{"id":14351,"date":"2019-06-25T11:17:49","date_gmt":"2019-06-25T09:17:49","guid":{"rendered":"https:\/\/www.dase-analytics.com\/blog\/?p=14351"},"modified":"2019-06-25T13:04:07","modified_gmt":"2019-06-25T11:04:07","slug":"jim-sterne-niektore-krajiny-su-v-analytike-lepsie-napriklad-slovensko","status":"publish","type":"post","link":"https:\/\/www.dase-analytics.com\/blog\/sk\/jim-sterne-niektore-krajiny-su-v-analytike-lepsie-napriklad-slovensko\/","title":{"rendered":"Jim Sterne: Niektor\u00e9 krajiny s\u00fa v analytike \u201clep\u0161ie\u201d. Napr\u00edklad Slovensko"},"content":{"rendered":"<p>Jim Sterne je medzin\u00e1rodn\u00fd re\u010dn\u00edk v oblasti digit\u00e1lneho marketingu a interakcie so z\u00e1kazn\u00edkmi. Je konzultant pre spolo\u010dnosti a podnikate\u013eov z Fortune 500, kde vyu\u017e\u00edva svoje 30-ro\u010dn\u00e9 sk\u00fasenosti z oblasti predaja a marketingu.<\/p>\n<p>Jim je zakladate\u013eom medzin\u00e1rodn\u00e9ho summitu eMetrics <a href=\"http:\/\/www.emetrics.org\" target=\"_blank\" rel=\"noopener\">www.emetrics.org<\/a> a je spoluzakladate\u013eom a s\u00fa\u010dasn\u00fdm predsedom Zdru\u017eenia digit\u00e1lnej analytiky <a href=\"http:\/\/www.DigitalAnalyticsAssociation.org\" target=\"_blank\" rel=\"noopener\">www.DigitalAnalyticsAssociation.org<\/a>.<\/p>\n<p>Bol menovan\u00fd ako jeden z 50 najvplyvnej\u0161\u00edch \u013eud\u00ed v oblasti digit\u00e1lneho marketingu spolo\u010dnos\u0165ou Revolution a rovnako za jedn\u00e9ho z 25 najlep\u0161\u00edch re\u010dn\u00edkov.<\/p>\n<p>Preto sa ve\u013emi te\u0161\u00edme, \u017ee sme mu mohli polo\u017ei\u0165 zop\u00e1r ot\u00e1zok. Pod slovensk\u00fdm prekladom n\u00e1jdete aj anglick\u00fd origin\u00e1l.<\/p>\n<p><img decoding=\"async\" loading=\"lazy\" class=\"aligncenter wp-image-14352 size-full\" src=\"https:\/\/www.dase-analytics.com\/blog\/wp-content\/uploads\/Jim_Sterne_09a.jpg\" alt=\"\" width=\"1000\" height=\"1000\" srcset=\"https:\/\/www.dase-analytics.com\/blog\/wp-content\/uploads\/Jim_Sterne_09a.jpg 1000w, https:\/\/www.dase-analytics.com\/blog\/wp-content\/uploads\/Jim_Sterne_09a-150x150.jpg 150w, https:\/\/www.dase-analytics.com\/blog\/wp-content\/uploads\/Jim_Sterne_09a-300x300.jpg 300w\" sizes=\"(max-width: 1000px) 100vw, 1000px\" \/><\/p>\n<p><strong>&#8222;Data-driven&#8220; je v s\u00fa\u010dasnosti jedn\u00fdm z najpou\u017e\u00edvanej\u0161\u00edch slov v biznis sf\u00e9re. Ako by ste definovali, \u010di je ur\u010dit\u00e1 spolo\u010dnos\u0165 zalo\u017een\u00e1 na d\u00e1tach alebo nie?<\/strong><\/p>\n<p>V\u0161etko je to ot\u00e1zka spr\u00e1vania. Ak sa rozhodnutia prij\u00edmaj\u00fa HIPPO (Highest Paid Person Opinion\/N\u00e1zor osoby, ktor\u00e1 plat\u00ed najviac) alebo t\u00ed, ktor\u00ed kri\u010dia najhlasnej\u0161ie, potom o d\u00e1tach nem\u00f4\u017ee by\u0165 re\u010d. Na druhej strane, \u010disto data-driven firma prehr\u00e1 kv\u00f4li nedostatku citu a jemnosti. D\u00e1vam prednos\u0165 informovan\u00fdm spolo\u010dnostiam, ktor\u00e9 robia v\u00fdskum, prin\u00e1\u0161aj\u00fa v\u00fdsledky, automatizuj\u00fa \u010do sa d\u00e1, a potom robia intuit\u00edvne rozhodnutia podlo\u017een\u00e9 d\u00e1tami.<\/p>\n<p><strong>Cestujete ve\u013ea. M\u00f4\u017eete porovna\u0165 analytick\u00fa vyspelos\u0165 v r\u00f4znych krajin\u00e1ch?<\/strong><\/p>\n<p>Komunity digit\u00e1lnej analytiky s\u00fa dnes v\u0161ade. Najjednoduch\u0161\u00edm sp\u00f4sobom, ako ich n\u00e1js\u0165, je pozrie\u0165 si zoznam na <a href=\"https:\/\/measurecamp.org\/measurecamp-calendar\" target=\"_blank\" rel=\"noopener\">https:\/\/measurecamp.org\/measurecamp-calendar<\/a>.<\/p>\n<p>Existuje p\u00e1r kampan\u00ed v ur\u010dit\u00fdch firm\u00e1ch, ktor\u00e9 vynikaj\u00fa analytikou. Av\u0161ak neexistuj\u00fa firmy, ktor\u00e9 by robili skvel\u00fa analytiku celkovo. V\u00fdborn\u00e9 pr\u00edpadov\u00e9 \u0161t\u00fadie n\u00e1jdeme v ka\u017edej krajine.<\/p>\n<p>Niektor\u00e9 krajiny s\u00fa trochu pozadu kv\u00f4li kult\u00fare, ktor\u00e1 sa vyh\u00fdba experimentovaniu a riziku. Vo v\u0161eobecnosti je v\u0161ak skvel\u00e1 digit\u00e1lna analytika roz\u0161\u00edren\u00e1 po celom svete. Niektor\u00e9 krajiny s\u00fa \u201elep\u0161ie\u201c ako in\u00e9 (alebo aspo\u0148 vidite\u013enej\u0161ie), ako napr\u00edklad USA, Spojen\u00e9 kr\u00e1\u013eovstvo, Slovensko, Rumunsko a Nemecko. Ale to je vec poh\u013eadu, nem\u00e1m to podlo\u017een\u00e9 d\u00e1tami. \ud83d\ude09<\/p>\n<p><strong>Kde vid\u00edte poz\u00edciu machine learning (strojov\u00e9ho u\u010denia) a AI (umelej inteligencie) v oblasti digit\u00e1lnej anal\u00fdzy?<\/strong><\/p>\n<p>Machine learning (ML) n\u00e1m pon\u00faka nov\u00e9 n\u00e1stroje a mo\u017enosti. Av\u0161ak rovnako ako kladivo a p\u00edla z v\u00e1s nesprav\u00ed tes\u00e1ra, ML z v\u00e1s neurob\u00ed d\u00e1tov\u00e9ho vedca a dokonca ani lep\u0161ieho analytika. ML preberie ve\u013ea pr\u00e1ce, ktor\u00fa dnes rob\u00edme manu\u00e1lne, je nevyhnutn\u00e1, av\u0161ak ne\u00fa\u010dinn\u00e1. Napr\u00edklad \u010distenie a integr\u00e1cia \u00fadajov, pokro\u010dil\u00e1 \u0161tatistick\u00e1 anal\u00fdza, segment\u00e1cia, klastrovanie a pod. Rovnako ako n\u00e1m tabu\u013eky <em>(Google Spreadsheet, pozn.redaktora)<\/em> poskytli nov\u00fd sp\u00f4sob pr\u00e1ce s \u010d\u00edslami, tak n\u00e1m ML d\u00e1va nov\u00fd sp\u00f4sob, ako sa hra\u0165 s d\u00e1tami.<\/p>\n<p><strong>Kde vid\u00edte bud\u00facnos\u0165 pr\u00e1ce v oblasti digit\u00e1lnej anal\u00fdzy? Ak\u00e9 zru\u010dnosti by si mal osvoji\u0165 za\u010d\u00ednaj\u00faci v\u00e1\u0161niv\u00fd analytik?<\/strong><\/p>\n<p>Analytik bude v\u017edy potrebova\u0165 urobi\u0165 tri veci:<\/p>\n<p>1. Rozhodn\u00fa\u0165, ak\u00fd probl\u00e9m potrebujete vyrie\u0161i\u0165 alebo ak\u00fa ot\u00e1zku polo\u017ei\u0165. To si vy\u017eaduje ve\u013ea poznatkov o biznise. Sk\u00fasenosti s\u00fa d\u00f4le\u017eit\u00e9! Ak\u00e9 ot\u00e1zky sa oplat\u00ed kl\u00e1s\u0165?<\/p>\n<p>2. Rozhodn\u00fa\u0165 sa, ktor\u00e9 \u00fadaje m\u00f4\u017eu by\u0165 najviac predikt\u00edvne a ako ich spracova\u0165. St\u00e1le existuj\u00fa n\u00e1klady na v\u00fdpo\u010dtov\u00fa techniku \u200b\u200ba &#8211; ako povedal Matt Gershoff &#8211; n\u00e1klady na istotu. Na vyrie\u0161enie probl\u00e9mu mo\u017eno nebudete potrebova\u0165 ML. Mo\u017eno budete potrebova\u0165 len jednoduch\u00fa \u0161tatistick\u00fa anal\u00fdzu. Analytik sa mus\u00ed rozhodn\u00fa\u0165, ak\u00e9 d\u00e1ta a ko\u013eko v\u00fdpo\u010dtovej sily je ochotn\u00fd vy\u010dleni\u0165.<\/p>\n<p>3. Rozhodn\u00fa\u0165, \u010di je model u\u017eito\u010dn\u00fd. &#8222;V\u0161etky modely s\u00fa nespr\u00e1vne, niektor\u00e9 modely s\u00fa u\u017eito\u010dn\u00e9&#8220; &#8211; George Box. Toto je test, ktor\u00fd vy\u017eaduje \u201csedliacky rozum\u201d a v\u0161eobecn\u00e9 znalosti. Ak po\u017eiadate stroj, aby zv\u00fd\u0161il engagement (zapojenie u\u017e\u00edvate\u013eov), zist\u00ed, \u017ee najlep\u0161\u00ed sp\u00f4sob, ako prin\u00fati\u0165 nov\u00fdch u\u017e\u00edvate\u013eov, aby reagovali, je posla\u0165 100 e-mailov za min\u00fatu 10 dn\u00ed po sebe. Z poh\u013eadu \u201csedliackeho rozumu\u201d to v\u0161ak ned\u00e1va zmysel!<\/p>\n<p><strong>\u010co sa v\u00e1m najviac p\u00e1\u010di na va\u0161ej pr\u00e1ci analytika a re\u010dn\u00edka?<\/strong><\/p>\n<p>M\u00e1m r\u00e1d cestovanie, stret\u00e1vanie sa s nov\u00fdmi \u013eu\u010fmi, dlhodob\u00fd kontakt s nimi a poznanie, \u017ee (niekedy) m\u00f4\u017eem pom\u00f4c\u0165 niekomu dosiahnu\u0165 viac, ne\u017e si myslel, \u017ee by mohl. To ma ve\u013emi nap\u013a\u0148a.<\/p>\n<h2>Jim Sterne interview<\/h2>\n<p><strong>Being data-driven is one of the most used buzz-words in the business sphere nowadays. How would you define if a certain company is data-driven or not?<\/strong><\/p>\n<p>It&#8217;s all a matter of behaviour. If decisions are made by the Hippo (HIghest Paid Person&#8217;s Opinion) or by the guy who shouts the loudest, then data isn&#8217;t even in the room. On the other hand, a purely data-driven company will lose the game due to a lack of nuance and finesse. I prefer data-informed companies that do their research, bring the results together, automate what they can, and then use data-informed intuition to make decisions.<\/p>\n<p><strong>You travel a lot, can you compare digital analytics maturity level in different countries?<\/strong><\/p>\n<p>Digital analytics communities are found all over. The easiest place to find a list is <a href=\"https:\/\/measurecamp.org\/measurecamp-calendar\" target=\"_blank\" rel=\"noopener\">https:\/\/measurecamp.org\/measurecamp-calendar<\/a>. While there are some campaigns in some companies that are doing amazing analytics, there are no companies that are doing a great job overall. There are great case studies in every country.<br \/>\nSome countries are a bit behind due to a culture that shuns experimentation and risk, but generally, great digital analytics is spread all over. Some countries are &#8218;better&#8216; than others &#8211; at least more visible &#8211; like the US, the UK, Slovakia, Romania, and Germany. But that is a perceptual difference &#8211; not one backed by data \ud83d\ude09<\/p>\n<p><strong>Where you see the position of machine learning and AI within the digital analytics industry?<\/strong><\/p>\n<p>Machine learning (ML) offers us new tools. Just as a hammer and a saw do not make you a carpenter, ML doesn&#8217;t make you a data scientist, or even a better analyst. ML is going to take on a lot of the grunt work that we have to do today that is currently necessary, but not impactful: data cleansing and integration, advanced statistical analysis, segmentation and clustering, etc. Just as the spreadsheet gave us a whole new way to play with numbers, ML gives us a new way to play with data.<\/p>\n<p><strong>Where do you see the future of digital analytics job? What \u201cmust-have\u201d skills to learn you would advise for new passionate digital analysts?<\/strong><\/p>\n<p>The analyst will always be needed to do three things:<\/p>\n<p>1. Decide what problem to solve or question to answer. That takes a good deal of domain knowledge. Experience counts! What questions are worth asking?<\/p>\n<p>2. Decide what data might be the most predictive and feed it to the machine. There is still a cost to computing and &#8211; as Matt Gershoff would say &#8211; a cost to certainty. You may not need ML to solve a problem. You might only need a simple statistical analysis. The analyst has to decide which data and how much compute power to bring to bear.<\/p>\n<p>3. Decide is the model is useful. &#8222;All models are wrong, some models are useful&#8220; &#8211; George Box. This is the Smell Test that requires common sense and general knowledge. If you ask the machine to improve prospect engagement, it will figure out that the best way to get strangers to respond to you is to send 100 emails a minute for ten days in a row. From a Smell Test perspective, that stinks!<\/p>\n<p><strong>What do you like the most in your job as an analyst as well as a speaker?<\/strong><\/p>\n<p>I enjoy the travel, meeting new people, staying in touch with them over a long period of time, and the knowledge that &#8211; sometimes &#8211; I can help somebody achieve more than they thought they could. THAT is satisfying!<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Jim Sterne je medzin\u00e1rodn\u00fd re\u010dn\u00edk v oblasti digit\u00e1lneho marketingu a interakcie so z\u00e1kazn\u00edkmi. Je konzultant pre spolo\u010dnosti a&#8230;<\/p>\n","protected":false},"author":66,"featured_media":14353,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[638],"tags":[],"_links":{"self":[{"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/posts\/14351"}],"collection":[{"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/users\/66"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/comments?post=14351"}],"version-history":[{"count":7,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/posts\/14351\/revisions"}],"predecessor-version":[{"id":14362,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/posts\/14351\/revisions\/14362"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/media\/14353"}],"wp:attachment":[{"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/media?parent=14351"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/categories?post=14351"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dase-analytics.com\/blog\/sk\/wp-json\/wp\/v2\/tags?post=14351"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}