DM_A_Practice

DM_A_Practice

1. Plan

   • Define
     • Objectives
       • clear
       • singualr
       • 5
         • brand awareness
           • Do consumers recall/recognize my brand? 
         • influence consideration
           • Do my products satisfy consumer's needs?
         • sales process improvement
           • Do my sales efforts result in wins for my brands?
         • Brand reposition
           • Do the experiences I deliver fulfill customer expectations?
         • Loyalty growth
           • Do consumers advocate for my brand?
         • choose only 1 to focus
     • questions
     • analysis type
       • resultant data
     • approach
       • data
         • required
       • tools
         • to use
     • Document

2. Collect

   • Measure what matters
     • locate
       • sources
     • utilyze
       • tools
       • techniques
     • select
       • data management system
     • limit
       • bias
   • Unstructured data
     • tools
       • bulk downloads
         • e.g. statistic websites
       • API
       • web scraping
       • R
   • Structured Data
     • look for links
     • tools
       • MS Access
       • STATA
       • SPSS
       • SAS

3. Analyse

   • Monitor and learn
     • "tidy"
       • datasets
         • column
           • 1 variable
         • row
           • 1 observation
         • 1 table/type of variable
         • multiple tables
           • key to link
         • change
           • copy file
       • errors
         • measurement
           • sample bias
             • when data collected from a sample is not reflective of the population
         • data entry
           • when data is recorded incorrectly
         • table integration
           • when data imported or joined from multiple databases does not integrate smoothly
         • calculation
           • when inacurate formulas or manipulations produce unintended outputs
     • address
       • data-quality
         • issues
         • concerns
     • trust data
     • make descisions
       • you comfortable
     • drill deeper
       • later
     • get comfortable
       • with
         • data
         • limitations
     • be consistent
   • perform
     • analysis techniques
       • draw conclusions
   • Categories
     • Predictive
       • Uses data on some object to predict the future value of others Depends heavily on having the right data and data quality Must remember that if X predicts Y, X does not cause Y Offer recommendations for actions that will successfully drive brand interest in the future
       • sources
         • Media optimization modeling Attribution modeling Other response modeling (e.g., customer & channel response)
     • Descriptive
       • First kind of analysis performed Commonly applied to raw data Generalizations typically not possible without modeling Determine whether consumers are or are not interested in your brand
       • sources
         • Web traffic data reports Web server performance data reports Web transaction data reports Competitive intelligence Clickstream analysis 
     • Inferential
       • Uses a small set of data to say something about larger set Commonly the goal of surveys Results depend heavily on population and sample Provide a description of what a typical consumer who is interested in our brand looks like and how they behave
       • sources
         • Usability studies Voice of customer studies Other surveys when a sample of a population is included
     • exploratory
       • Seeks to discover connections and patterns in the data Good for defining future studies Should not be used alone for generalizing or predicting Identify when interest started for the brand, when it ended, and other interesting patterns in the data
       • sources
         • Outcomes analysis Other regression and performance correlation analyses
     • Causal
       • Seeks to determine what happens to one variable when a change is made to another Usually seen as average effects Usually the “gold standard” for data analysis Explain what has been done in the past that successfully drove brand interest and other influential brand attributes
       • sources
         • Experimentation and testing Multivariate & A/B testing Site optimization Other campaign optimization (e.g., “Heavy up” media tests)
     • 3 rules
       • Three rules of context ensure impact in your analysis • Never produce an analysis that reports a metric all by itself … period • Use benchmarks (internal or external), goals, or even prior performance to give some kind of context • Include insights — in words — to summarize performance and recommend actions in every analysis

4. Report

   • Communicate and act
     • visualisation
       • quick
       • efficient
       • atributes
         • preattentive
           • how many 9's you can see: 1) among lots of other numbers 2) still among numbers but highlighted
           • Highlight your message and eliminate distractions
           • Use visual cues to help lead your audience through your insight
           • Use contrast (size, color) to capture the reader’s attention (and imagination)
       • test
         • Would eliminating this change anything? If the answer is no, get rid of it. Earn your reader's trust by giving them only what they need. Where is your eye drawn? Look away from your visual for 5 seconds, then back to it. This is most likely the place your audience's eye will be drawn as well. “The colleague test”. Think your visual is perfect? Have a colleague take a look. Give them minimal context and 10-15 seconds to process and ask what they would take away.
     • recommendations
       • clear
       • concise
     • story
       • passion
       • 60 sec
   • presentation
     • 1 BIG idea
       • 1 sentence
         • might say, “It’s about why production will keep missing key deadlines unless we upgrade our workflow management software.”
       • add
         • supporting ideas
     • from audience's interest
     • glance test
       • These five design principles will help you simplify your slides so they’ll pass the glance test: Flow. You can direct people’s eyes to certain areas of a slide to emphasize important points. Here, for example, your eye takes in the cluster of grapes, then moves to the message about quality, and then focuses on one beautiful grape from the “yield”: Contrast. Our eyes are drawn to things that stand out, so designers use contrast to focus attention. Create contrast through your elements’ size, shape, color, and proximity. In this example, a presenter compared cross-sections of skin and soil to show that tending to both requires an understanding of the microbiological activity beneath the surface. The blurred backgrounds set off the stark white illustrations for quick visual processing: White space. White space is the open space surrounding items of interest. Presenters are often tempted to fill it up with additional content that competes for attention. But including a healthy amount of white space sharpens viewers’ focus by isolating elements. In this example, if we’d paired the text with a larger or more detailed image, your eye wouldn’t know where to begin, and the quote would have lost its power: Hierarchy. A clear visual hierarchy allows viewers to quickly ascertain a slide’s most important elements: Unity. Slides with visual unity look as though the same person created them and make your message feel cohesive. You can achieve this through consistent type styles, color, image treatment, and element placement throughout the slide deck. 
       • comprehend
         • So when adding elements to your slides, have a good reason: Does the audience need to see your logo on each slide to remember who you work for? Does that blue swoosh add meaning? If not, leave it off. The same goes for text. Keep it short and easy to skim. Scale the type as large as possible so the people in the back of the room can see it.
         • 1 slide / 3 sec
       • consistency
         • It’s also important to stick to a consistent visual style in your slide deck. Select one typeface — two at most. Use the same color palette throughout (limit yourself to three complementary colors, plus a couple of neutral shades, like gray and pale blue). Photos should be taken by the same photographer or look as if they are. Illustrations should be done in the same style.
     • meaningful and action dashboard
       • When Presenting Your Data, Get to the Point Fast
       • 1. chart
       • 2. translate chart into english
       • 3. Recommendation for actions
       • 4. Outcomes/threats

Imágenes y archivos del mapa