Tag Archives: analytics

Cognitive Analytics: Enabling assisted intelligence in human resources recruiting and hiring by Noel Webb, @CognitiveHR, Karen.ai, #BigDataTO #BigData #AI

Notes from the #BigDataTO conference in Toronto

  • He realized that HR teams were spending too much time prescreening resumes before they could even meet with the best candidates
  • Recruiters only spend 6 seconds reviewing a resume which means they end up accidentally discarding some of the best ones. Time crunches mean they may only be able to get through 20% of candidates. ML can solve these problems .
  • 75% of candidates who apply to jobs do not hear back from the company because there are simply too many candidates and not enough time to do so. NLP and chatbots can solve this problem.
  • AI will not steal all jobs but it will automate processes and allow you to engage with potential hires in a more meaningful way.
  • Shortlisting is a huge challenge for HR as reducing a huge list of resumes into a screened list takes a lot of detailed attention. Technology such as direct keyword matches aren’t the best option as they eliminate people with relevant skills but not the exact words. For instance, know R is just as good as knowing SAS but a keyword search wouldn’t know that. NLP would work much better.
  • Personality insights can also be collected using sentiment analysis to get a functional understanding of the Big 5 Personality traits. [Wow, I can’t imagine how valid it is to do personality assessments with resumes which are often written by third parties and without traditional grammar and style]
  • Chatbots can take an applicant through hiring and onboarding processes by answering questions that would normally be asked of an operations officer. [imagine how many stupid questions the chatbot would be asked that new hires are too scared to ask people]
Advertisements

Co-creating the future #IIeX #MRX 

Live note taking from #IIeX in amsterdam. Any errors or bad jokes are my own.

Using insight and analytics to steer change in an organization by Melissa Gill, Credit Suisse

  • The five stages of grief – you show us things we don’t we to hear, you make us question our sense of purpose, create a sense of loss for ourselves and our team, people go through denial as they read the results, it takes a while to accept the final results
  • Mission statements are a call to action, be part of the change, have clear direction, creating a vision is less functional
  • Define the problem you want to solve next year, create a team to act on a critical issue, don’t make people boil the ocean, get value this year so you can create a budget for next year
  • Assess the ideal situation, work requires part of a person not a whole person and it’s expensive, find out where the data is, where the expertise is within your company
  • Plan for a win, get results in 6 to 9 months minimum, don’t demonstrate technical competition yet, demonstrate value for the business within the budget timeframe
  • Solve your mission problem with incremental wins
  • Communicate, don’t communicate your wins or learning, also communicate what you can do better, help other people see other areas where they can contribute or add to your success
  • Create the right environment, who is the lead – the CRO, CEO, or someone else, the spehere of influence is quite broad
  • Its better to focus on learning how to improve things rather than asking questions that reflect your KPIs

Made you look! Using eye-tracking to see digital advertsing in a new way by Colin Deller and Maria Sealey, AIMIA

  • 46% of impressions are not viewable, this costs 1.6 billion pounds per year for UK companies
  • What do people actually look at online, do they look at native as much as display advertising
  • Infra-red eye tracking technology is available
  • Eye tracking showed that people weren’t pay much attention to a text ad, indexed it to a penguin ad and you could see much more attention paid to the image ad versus the text ad, more people viewed, for longer time
  • Only 35% of ads we purchase are viewed by people, only 9% of people look at them for more than one second
  • This is low in comparison to traditional ads, press gets more views because there is no load time and it’s viewed longer, 40% look at the ads for more than one second, 2.2 seconds average for print
  • But you cant just lift the print ad and use it is a a digital ad
  • Ads aren’t always immediately viewable and sometimes they switch to a different ad immediately
  • Viewable does not mean its being viewed
  • Think like.a poster not like direct mail
  • User cost per view, not cost per thousand

FinTech and Disruptive Innovation in Financial Services by Anthony Michelini, Citigroup

  • Citigroup is over 200 years old
  • Ingenuity has been art of their DNA 
  • Banking isnt just credit cards, it’s advising, investments, money transfers, bill payments
  • The bank is the hub of people’s money but there our spokes coming out of it
  • Half of millennials are already using financial tech solutions, 60% are happy with it, two thirds will use more going forward
  • Only half of millennials would be happy if they only had financial tech, but the trend is similar for older people, B2B are the furthest behind but still want digital
  • We can’t be all things to all people
  • Six markers of progress makers – optimistic, driven, resilient, future focus, worldly outlook, generosity of spirit
  • These come from motivations as people and how they view their lives, not just how they use banks
  • They use communities, advisory boards, innovation labs around the world

Tipping the sacred cows of MR #IIeX 

Live note-taking at #IIeX in Atlanta. Any errors or bad jokes are my own.

Will Watson replace researchers? By Bruce Weed

  • Health data will grow 99%; Insurance data will grow 94%; Utilities data will grow 99%, and more than 80% of that data will be unstructured
  • Machines don’t make up answers, they will give the answer you teach it to give
  • Now we teach machines to read images like MRIs, a doctos can’t remember an MRI from ten years ago but a machine will
  • Machines understand, reason, learn. They can learn multiple languages too. Can teach it how to read, hear, see, and 9 languages 
  • Showed all the Ted talks to Watson and now it will find the relevant part of the video you want to see
  • Teach machines to do more than a keyword search, teach it to learn and understand
  • Machines are listening in to call Centers and helping the agents give better answers
  • Machines learning will give us crime and threat detection, early detection of diseases, understanding customers, new product development
  • Machine learning makes humans smarter because it gives us capacity

Co-Creating a tailored experience to identify relevant insights leveraging advanced cognitive text analytics by Sion Agami and David Johnson

  • There are lots of five star ratings out there but not all five stars are created equally
  • Can’t approach analytics from a single dimension
  • Corpus linguistics – how people communicate
  • Olden days used to be keyword, Boolean, taxonomies
  • Now it’s NLP, machine learning, topics modeling – these are probabilistic models – 65% confidence that this is what you wanted, what if 4 different models are 65% confident?
  • Next is leveraging all methods in parallel – focus on emotions and cognitive states
  • Emotions, persona, experience, purchase path, topics are all important
  • How do you rate BOO, Not like, disappointed, like, good, WOW, and then add the emoticons into the scale
  • Algorithms can pick apart which products really are a 5
  • Fix the social media comments that are filled with emotion
  • How do identify WOW experiences before launching a products? What is the best question to ask consumers so they can share emotions, how accurate does your model need to be, can you measure what moved the needle from consumers with confidence
  • Put new tools in front of people who are passionate, those with project specific challenges
  • Watch out for groups who think they can already do something, maybe it’s time to work together OR let the people are ARE doing being the people who DO

The Perils and Pitfalls of Recall Memory: How flawed recall and memory bias pollute market research with David Paull, Elizabeth Merrick, Andrew Jeavons and Elizabeth Loftus

  • [I did an entire class in graduate school on unconscious and flawed memory. I’m totally on board with this session. Love this topic. Wish I could remember more of it. Ha ha. I really do.]
  • Market research has made a lot of assumptions about how memory works, completely contrasting academic research, we can’t remember names so how can we remember the past
  • [we need more true academics in market research ]
  • We assume what you did in the past will predict what you did in the future, or that we can predict
  • Our goal is to make money, we want to know allocation of marketing dollars so we ask about recall, we just don’t have better tools though good tools are on the horizon
  • There are lots of false positive and false negatives in recall data, 15% of people misremembered receiving something [This is NOT a bad respondent or a cheater or fraud. This is real human behavior.]
  • There is more to memory than forgetting, false memories are a huge part of memory
  • It’s very easy to expose people to leading questions, misinformation, erroneous versions and to contaminate or transform people’s memories
  • You can plant entirely false memories for things that didn’t happen, it has consequences, it affects their thoughts intentions and behaviours, memory is malleable
  • They planted memories that people got sick eating something as a child and people no longer wanted to eat those foods, they planted positive memories and got people to like yuck foods more
  • Should we take advantage of this to make people happier and healthier, or use them for marketing purpose
  • Sounds like advertising, we find a feeling like nostalgia so we put that into an ad
  • [we need more academics on stage. Most market researchers just don’t have the relevant psychology/sociology background]
  • Manipulation feels creepy but that’s a practical application
  • What is ethical – a therapist helping someone eat better, maybe not; What about a parent doing it with an overweight child?… Hello Santa Clause. WHich would you rather have, an obese child with heart problems or a child who remembers broccoli with grandma when that never happened
  • People have a lot of fiction mixed in with their facts
  • Memory includes “what you bought at the store last week” but memory also includes meaning “I remember the brand Uber” but I may now remember going to the store
  • Semantic memory helps us build great products
  • Memory applies to doors – we expect a pull door to look in a certain way different from a push door
  • We know we shouldn’t have long questionnaires, cognitive load is a problem, that hurts recall, we need to make it easy for people to recall episodic memories, it’s very shaky to ask people to remember the past
  • At least get the recollection as soon as possible, as they’re happening, need to get it before they interact with other people, responses influence each other, doesn’t matter if it’s a focus groups, early responses affect what people say later on
  • Automated systems can help remove some biases, qualitative is less and less reliant on humans
  • think about biases of respondents and and yourself
  • If people know they can look up the information later, they won’t try to remember it, we no longer bother to remember phone numbers, passwords are a huge problem
  • Does the precise memory matter more than the feeling, we can alter the feelings people have about products
  • “You told us you were a 4 on that scale” and many people won’t remember that they originally said it was a 2
  • Must think carefully about the outcome you need, be realistic about when you need precise memories vs insightful memories, knowing it was the 37th flow of a building may not matter because all that was important was that it was high

Social Disruption: The vertical network arrives by Ashlyn Berg

  • There are many social networks specific to  careers – ResearchGate is for PhDs, Github, ZumZero, SpiceWorks
  • Community aspect – online home for professionals to interact with their peers
  • Content – users share millions of original and shared content to stay up to date on trends and do their jobs better
  • Apps and tools – help people get their job done
  • Mostly for free so people engage for a long time
  • 1-stop shop for marketers, place to build relationships, platforms for research
  • Easy place to investigate Ned’s and challenges of your audience
  • Better platform for research over other alternatives
  • Rich projected information, not just into about their company, know which apps they use, hardware and software they use, massive amount of Behavioral insight
  • Vertical network is very clean data, social behavior is clean, it is the real audience you want to talk to, not someone who wanted an incentive 

Location and the Art of Business Analytics by Simon Thompson #FOCI14 #MRX

Live blogging from the #FOCI14 conference in Universal City. Any errors or bad jokes are my own.foci14

Location and the Art of Business Analytics 
Simon Thompson, Director, Commercial Solutions, ESRI

  • language evolved in order to map things in our head
  • we implicitly understand location, our smartphones understand our exact location to 6 decimal places of latitude and longitude which is 100 meters
  • we use location to add context – addresses, activities
  • if you buy a phone with an air pressure sensor on it, then you can tell the elevation, which floor of a building – time, location, height means you’re at this marketing research conference
  • i will target certain things to get a coupon but they don’t need to know where I am all the time
  • context trumps content
  • in 2020 there will be 50 billion connect devices and they are all location aware
  • 55% of all retail internet time originated on smartphones and tablets
  • the death of things – paper maps, notebooks, business cards, pedometer, tape measure, scanners – all dying because of smartphones
  • people think about geography as getting coupons – is this minority report?
  • 80% of purchases are made within 10 miles of home or work
  • combine social media mining with location monitoring
  • locavation – reason for being in or at a place, act of giving someone a reason or incentive to be somewhere, commitment that makes someone want to be somewhere, behaving in a particular way in a particular place
  • location is the new cookie – place impressions out perform page impressions
  • we live in the internet, not on the internet
  • recall the case when Target identified an underage girl as being pregnant, and then sent advertising to her? Why didn’t Target recognize she was a GIRL, not an adult? They had all the data to identify that datapoint as well but they ignored it?
  • “if you build it, you can monetize it” — that’s scary! build a technology and then try to use it
  • the shopping journal is just like a shopping journal online – except online you can also get a welcome back, personalized offers, recommend products, tailor the content, analyze behavior
  • can a physical store do this? with location devices you can localize products for an area, detect arrival at a location, tailor offers by proximity, analyze raw data, concierge services for premium customers, validate effectiveness of window displays, modify store layout, analyze checkout queue times
  • indoor mapping with transform retail in the same way that freeways did in the 1950s
  • smartphones make it hyper hyper local

Other Posts

Storytelling Through Digital Analytics by Scott Vanderbilt, NPR #MRA_National #MRX

… Live blogging from Disney Orlando, any errors are my own…

Storytelling Through Digital Analytics

Scott Vanderbilt, Digital Research Manager, NPR; Sarah Withrow, Senior Research Analyst, NPR

MRAInsights

  • NPR is a news media organization “National Public Radio”
  • Afternoon show “All Things Considered”, how does a radio show translate to digital
  • Key issues
    • What is doing well or not in their content, need it real-time, there are dips on certain days and certain hours, as well as certain shows
    • How to better engage with afternoon listeners and increase the stickiness of the content?
  • Digital metrics of audience behaviour gives us the what- when do they tune in and out, which stories do they tune in and out of, amount of time listening
  • Diary studies gives us the how – archetypes of current listeners, grouped by listening setting, importance of the stories,where they listen to the show whether at home or car or work, what platform they listen on, find out what they’re looking for, what they turn to after the show
  • Focus groups gives us the why – motivations and perceptions, they like particular hosts for particular stories
  • English: The logo of National Public Radio's A...Why do they tune out?  it’s not length as long as the story appeals to them.
  • Are they expecting something different? they expect eclectic stories, they love not knowing what the stories will be
  • Does the flow of the show match their needs?  i.e., the order of segment. changing the flow is fine if it’s done gradually
  • [this show sounds like “Sunday Morning” which I love]
  • What would increase the stickiness? Provide the full show or pieces of the show digitally
  • Does the tone match what they want? do they want shorter, snappier news?
  • Which shows draw the largest audience, which content kept them on the website longer?
  • Use voting buttons to see which shows viewers like on their website

Tom Anderson: Web Analytics #Netgain #mrx

php|architect's Guide to Web Scraping

Image by CalEvans via Flickr

What follows are some of my silly musings and key take-aways of the session.
Tom Anderson – Web Analytics
– 85% of all data stored is unstructured, it doubles every three months, 7 million web pages are added every single day
– First, tracking survey case study, analysis of guest satisfaction survey which has 10 point scales and permits verbatim responses
– Funny thing is the checkbox answers were different from the verbatims. Checkmarks related to the room and the bed but the verbatim was about the food that made her throw up. The verbatims MUST be read! (people assume you’ll look after the problems and use the comment box for stuff you forgot to ask about, at least that’s what i do)
– Problem with manual coding is code frame changes over time, some codes are missing, some codes become irrelevant, inter-rater reliability (different people and same person would code it differently)
– ooooh, CHAID results, and regression equation 🙂
– Future – surveys might look like a blank post card, thumbs up or down and then write in all your comments
– Second case study, five hotels within a travel website
– Indexing might be the new word for webscraping (it’s a tech term that’s nicer than scraping!)
– 20% of the users are responsible for 80% of the posts, pareto principle, most people make just one or two posts in the last year or so
– “online introverts” folks who are listening but don’t say too much
– People posting on multiple hotel boards are looking for cheaper rates, free nights
– Loyalists who focus on one hotel board are more positive about the hotel
– Had a board lurker who interacted with posters, he knew specific people (slippery slope, researchers can’t do this but the client was the lurker so he was able to)
– Was able to see client’s promotional schedule in the text analysis, nice validation
– 60% of online population uses a social network, anyone under 24 is on a social network usually facebook
– WW2 generation is showing the fastest growth particularly to stay in touch with their family, photos of the grandkids and such (ah, isn’t that sweet, STOP following me gramma!)
– LinkedIn has 65 million users (hey, LinkIn with me!)
– Social networks let people raise their hands that they like a certain brand
– Text analytics predict income and purchasing/spending power on LinkedIn
– Qualitative analysis is a sample of information, text analytics can measure entire population

Related Links
#Netgain5 Keynote Roundup: Last Thoughts
Brian Levine: Neuroscience and Marketing Research
Brian Singh: Insights from the Nenshi Campaign
Monique Morden: Online Communities, MROCs
Ray Poynter – Overview of Online Research Trends
Tom Anderson: Web Analytics
Will Goodhand: Social Media Research and Digividuals

%d bloggers like this: