort deal signed after a six-month standoff
- MarketsWhy the banks are hoarding cash
- PolicyElection funding: the new cap, explained
In the newsletter
One edition, a different ‘For you’ for every reader
Add a personalised feed to a Publications template and each recipient’s slot fills at send with their own stories, ranked by the same engine that ranks your homepage. Your editors build the rest of the edition as they always do.
- Filled per recipient through Braze, from what that reader has read on your website.
- A feed can name the same placement as your homepage, so its filters and boosters carry over.
- The HTML, plain-text and AMP versions of one email show the same stories.
To: LenaThe Morning Brief · Mon 8 June
For you Election funding: the new cap, explained
To: TomThe Morning Brief · Mon 8 June
For you Paris shows: the five collections that mattered
To: AmaraThe Morning Brief · Mon 8 June
What the tariff ruling changes from today
For youPersonalised feed
Unsubscribe · View in browser
One reader
Recommendations from the first read, signed in or not
A reader who has not signed in gets recommendations from what they have read on that device. When they sign in, that history moves to their account, so their phone, their laptop and their newsletter all draw on the same reading.
- Learned from the events EdgeKit already collects on your domain, not from a second tracking tag, and updated on every read.
- For signed-in readers, profile fields you choose from one title travel with them, so a placement can use what they have told you.
Phone · not signed in
Laptop · signed in
Friday’s Morning Brief
Also in it
Built for a newsroom, not a shop window
Clicks it learns from
Every recommended link carries its recommendation, on the page or in the inbox, so the model learns what was clicked.
One engine for the group
A signed-in reader’s reading on any of your titles shapes what each title recommends to them.
Nothing in the reader’s browser
No engine tag on your pages, and its key never reaches a browser. Pages ask EdgeKit, on your own domain.
How it works
Your stories and events in, ranked stories out
White-glove onboarding: we set it up with you, start to finish.
- 01
Your stories sync
EdgeKit sends what it already syncs from your CMS: headline, summary, image, link, sections, tags and authors.
- 02
Your events teach it
The article views and sign-ins you already send are mapped once, so the engine knows what each reader read. Purchases teach it what converted once your payment event names the article the checkout started from.
- 03
Your team sets the placements
In your organization’s engine account, your team defines each box once: its logic, its filters, its boosters.
- 04
Your website draws the boxes
Your website asks for a placement and gets back stories ready to render. The newsletter asks for the same placements through Braze.
One request per box. Your website asks for a placement and how many stories it holds. Each story comes back with its headline, summary, image, link and publish time, ready to draw in your own design.
Tariff ruling: what changes on Monday
More like this
- Bond yields jump as the ruling landsTraders priced the ruling in before noon.2 h ago
- Inside the new rules on market dataWhat the regulator now asks of every exchange.5 h ago
- Election funding: the new cap, explainedParties face a lower limit from January.Yesterday
Beside what you run
It adds a ranking, not a platform
Your CMS, your website and your email platform stay where they are. Personalization reads stories from your CMS, answers your website and Braze, and never decides access: that stays with whichever paywall you run, EdgeKit’s or your own.
In
Your CMSPublished stories onlyYour domainArticle views and sign-insReader profileFields you choose, signed-in readersOut
HomepageFor you · Trending · Editors’ picksArticle pageMore like thisNewsletterFor you, through Braze- CMS
- Stays. Editors keep working where they work; EdgeKit reads what they publish.
- Website
- Stays. It renders the stories EdgeKit returns, in your own design.
- Marketing stack
- Stays. The personalised slot fills inside Braze at send.
- Paywall
- Still decides who may read each recommended story, whether it is EdgeKit’s or yours.
Questions
What a head of digital asks first
Can editors overrule it?
Yes. They pin a story and boost a section or fresh reporting. The model ranks; your rules decide what is eligible.
Does it decide who can read a story?
No. It chooses which stories to show, never who may read them. A click on a recommendation goes through your paywall’s rules like any other read.
Does it work for readers who never sign in?
Yes. A reader who has not signed in gets recommendations from what they have read on that device. When they sign in, that history moves to their account.
Can we take it without the rest of EdgeKit?
Yes. It needs EdgeKit’s event collection on your domain, which is what it learns from, and the content sync from your CMS. Recommendations that follow a reader across devices and into the newsletter also need EdgeKit ID, so a signed-in reader’s reading joins one account. Your CMS and your email platform stay as they are, and access stays with your paywall.