Doporu\u010dovat spr\u00e1vn\u00e9 produkty v ko\u0161\u00edku a u pokladny je kl\u00ed\u010dem k ziskov\u00e9mu e-shopu. Poj\u010fme se pod\u00edvat, jak z t\u00e9to funkce vyt\u011b\u017eit maximum. Existuj\u00ed t\u0159i hlavn\u00ed typy doporu\u010den\u00ed, o kter\u00fdch budeme mluvit. Nejprve doporu\u010den\u00ed zalo\u017een\u00e9 na historii, d\u00e1le doporu\u010den\u00ed zalo\u017een\u00e9 na ko\u0161\u00edku a nakonec doporu\u010den\u00ed zalo\u017een\u00e9 na trendech. Tyto t\u0159i typy doporu\u010den\u00ed spolu dob\u0159e funguj\u00ed a jejich kombinace poskytuje ty nejlep\u0161\u00ed v\u00fdsledky.<\/p>\n\n\n\n
DOPORU\u010cEN\u00cd ZALO\u017dEN\u00c9 NA HISTORII<\/p>\n\n\n\n
Doporu\u010den\u00ed zalo\u017een\u00e9 na historii je nejkomplexn\u011bj\u0161\u00ed a nejslo\u017eit\u011bj\u0161\u00ed, ale tak\u00e9 v\u00e1m d\u00e1v\u00e1 nejv\u011bt\u0161\u00ed p\u0159\u00edle\u017eitost z\u00edskat n\u00e1skok p\u0159ed konkurenc\u00ed. Tento typ analyzuje historii n\u00e1kup\u016f z\u00e1kazn\u00edka, v\u010detn\u011b pr\u016fm\u011brn\u011b utracen\u00e9 \u010d\u00e1stky, zakoupen\u00fdch produkt\u016f, opakovan\u00fdch objedn\u00e1vek, kategori\u00ed produkt\u016f a pr\u016fm\u011brn\u00e9ho po\u010dtu polo\u017eek v n\u00e1kupn\u00edm ko\u0161\u00edku. To p\u0159edstavuje spoustu dat, kter\u00e1 mus\u00ed predik\u010dn\u00ed algoritmus zpracovat. Algoritmus identifikuje n\u011bkolik produkt\u016f s nejvy\u0161\u0161\u00ed pravd\u011bpodobnost\u00ed n\u00e1kupu a doporu\u010d\u00ed je z\u00e1kazn\u00edkovi. Samotn\u00fd algoritmus je \u010dern\u00e1 sk\u0159\u00ed\u0148ka a jeho p\u0159esn\u00fd provoz je obvykle obest\u0159en tajemstv\u00edm. Pod\u00edvejme se v\u0161ak na principy, kter\u00e9 predik\u010dn\u00ed n\u00e1stroj pou\u017e\u00edv\u00e1.<\/p>\n\n\n\n
Nejprve algoritmus identifikuje jak\u00e9koliv opakuj\u00edc\u00ed se n\u00e1kupy a jejich frekvenci. Jakmile se bl\u00ed\u017e\u00ed \u010das dal\u0161\u00edho n\u00e1kupu, syst\u00e9m doporu\u010d\u00ed produkt.<\/p>\n\n\n\n
D\u00e1le se d\u0159\u00edve zakoupen\u00e9 produkty pou\u017e\u00edvaj\u00ed k nalezen\u00ed z\u00e1kazn\u00edk\u016f s podobn\u00fdm vkusem. Algoritmus pak hled\u00e1 opakovan\u011b nakupovan\u00e9 produkty podobn\u00fdmi z\u00e1kazn\u00edky a doporu\u010duje ty nejvhodn\u011bj\u0161\u00ed.<\/p>\n\n\n\n
Nakonec algoritmus posuzuje, zda je k d\u0159\u00edve zakoupen\u00e9mu produktu dostupn\u00fd upgrade nebo vhodn\u00e9 p\u0159\u00edslu\u0161enstv\u00ed a zda je pravd\u011bpodobn\u00fd n\u00e1kup takov\u00e9ho produktu.<\/p>\n\n\n\n
V\u00fdvoj algoritmu vy\u017eaduje mnoho zdroj\u016f a \u010dasu, ale i provoz algoritmu vy\u017eaduje vysok\u00fd v\u00fdpo\u010detn\u00ed v\u00fdkon. Proto v\u011bt\u0161ina e-shop\u016f vyu\u017e\u00edv\u00e1 extern\u00ed cloudov\u00e9 slu\u017eby, aby se vyhnula n\u00e1klad\u016fm na v\u00fdvoj a z\u00e1rove\u0148 udr\u017eela z\u00e1t\u011b\u017e serveru mimo samotn\u00fd e-shop.<\/p>\n\n\n\n
DOPORU\u010cEN\u00cd ZALO\u017dEN\u00c9 NA KO\u0160\u00cdKU<\/p>\n\n\n\n
Doporu\u010den\u00ed zalo\u017een\u00e9 na ko\u0161\u00edku je mnohem jednodu\u0161\u0161\u00ed, ale p\u0159esto poskytuje v\u00fdznamn\u00e9 v\u00fdsledky, kter\u00e9 udr\u017euj\u00ed e-shopy ziskov\u00e9. Jeho funkce je podobn\u00e1 hled\u00e1n\u00ed z\u00e1kazn\u00edk\u016f s podobn\u00fdm vkusem, ale zde se jedn\u00e1 o hled\u00e1n\u00ed podobn\u00fdch ko\u0161\u00edk\u016f. Identifikuje nej\u010dast\u011bj\u0161\u00ed produkty v podobn\u00fdch ko\u0161\u00edc\u00edch a doporu\u010duje nejlep\u0161\u00ed shodu.<\/p>\n\n\n\n
Velkou v\u00fdhodou tohoto typu doporu\u010den\u00ed je jeho jednoduchost. Algoritmus nemus\u00ed rozum\u011bt produktu nebo jeho vztahu k jin\u00fdm produkt\u016fm; pouze je seskupuje a doporu\u010duje chyb\u011bj\u00edc\u00ed polo\u017eky v aktu\u00e1ln\u00edm ko\u0161\u00edku. Je to nejlep\u0161\u00ed volba pro nov\u00e9 z\u00e1kazn\u00edky, kde prvn\u00ed algoritmus nem\u016f\u017ee generovat \u017e\u00e1dn\u00e1 doporu\u010den\u00ed.<\/p>\n\n\n\n
DOPORU\u010cEN\u00cd ZALO\u017dEN\u00c9 NA TRENDECH<\/p>\n\n\n\n
Doporu\u010den\u00ed zalo\u017een\u00e9 na trendech je nejjednodu\u0161\u0161\u00ed. Nem\u011blo by chyb\u011bt ve va\u0161em mixu doporu\u010den\u00ed, ale ani by nem\u011blo b\u00fdt jedin\u00fdm \u0159e\u0161en\u00edm. Efektivita tohoto doporu\u010den\u00ed spo\u010d\u00edv\u00e1 v pravidle 80\/20, kter\u00e9 \u0159\u00edk\u00e1, \u017ee p\u0159ibli\u017en\u011b 20 % va\u0161ich produkt\u016f generuje asi 80 % va\u0161ich zisk\u016f.<\/p>\n\n\n\n
To je nejlep\u0161\u00ed p\u0159\u00edstup pro nov\u00e9 z\u00e1kazn\u00edky s pr\u00e1zdn\u00fdm n\u00e1kupn\u00edm ko\u0161\u00edkem. Nejsou k dispozici \u017e\u00e1dn\u00e1 data, kter\u00e1 byste mohli pou\u017e\u00edt pro v\u00fd\u0161e zm\u00edn\u011bn\u00e9 algoritmy, tak\u017ee doporu\u010den\u00ed bestseller\u016f je va\u0161\u00ed nejbezpe\u010dn\u011bj\u0161\u00ed mo\u017enost\u00ed. Toto doporu\u010den\u00ed tak\u00e9 nejl\u00e9pe funguje na hlavn\u00ed str\u00e1nce nebo na za\u010d\u00e1tku kategorie, ale jeho u\u017eite\u010dnost se sni\u017euje v ko\u0161\u00edku nebo u pokladny. Tam doporu\u010dujeme pou\u017e\u00edvat pouze prvn\u00ed dva typy doporu\u010dovac\u00edch metod.<\/p>\n\n\n\n
Jak vid\u00edte, funkce doporu\u010den\u00ed je slo\u017eit\u00e1 a vy\u017eaduje zna\u010dn\u00e9 \u00fasil\u00ed pro spr\u00e1vnou implementaci. Nejlep\u0161\u00ed je si zvolit siln\u00e9ho partnera se zku\u0161enostmi a zdroji, kter\u00fd v\u00e1m pom\u016f\u017ee funkci implementovat. CareCloud poskytuje \u00fa\u010dinn\u00e1 doporu\u010den\u00ed ji\u017e testovan\u00e1 v mnoha odv\u011btv\u00edch a zem\u00edch. Nav\u00edc zku\u0161en\u00ed konzultanti porad\u00ed nejlep\u0161\u00ed kombinaci a um\u00edst\u011bn\u00ed typ\u016f doporu\u010den\u00ed p\u0159\u00edmo pro v\u00e1\u0161 business.<\/p>","protected":false},"excerpt":{"rendered":"
Recommending the right products in the cart and during checkout is the way to the profitable e-shop. Let’s look at how to get the most from this feature. There are three main recommendation types that we will be talking about. Firstly the history-based recommendation, secondly the cart-based recommendation, and thirdly the trend-based recommendation. Those three…<\/p>","protected":false},"author":3,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_wp_rev_ctl_limit":""},"categories":[12],"tags":[14,13,22],"class_list":["post-2916","post","type-post","status-publish","format-standard","hentry","category-tips-tricks","tag-e-commerce","tag-emailing","tag-retail"],"_links":{"self":[{"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/posts\/2916","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/comments?post=2916"}],"version-history":[{"count":12,"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/posts\/2916\/revisions"}],"predecessor-version":[{"id":3541,"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/posts\/2916\/revisions\/3541"}],"wp:attachment":[{"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/media?parent=2916"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/categories?post=2916"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.crmcarecloud.com\/cz\/wp-json\/wp\/v2\/tags?post=2916"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}