Pick Rebuy if the job is cart and post-purchase upsell on a Shopify theme and you can live with percentage-of-revenue pricing. Pick Nosto if you need onsite recommendations plus merchandising control for a catalog big enough to justify an enterprise contract. Pick Searchspring if search and category merchandising are the real problem and you want merchandiser-friendly rules. Pick Malachite if you want a newer personalization layer and are willing to test a less proven vendor instead of paying enterprise rates.
The short comparison
 Rebuy |  Nosto | Searchspring | Malachite | |
|---|---|---|---|---|
| Built for | Cart, PDP and post-purchase upsell on Shopify | Onsite recommendations, personalization and category merchandising | Site search, autocomplete and category sorting | Automated personalization and recommendations |
| What else it sells | Smart carts, widgets, email and SMS upsell blocks | Personalized content, segments, onsite experiences | Merchandising rules, recommendations | Recommendation and merchandising automation |
| Where it is strongest | Fast to launch widgets without dev work | Breadth across recommendations and merchandising in one contract | Rule building that a merchandiser can run alone | Promise of working without a long tuning cycle |
| Reason brands leave | Percentage-based pricing plus recommendations that stay generic | Price can scale with revenue, so growth raises the bill | Slow product movement and support that does not keep up | Too new to have a settled leaving pattern; evaluations stall |
| Typical switch | To a search-led tool or to custom recommendations | To a cheaper recommendation app or to in-house logic | To Algolia, Nosto or a newer search vendor | Usually arrives as the replacement, not the thing replaced |
Rebuy
Rebuy is a common pick when the brief is "add upsells to the cart this month." Widgets go live quickly, the Shopify integration is direct, and a merchandising team can run it without engineering time. That speed is real and it is why the app is widely installed.
Two things sour it. The first is pricing: a percentage charge on attributed revenue means the tool gets more expensive exactly as it works, and the attribution is the vendor's own. The second is recommendation quality. Rebuy leans on the data you map into it. If nobody maintains that mapping, results drift toward generic bestsellers, and widgets that sit on every template can pull conversion down rather than up. Measure the widgets against a clean holdout before you renew. Brands that outgrow Rebuy do not always replace it with another upsell app; the logic can move into the theme or into a recommendation engine already being paid for.
Baltic Born runs Nosto on its storefront, which is the alternative shape of this decision: one broader personalization contract instead of a stack of widgets.
Nosto
Nosto covers more ground than a cart app. Recommendations, onsite content personalization and category merchandising sit in one platform, which is why larger apparel and home brands end up there. Coming off a weak search or recommendation tool, it reads as a step up.
The friction is cost and fit. Nosto prices like enterprise software and can tie to revenue, so a growing brand sees the bill climb without a change in what it uses. Smaller catalogs rarely clear that bar. The other issue is implementation on headless or React storefronts, where the install stops being a script tag and starts being a project. Ask who owns that work before you sign.
Searchspring
Searchspring is a search and category merchandising tool first. Its appeal is that a merchandiser can build rules, pin products and control sort order without filing a ticket. For catalogs where the real loss is shoppers failing to find a product, that is the right shape of tool.
The complaints are consistent: the product has moved slowly, results feel one dimensional next to newer search vendors, and support has not kept pace. Brands tend to stay for years, get frustrated, then look at Algolia or a newer search product and decide based on price. Algolia carries its own trap, which is large contracts signed and then never properly configured, so the spend produces nothing. Whichever way you go, budget implementation time, not just license fees.
Malachite
Malachite is the newer name in this set and it comes up when a team is unhappy with what it has and unwilling to pay enterprise prices. The pitch is personalization that works without a long testing and tuning program. That is attractive after a year of maintaining rules by hand.
Treat it as a test, not a replacement decision. Evaluation here can stall because the promise is hard to verify quickly: if a tool claims it needs no testing, you still need a holdout to know whether it did anything. Run it on one surface, keep your existing setup live elsewhere, and set a date to decide.
How to decide
Questions people ask
Is Rebuy worth it if I already have Nosto?
Usually not. Nosto covers recommendations across the site, and adding Rebuy on top means paying two vendors for overlapping carousels. The exception is cart and post-purchase offers, where Rebuy's widget set is quicker to configure. If you do add it, limit it to those surfaces and test against a holdout.
Why do brands leave Searchspring?
Two reasons: the product has been slow to improve compared with newer search vendors, and support quality has slipped. Brands tend to tolerate it for a long time before a contract renewal forces a look at alternatives, then move to Algolia, Nosto or a smaller search specialist.
How should I price-compare personalization tools?
Normalize everything to annual cost at your projected revenue, not current revenue. Percentage-of-attributed-revenue pricing looks cheap at signing and expensive later. Then subtract the lift you can prove with a holdout. A tool that charges a flat fee and lifts less can still be the better deal.
