Archive for ‘Thema4 Intelligence de marché’

20 mai 2018

Second-generation p-values: Improved rigor, reproducibility, & transparency in statistical analyses

par marketingthema

Statisticians often try to come up with alternatives to p-values. Here’s a recent attempt called ‘second generation p-values’.

Verifying that a statistically significant result is scientifically meaningful is not only good scientific practice, it is a natural way to control the Type I error rate. Here we introduce a novel extension of the p-value—a second-generation p-value (pδ)–that formally accounts for scientific relevance and leverages this natural Type I Error control. The approach relies on a pre-specified interval null hypothesis that represents the collection of effect sizes that are scientifically uninteresting or are practically null. The second-generation p-value is the proportion of data-supported hypotheses that are also null hypotheses. As such, second-generation p-values indicate when the data are compatible with null hypotheses (pδ = 1), or with alternative hypotheses (pδ = 0), or when the data are inconclusive (0 < pδ < 1). Moreover, second-generation p-values provide a proper scientific adjustment for multiple comparisons and reduce false discovery rates. This is an advance for environments rich in data, where traditional p-value adjustments are needlessly punitive. Second-generation p-values promote transparency, rigor and reproducibility of scientific results by a priorispecifying which candidate hypotheses are practically meaningful and by providing a more reliable statistical summary of when the data are compatible with alternative or null hypotheses.

http://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0188299

Publicités
18 janvier 2018

Looking for new trends? #trendhunter is a great source to use for teaching and/or innovation

par marketingthema

TrendHunter.com is the world’s largest, most popular trend community. A great ressources if you want to better understand the key trends emerging and have new ideas. I use it as a ressource in my courses to help students choose innovations they can work on for their projects and marketing approach, and to spot new consumer behaviours. This also can inspire academics for some research topics in consumption phenomena. The marketing professionals can also find ideas for their business and monitor the market evolution. A great ressource, and above all it’s for free 😉

YB

14 avril 2017

Forrester vs Gartner on Data Science Platforms #Ranking

par marketingthema

Who leads in Data Science, Machine Learning, and Predictive Analytics? We compare the latest Forrester and Gartner reports for this industry for 2017 Q1, identify gainers and losers, and strong leaders vs contenders.

Last month a leading analyst firm Forrester released their « Forrester Wave™ »: A report on Predictive Analytics and Machine Learning Solutions for Q1 2017, written by Mike Gualtieri. Predictive Analytics and Machine Learning are among the most important technologies now, as KDnuggets readers no doubt know, and Forrester forecasts a 15% compound annual growth rate (CAGR) for the PAML market through 2021. The report examines and evaluates 14 firms in terms of strategy, current offering, and Market Presence. The results are summarized in Fig. 1.

Forrester Wave Big Data Predictive Analytics 2017
Fig. 1: Forrester Wave™: Predictive Analytics And Machine Learning Solutions, Q1 2017
The leaders:

By , KDnuggets.

Full article here

13 avril 2017

#Trending : Image Analytics & simple clustering using machine learning #Orange #Demo

par marketingthema

La classification des images via des algorithmes intelligents atteint des résultats de plus en plus satisfaisants. Nous vous présentons ici le cas d’une classification simple avec le logiciel orange datamining (open source) qui permet justement de faire un classement simple des animaux en se basant sur plus de deux cent critères générés depuis chaque photo automatiquement à travers l’usage de classes d’objets prédétérminées, et de la technologie machine learning. Ce type d’applications peut être enrichissant à appliquer pour des démarches de classification automatique de masse pour des projets de recherche marketing, netnographies, classification du type de partages et d’images pour tout ce qui est User Generated Content sur les réseaux sociaux, une classification des catalogues d’images des sites web commerciaux et l’étude de leur style graphique et ses effets sur les comportements, etc …. les applications sont théoriquement  infinies.

For source and additional detail visit the official website of Orange

 

8 février 2017

The Data Economy Is Going To Be Huge. Believe Me. #Forrester

par marketingthema

by Jennifer Belissent, Ph.D.

Are they serious? I’ve just finished reading the recent Communication on Building a European Data Economy  published by the European Commission. And, it’s a good thing they’re seeking advice. The timing is perfect. I’m in the thick of my research for a new report on data commercialization. When I first published It’s Time To Take Your Data To Market the idea was merely a twinkle in people’s eye. Today that twinkle is much brighter than it has been, and it’s growing fast…really, really fast. Believe me. It’s going to be huge.

From what we see and hear, companies know they are sitting on a wealth of valuable data. And they know they aren’t the only ones who can benefit from it. According to Forrester’s Business Technographics, one-third of firms reported selling their data in 2016, up from only 10% in 2014. And, that trend extends across industries, and companies of all sizes.

The Full paper here

7 février 2017

#BCG Making Sense of the Marketing Measurement Mess

par marketingthema

by Nicolas De Bellefonds, Dominic Field, David Ratajczak, Neal Rich, and Jody Visser

Winning means keeping score, and keeping score requires a scorecard. Yet in our experience, few marketing organizations are able to quantify and communicate their contributions to such critical corporate goals as raising revenue and increasing brand recognition and advocacy. It’s not for lack of actual success. Today’s paradox is that with an unprecedented amount of data, tools, and analytics at their disposal, marketers are finding it more and more difficult to demonstrate the value that they create.

Measuring the value of marketing has never been easy. And the digital revolution has compounded the complexity, drastically increasing the number of touch points between companies and consumers while raising the expectations of senior management, among others, that everything is measurable. New tools and techniques are invented daily—the number of marketing technology vendors, each with its own way of improving reach, engaging consumers, and measuring effectiveness, has exploded from 150 in 2011 to almost 4,000 today. With marketing budgets running into the billions, it is more important than ever to know—and be able to show—what’s working.

Marketing leaders who want to demonstrate marketing’s value should take a step back, look objectively at the tools and metrics in place, and ask a few simple questions:

  • Do the metrics and tools capture the short- and longer-term value of marketing?
  • Do they produce answers and insights that can be acted on?
  • Are they readily understood by and credible to the CEO, the CFO, and the broader organization?

If the answer to any of these questions is no, a reassessment is in order, but as they proceed, marketing leadership should avoid the temptation to aim for perfection rather than confidence. That is, instead of striving for the ideal, it is far more effective and practical to apply the necessary ­resources to develop and build consensus around a simple set of metrics and tools that do the job well and that will consistently demonstrate value and improvement. Here are five rules, based on our experience with hundreds of marketing organizations in many different industries worldwide, that help link measurement to real business outcomes.

The full paper can be found here

28 janvier 2017

Marketing politique : Démocra-ciblée #DATAGUEULE 68

par marketingthema

4 novembre 2016

Étude : le potentiel de la publicité programmatique à la télévision

par marketingthema

La publicité programmatique, autrefois réservée au web via les écrans d’ordinateur et les mobiles, est en passe d’arriver à la télévision. Pour évaluer les possibilités offertes par la publicité TV programmatique dans les 10 prochaines années, Enders Analisys a réalisé une étude pour le compte de Google/DoubleClick. Elle s’intéresse au potentiel de ce type d’annonces à la télévision dans plusieurs pays européens dont la France. Les principaux enseignements de l’étude Enders Analysis sont les suivants.

Les tendances de la consommation TV en Europe

L’institut a analysé les habitudes des européens vis-à-vis de la télévision. Sans surprise, on assiste à une baisse globale de la consommation TV, notamment pour les tranches d’âge les plus jeunes. Leur consommation vidéo est beaucoup plus fragmentée entre les différents terminaux et les différentes sources de contenu. Comme il est difficile d’atteindre les jeunes populations à la TV, les prix des publicités qui ciblent les jeunes ont fortement augmenté – bien qu’il soit aujourd’hui plus difficile de caractériser aussi finement une audience à la TV que sur Internet.

read more »

15 octobre 2016

JASP: Expliquer les statistiques et les analyses de façon visuelle et simplifiée

par marketingthema

Aujourd’hui, je vous parle assez rapidement d’un outil intéressant qui permet de visualiser les statistiques de façon assez originale. Le pack n’est pas complet et ne contient pas beaucoup d’analyses disponibles dans des outils classiques comme R, SAS ou SPSS mais l’essentiel est là! Statistiques descriptives, les tests d’hypothèses et aussi l’analyse factorielle. Cependant, pour moi, le plus intéressant c’est l’aspect sur la statistique bayésienne souvent pas très intuitive à comprendre ou parfois non disponible dans certains outils (SPSS ne propose pas d’analyse bayésienne par exemple, sauf via un module complémentaire à installer).

L’outil propose de façon assez simple une alternative bayésienne aux tests d’hypothèse classiques, et on peut ainsi voir par exemple la robustesse du test à travers l’étude de la taille de l’effet statistique et le degré auquel on peut considérer si les résultats sont robustes ou non.

JASP est visuel, avec une approche (live) vous voyez en temps réel l’effet de vos choix de paramétrage sur les résultats! (Ca devient de plus en plus à la mode dans les outils statistiques, je pense notamment à Orange par exemple).

L’approche NHST est de plus en plus critiquée, on a par exemple les travaux de G. Cummings ainsi que de plusieurs auteurs qui expliquent que la valeur P correspondant à la significativité, peut être trompeuse et due purement au hasard, et les chercheurs spécialisés recommandent d’exprimer les résultats en terme d’intervalle de confiance qu’on peut comparer via des méta-analyses sur un phénomène étudié (Pour en savoir plus vous pouvez lire un précédent billet que j’ai écrit sur le sujet).

JASP est disponible pour Mac, Linux et PC et c’est gratuit! alors ne vous en privez surtout pas! 🙂

YB

Download

25 septembre 2016

MROC : Marketing Research Online Communities #Qualitative #EmergingNewMethods

par marketingthema

By: Kristopher Arcand

This past summer, we at Forrester continued to explore new and innovative methodologies. One of my highlights was visiting the IIeX conference in Atlanta back in June. And although I was impressed by the variety of new (qualitative) methodologies, it’s rarely a matter of choosing one or the other. The recent GRIT report by GreenBook shows, for example, that many market research online community (MROC) vendors dropped a few places in terms of innovation, but I agree with Andrew Leary from Ipsos SMX that these online communities will continue to play a relevant (and innovative!) role thanks to their flexibility and variability when it comes to size, duration, integration, and scale.

I recently researched the MROC space, interviewing all the major players to understand their capabilities and how they support organizations. I found that there are a number of ways that MROCs aid customer insights professionals, including:

read more »

6 septembre 2016

The absolutely epic Periodic Table of « Marketing Signals » by chiefmartec #Marketingthema

par marketingthema

 

marketing_signals

The following is a guest post by Steven Wastie, the CMO of Origami Logic. While it naturally aligns with his company’s solution, I think his Marketing Signals Framework — and this absolutely epic Periodic Table of Marketing Signals — transcends their own product and is a powerful concept for marketing technlogy management. CLICK ABOVE FOR A LARGER VERSION

As a modern marketer, you’re overwhelmed with information, all the time — and there’s no end in sight as complexity and the pace of change shows absolutely no sign of letting up.

Your job looks nothing like it did five years ago — even one year ago. Yet despite this permanent state of change, you are, more than ever, accountable for knowing exactly what’s happening across your organization’s eclectic mix of marketing activations at all times. So you’d have no problem explaining how your campaigns are performing right now, today. Right?

It’s a simple question, but for most, it’s incredibly difficult — if not impossible — to answer. Mastering measurement has always been hard and, for many, even aspirational. As measurement vectors and approaches become increasingly diverse and sophisticated, many marketers are more than a little overwhelmed.

Read more from the source here

14 juillet 2016

What if most scientific papers & rules were false? #P-value #ABTesting #Marketingthema

par marketingthema

This is not a joke nor a teasing title! Medical studies, humanities, business research, psychology, etc … everybody is concerned! Actually, it’s a very serious issue about the flaws of P-value used as a scientific objective mean for rejecting the null hypothesis. This kind of statistical validation is being held in all scientific fields as the mainstream way to establish a rule or a scientific law or make a result accepted and published. The way we report our experiments via P-value gives as the false illusion of precision and validity!

In a famous paper published by Loannidis in 2005 (Why Most Published Research Findings Are False) the author conducted simulations proving that most scientific results are more likely to be false than true. The chase of statistical significance is not a warranty that the results are true. According to the author the bias could be coming from many elements such as : weak effects, experimental design flexbility, sample sizes, etc… Obtaining a significant P-value can be the result of chance and not the evidence of an effect (the opposite is also true, where research can be unable to observe a true hidden effect).

Pr. Geoff Cumming defends the idea of a « New statistics » way of working to ensure that we report in research  the confidence intervals instead of the misleading P-value. The above video is a simulation explaining how P-value is also in someway a kind of -random metric- depending on other parameters! and that the only way for science to advance is to express results as an estimation or likelihood confidence interval to compare. 

Researchers should conduct many experiments and encourage metanalysis that aim to make those confidence intervals smaller (more precise) by replication and comparing if those effects can validate (or invalidate) our expectations or research conjectures.

This approach will reset the way we do research by creating incentives to search for the cummulative truth, instead of looking for an artificially probably biaised P-value. It’s also a more humble, honnest and precise way to share scientific research findings and to boost the cummulative aspect of science instead of seeking unreliable novelty or working with a silo research mindset. Our P-values gives us very poor information and exposes heavily research to the danger of false inferences. 

Ressources to consult :
www.thenewstatistics.com : The excellent website of Pr. Cumming, you can download here ESCI software (Excel files) that allow you to run experiments and simulations manually & discover how P-value can be tricky and misinform research.
Buy or rent the book : Cumming, Geoff. Understanding the new statistics: Effect sizes, confidence intervals, and meta-analysis. Routledge, 2013. (A new version will be published in August 2016)

Author: Yassine El Bouchikhi

13 juillet 2016

Pokémon GO: 21 millions d’usagers en 24H! Quelles perspectives pour la réalité augmentée? #Marketingthema

par marketingthema

Le nouveau lancement du jeu sur smartphone Pokemon Go est un succès planétaire! En effet, selon NBC News Le jeu de Nintendo aurait pulvérisé tous les records en 24h de lancement! Rien d’étonnant quand on observe à quel point les smartphones et les applications mobiles sont devenus des produits d’information hautement addictifs! Il s’agit d’une vraie extension digitale de soi comme le disait Russel Belk.

Ce qui est intéressant dans le Pokémon Go n’est pas tant son originalité, ou il faut chercher des Pokémons un peu partout dans la ville pour les collectionner et les échanger! Le plus inspirant à mon avis, c’est qu’est ce qu’on peut faire et créer comme nouvelles applications de cette réalité augmentée. Google, avait lancé par ailleurs bien avant Nintendo un concept similaire qui s’appelait Ingress et qui permettait de chercher des portails aussi dans toute la ville, en suivant les indications géolocalisées de son smartphone et d’entrer ainsi en compétition avec d’autres joueurs d’équipes adverses. J’ai eu l’occasion de l’essayer en 2015 au Maroc, c’était assez amusant! mais pas assez attractif pour m’investir et marcher dans la ville avec un smartphone à la main pour chercher des portails un peu partout! En plus au Maroc, il fallait chercher le portail (virtuel) et se protéger du risque d’agression de voleurs (bien réels) 🙂 Jouer deux jeux en même temps, j’étais pas très motivé encore! 😉

Avant Pokémon Go, Google avait déjà lancé Ingress bien avant! 

La fusion immersive entre la réalité et l’aspect digital peut créer un aspect cognitivement absorbant et émotionnellement stimulant qui peut devenir extrêmement addictif. 

Read more ……

read more »

13 juillet 2016

#Tendances : Le recrutement prédictif déniche le meilleur salarié grâce aux algorithmes

par marketingthema

Source illustration: http://rmsnews.com/

Présents dans la finance et dans le marketing, les algorithmes commencent à s’imposer dans le secteur du recrutement.

 Le BHV Marais a trouvé la solution pour trouver la perle rare et éviter les erreurs de casting : le recrutement prédictif. « Il s’agit d’une technique qui permet, grâce à des algorithmes, d’anticiper de manière fiable les probabilités de réussite d’un salarié. Par réussite on entend sa performance et le fait d’être heureux et engagé dans son travail », explique David Bernard, directeur d’AssessFirst, une entreprise spécialisée dans la conception de plateformes de recrutement prédictif.

« L’utilisation de modèles prédictifs basés sur des algorithmes existe depuis une vingtaine d’année dans la finance ou dans le marketing pour prédire le comportement du consommateur. Depuis 2014, ce type de modèle commence à séduire le monde du recrutement car il réduit les incertitudes », estime David Bernard. L’offre de recrutement prédictif d’AssessFirst existe depuis 18 mois et a déjà réalisé 18 000 recrutements.

Aujourd’hui de plus en plus d’entreprises se présentent sur ce marché, des mastodontes des ressources humaines comme Cornerstone aux entreprises spécialisées dans les tests psychométriques comme CEB ou encore le québécois D teck.

Lire la suite sur le journal du Net

13 juillet 2016

Research in the Crowdsourcing Age, a Case Study #Amazon_Mechanical_Turk #Trends #Marketingthema

par marketingthema

After the goods and services platforms, here is the ultimate uberisation of human intellectual forces … 

How scholars, companies and workers are using Mechanical Turk, a ‘gig economy’ platform, for tasks computers can’t handle

BY

Digital age platforms are providing researchers the ability to outsource portions of their work – not just to increasingly intelligent machines, but also to a relatively low-cost online labor force comprised of humans. These so-called “online outsourcing” services help employers connect with a global pool of free-agent workers who are willing to complete a variety of specialized or repetitive tasks.

Because it provides access to large numbers of workers at relatively low cost, online outsourcing holds a particular appeal for academics and nonprofit research organizations – many of whom have limited resources compared with corporate America. For instance, Pew Research Center has experimented with using these services to perform tasks such as classifying documents and collecting website URLs. And a Google search of scholarly academic literature shows that more than 800 studies – ranging from medical research to social science – were published using data from one such platform, Amazon’s Mechanical Turk, in 2015 alone.1

The rise of these platforms has also generated considerable commentary about the so-called “gig economy” and the possible impact it will have on traditional notions about the nature of work, the structure of compensation and the “social contract” between firms and workers. Pew Research Center recently explored some of the policy and employment implications of these new platforms in a national survey of Americans.

Read more here …..