Week 2: Data Collection

Data Collection - MOOC Summaries - University of Illinois (Urban Champaign) - Digital Analytics for Marketing Professionals: Marketing Analytics in Practice

Week 2: Data Collection

“Preparing for the Analysis Journey, Part 1…Preparing for the Analysis Journey, Part 2,..Data Collection Part I: Unstructured Data…Comments by Prof Rhiannon Clifton…”
(Source)

Summaries

  • Lesson 2: Preparing for the Analysis Journey, Part 1
  • Lesson 2: Preparing for the Analysis Journey, Part 2
  • Lesson 3: Data Collection Part I: Unstructured
  • Comments by Prof Rhiannon Clifton

Lesson 2: Preparing for the Analysis Journey, Part 1

  • Those five are build awareness, influence consideration, improve sales processes, reposition your brand, or grow loyalty.
  • For building awareness, ask yourself do consumers recall and recognize my brand.
  • For influence and consideration, ask do the products I have satisfy consumers’ needs.
  • If they don’t, that means consumers are choosing other products, competitive products to yours and, as a result, you need to better influence consideration and drive consumers to your products.
  • Third question for the third business objective is do my sales efforts result in wins for my brands.
  • For an e-commerce company, once someone gets a product into my checkout process, are they completing that process or do I have large cart abandonment.
  • Is there something in that process that is preventing them from becoming a customer? Dependent upon the answer that you have to that question, that’s going to tell you if your brand needs to improve its sales processes.
  • Do the experiences I deliver fulfill customer expectations? If they don’t, I’m going to need to reposition my brand.
  • I’m going to need to do something that will allow me to produce products or set the expectations in consumers’ mind that the products that I have, the brand that I promote, actually do fulfill whatever needs they have and fulfill their expectations.
  • A fifth question that you can ask is do consumers advocate for my brand? If they do, I probably have strong loyalty.
  • If consumers don’t advocate for my brand, well, then loyalty could be an issue.
  • A great framework to help here is the consumer decision journey that is produced by McKinsey and what that CDJ does is it takes the whole process for consumers from the time they are triggered for a product need all the way through to the point that they make that purchase and then the experiences afterwards.
  • Everything begins with a trigger, either a customer sees an advertisement and says wow, I really want that product and now that’s instilled a need and the more they run out of a product that they have.
  • Some kind of trigger has happened for them to say I want to get that product.
  • The first thing they’ll start to do is think about what brands could deliver that product.
  • You want customers to think of you when they need a product that you sell.
  • It’s when consumers are collecting a lot of information to evaluate the different product choices and brands that they have available to them.
  • Now they have actually purchased a product or they’re at that point where they’ve done all their analysis, they’ve done all their evaluation and they’ve arrived at a product and actually made that purchase.
  • This is was P&G calls that first moment of truth when a consumer is standing in a grocery store or a retail outlet and sees the different products in front of them and has to make a choice between those where brands are competing for attention.
  • You get the product home, you have to see do I like this, does it fit the expectations that I had going in.
  • Then you’ve got this little thing called the loyalty loop and what this is is the place that every brand wants to be.
  • It’s basically a shortcut of the entire consumer decision journey and it says that if that trigger arises again, if I’m a satisfied customer and I like that brand and the product that you produce fits my needs, I will go right to the moment of purchase.
  • I’ll jump in that loyalty loop and avoid the possibility of buying some other brand’s products.
  • The recall, the brand awareness, fits very well with the initial consideration side.
  • If you are not- if your brand is not on that initial consideration set, you’ve got an awareness problem.
  • Do my products satisfy customer needs? If they don’t, they’re going elsewhere for that product, you need to influence consideration.
  • You can see how those different brand objectives really measure up very well to the areas in the consumer decision journey.
  • That’s an understanding the customer challenge and very important for all sorts of brands, but if you don’t know that, it’s not necessarily a marketing challenge that you face, it’s just you don’t have the depth of understanding of the customers that you want.
  • That’s why I haven’t mapped that one back to a consumer- a brand objective.
  • For the trigger, a click stream analysis can be very important to see how consumers are moving around my website, where they’re going, what’s important to them, what is not.
  • Outcomes analysis is a fantastic way to say here was a purchase, what were the factors that led up to that, or here’s a customer that didn’t buy my product, what were the factors that caused them to go elsewhere.
  • What do I think about the product? As well as collecting data around brand advocacy.
  • Are consumers advocating for my brand? That’s a great analysis to learn more about that loyalty move and your customer loyalty.

Lesson 2: Preparing for the Analysis Journey, Part 2

  • Tools that we have available to us are many and here’s a sampling of some of the more prominent ones that are either free or very low cost and so any analyst, no matter where they are, can use these kind of tools to conduct those different types of analyses that we just talked about.
  • You see how we’re kind of layering in and building together that objective that I’m on against the key questions that I need to ask to answer that objective, the tools that I’ve got and the analysis techniques that I’ll use.
  • You want to document your business objective, tie that to a key question, and then identify data and sources.
  • One of the questions that we’ll want to ask then is how consumer interest in our brand trended over time has.
  • More search volume means more people looking for me, more people interested in the brand.
  • Those are two sources that we’ll use to answer that key question.
  • A second key question is what consumer group is our strongest advocate.
  • Now Twitter is not going to tell us which of our- which of their Twitter accounts line up to which of our consumer segments, so we’re going to be clever marketers and let’s say that we’ve used hash tags in some way to tie people to different consumer segments.
  • Third question would be which marketing programs have grown advocacy for us.
  • We’ll use our company internet site, which has the calendar of events that we’ve run and here again we’ve probably asked consumers, because we’re clever marketers, to use hash tags around those different campaigns.
  • Then we’ll just go and check out Topsy, which can provide for us the hash tag usage so that we get a sense for when the campaigns- what volume- relative volume of Twitter activity did each campaign kind of create.
  • With this in hand, now we are ready to start going out and identifying the data sources and pulling those data source- pulling the data out of those sources and then starting to conduct some analysis to answer those key questions, but without this phase, without this very important document, we would be heading off on a journey without a map and, as you know, those never end well.
  • The key questions that we’re going to ask of the data must tie directly back to those key questions or business objectives and must answer them completely.
  • McKinsey’s CDJ is a very helpful framework to help us in that process and then there are many, many tools available to us that are free or very low cost, which makes accessibility of this sort of analysis very, very global.

Lesson 3: Data Collection Part I: Unstructured

  • Data Collection part 1, Unstructured Data, so five things we’re going to talk about in this lesson.
  • The first one is that linking to your plan is critical for efficient data collection.
  • The second is that unstructured data and structured data differ in very key ways.
  • Third is that unstructured data, despite its name, can still have very neat headings and rows but we’ll talk about what exactly makes it unstructured.
  • Then fourth thing is that there are three primary ways to collect unstructured data on the web, all of them very important and very important tools in the marketing analyst tool belt.
  • Then the fifth thing is that some of those tools are very, very powerful and once you understand how to use them it really opens up the entire worldwide web to you as a data source.
  • We are at the first part of collect the unstructured data.
  • What we’re going to talk about here is this unstructured data source, a customer service rep database that has some free form text entry about what the consumer is calling about.
  • Then the Twitter API and Topsy, both of them which will pull Twitter information for us, Tweet data.
  • So what makes the difference between unstructured data and structured data? What makes unstructured data unstructured? The first is that, and the primary difference, is that unstructured data does not have a predefined data model.
  • The reason it is so important is that it has been thought that unstructured data might account for 70% to 80% of all data in organizations.
  • R is a fantastic program and fantastic tool to use to get to unstructured data.
  • What makes this unstructured data is right here, it’s the actual tweet that has been collected.
  • There is, even though Twitter confines you to a certain number of characters, there’s still no defined length in the data here.
  • Three tools that we’ll use to collect unstructured data include bulk downloads, the ability to download big volumes of data, API’s that we’ll tap and then web pages that we’ll scrape.
  • There are so many sources of information out there that will include large volumes of data for you to collect.
  • The trick is just learning how each of those provide data for you and then which ones are relevant to the work you are doing and which ones aren’t.
  • You can kind of build for yourself a set of go to data sources.
  • As an analyst, take a little time, watch some of those videos, complete some of those courses and get familiar with that capability because once you do that, literally any web page that has information on it becomes a data source for you, an accessible data source.
  • Unstructured Data, what did we talk about? What did we learn? First was that linking to your plan is critically important.
  • Some of those tools, particularly R plus R Studio, open up the entire World Wide Web to you as a data source.

Comments by Prof Rhiannon Clifton

  • I teach the next two courses in the specialization, digital marketing channels.
  • The landscape in which we look at the entire digital landscape in terms of channels that we can use to achieve our marketing objectives.
  • We then switch gears to digital marketing channels planning, where we look at each channel individually and together to see how we can use them to achieve our marketing objectives.
  • How these things can translate to our marketing tactics.

Return to Summaries.

(image source)

 

Leave a Reply

Your email address will not be published. Required fields are marked *