The leak

Where the second purchase leaks.

Returning shoppers are treated like strangers, so the second purchase never gets asked for. Retention stays the least funded line in the budget even though it is the cheapest revenue on the table.

Personalization built on behavioral signals typically accounts for 12-31% of an e-commerce company's total revenue. Barilliance

5-25x
Cheaper to retain a customer than acquire a new one McKinsey
60-70%
Repeat purchase probability, against 5-20% for a new prospect McKinsey
5.2x
Average return on retention programs McKinsey

Coverage

Three of the eight algorithms work the next purchase.

All eight are live from day one across product, category, home, search and cart. These three carry the shopper who has already bought from you.

Returning and mid-session

Recently Viewed

A deterministic recall of the shopper's own view history by recency, held across the session and across return visits. It removes the work of re-finding a product and reinforces what is already under consideration, as a persistent strip, as the homepage on a return visit, and as a cart-page reminder of considered-but-not-added items.

Post-purchase

Replenishment

A predictive timing model for consumables and repeat-purchase products. It estimates each shopper's own reorder interval from purchase history and product-level consumption patterns, then surfaces the reorder prompt as the predicted depletion date approaches, on the homepage, on order history, on the product pages for what they bought, and as a one-click add in the cart.

Cart abandonment

Abandon Cart Triggers

An event-driven recommender on real-time abandonment signals: an add to cart followed by exit intent, session inactivity, or a session that ends without checkout. It re-presents the cart plus relevant complements while the shopper is still on the site, and offers a resume-checkout entry point if they land again with the cart still open.

Cart recovery

The most recoverable moment in the funnel.

A populated cart plus exit intent, inactivity or an unfinished checkout is the strongest signal a shopper gives you. Fulcrum acts on it where the shopper still is, on the site and inside the session.

56.5%
Purchase rate on a cart exit prevention campaign One Fulcrum client engagement, validated across 1,127 users
$210,089
Revenue recovered, from visitors who were leaving the site The same
engagement
+2.3%
Conversion on an exit-intent recovery campaign Another engagement, against a holdout, 3,401 users

Beyond the suite

A customer you already have is a customer you can reach.

A customer has resolved identity by definition, which is what puts the lifecycle triggers here rather than earlier in the funnel. Email personalization runs on the same behavioral profile as the on-site suite, as a separate capability alongside it, and a loyalty tier is a targeting rule on an on-site placement.

After the order

Post-purchase cross-sell

An order is a trigger. Fulcrum builds the post-purchase audience from it and fills the recommendation zones inside the email your own service provider sends. The content resolves when the customer opens it rather than when you sent it, so what they see reflects the order they just placed and what is available now.

Lapsed

Win-back

A customer who has stopped coming back still has a behavioral history. The win-back audience is built from it, and the email a lapsed subscriber opens is personalized from their own profile rather than merchandised once for the whole list.

Loyalty tiers

VIP merchandising

A placement can be scoped to a loyalty tier. The pattern Fulcrum deploys is a VIP-only carousel: the same personalization engine and the same catalog, restricted so the customers in the tier see merchandising the rest of the site does not.

Proof

Repeat purchase, against your own control group.

Every reported lift comes from a randomized holdout: a fixed 10% of traffic that never sees personalization for the whole contract term. Lift is the difference between the groups, calculated once at term end and reported conservatively from the lower bound.

Supplements

National vitamin and supplement retailer

+29%
Repeat-buyer
rate vs. control
group
+13%
Average order
value vs. control
group
-9%
Bounce
rate vs. control
group

Home and gift

Home and gift retailer

+269%
Revenue per user, returning visitors vs.
holdout
+187%
Revenue per user, independent re-run vs. holdout, larger
group
24.5%
Purchase rate, returning-shopper recommendations Returning
shoppers

Memorabilia

School memorabilia and gifting brand

2.2x
Purchase rate, engaged vs. not engaged Returning
shoppers

Questions

What buyers ask about the next purchase.

Does a customer have to log in before you recognise them?

No. Fulcrum works from behavior: what a shopper is doing and where they sit in the journey, not whether they have identified themselves. A returning shopper is recognised from their own behavioral profile, which persists across the session and across return visits.

How does Fulcrum know when someone is due to reorder?

Replenishment estimates each shopper's own reorder interval from their purchase history and product-level consumption patterns, then surfaces the prompt as the predicted depletion date approaches. It appears on the homepage and order history on a return visit, on the product and category pages for what they already bought, and as a one-click add in the cart.

What happens when a customer abandons a cart?

The recovery happens on the site, in the session. Abandon Cart Triggers fire on an add to cart followed by exit intent, on session inactivity, or on a session that ends without checkout, and they re-present the cart plus relevant complements. If the shopper lands again with the cart still open, the homepage offers a resume-checkout entry point.

Are the recommendations recomputed, or set once at onboarding?

Not all of them. Six are rebuilt from scratch every night, so what is served today reflects yesterday's purchases and today's catalog and pricing. Recently Viewed reflects a shopper's own browsing history directly, so there is nothing to recompute, and Abandon Cart Triggers are evaluated at session boundaries rather than on the nightly cycle.

How much of this does my team have to run?

None of it. Fulcrum connects to your catalog and purchase graph, configures every algorithm against it, and shows you the placements before anything goes live. After that the suite recomputes nightly and adjusts in real time as behavior comes in. Nobody on your team operates it.

What does the guaranteed figure actually measure?

Holdout lift: the difference in revenue per user between your personalized traffic and a randomized 10% holdout fixed for the life of the contract. It covers the whole order value rather than only the recommended product, with no cap per customer and no de-duplication, calculated once at the end of the term and reported conservatively from the lower bound.

Test for free for 45 days. Fully managed by Fulcrum.

Test before you invest. Fulcrum provides your company with a fully provisioned and fully managed service for 45 days. Cancel any time, nothing owed.

Start Your Free Pilot