Showing posts with label digital analytics. Show all posts
Showing posts with label digital analytics. Show all posts

Monday, November 10, 2014

Shopper Marketers: Understand the Data Today or Lose the Market Tomorrow

Today shopping has transformed into an omni-channel experience integrating online to offline at multiple touch points. As a graduate student in Northwestern University's Integrated Marketing Communications (IMC) focusing on marketing analytics,   I have found the following two articles insightful for shopper marketers. They respectively explain the big data utilization in shopper marketing and precision marketing analytics.

“Big data is a journey that every company must take to close the gap between the data that’s available to them, and the business insights they’re deriving from that data.” quoted from the article Shopper Marketing Filling the Cart with Big Data which draws the line form  Lisa Klauser , president of shopper marketing at IN Marketing Services. According to the article, the integration of online activities and offline in-store shopping experiences creates tons of opportunities for brick-and-mortar marketers to track and measure the shoppers in more aspects than ever. By collecting and connecting more data variables together, marketers can more precisely detect the causality. Tracking the patterns in data is almost like keeping abreast of millions consumers from their online engagement to their offline purchase journey, while linking the patterns together could even lead to a map of specific decision-making process of each target segments. More importantly, technology enables marketers to track individual shoppers and connect multiple devices to in-store shopping. In other words, understanding data would help us to unveil the real identity of the consumers, estimate consumer lifetime value and take immediate actions.

                                      Illustration: Mitch Blunt
Source:http://www.adweek.com/sa-article/shopper-marketing-filling-cart-big-data-158332 

More often than not, marketers use constant parameters regression (CPR) modeling to estimate advertising effectiveness, which is acceptable in revealing an overall effect which has been averaged out over time and different instance. In the article Precision Marketing Analytics: Working withTime-Varying Parameters from Accenture, the author compares the CPR modeling with the more advanced time-varying parameters (TVP) modeling. TVP modeling is able to discriminate the actual difference in advertising effectiveness at a granular level over time. Thus, it allows marketers to measure the real ROI of certain marketing activity separated from other actions across the time. As the example of a fictional FCMG company given in the article, using TVP model can especially help retailers and CPG brands attribute marketing efforts to real sales in a more accountable way considering the seasonality, geographic differences and other factors resulting in the volatility of this industry.

Connecting these two articles with my studies of marketing analytics at Northwestern, there are three actions you should consider in your next analytics project. They are as following:

1.     Leap to insights - This not only requires the analyst to carefully examine the availability and quality of data, but also put the analysis in certain context and even from a personalized perspective. The leap from data to insights is usually propelled by finding the connections between consumers’ wills and their actual behaviors.
2.     Deal with complexity. Simply average out the effect in analytics is dangerous in terms of both precise prediction and appropriate interpretation. Including time varying parameters is vital for marketers to discover the real unusual patterns.
3.     Anchor analytics to strategy. Actually, the most effective marketing analytics is an eco-system creating the self-sufficient circle from data collection, personalized interaction with consumers through multiple channels, to measuring marketing effectiveness.

The transformation that is taking place in the CPG and retail industry is just as fast as the increase of data everyday. Always being innovative and analytical, professionals in shopper marketing would never let go of such a great opportunity of boosting their business to another level with data.






Amy Zhou is a marketing analyst who has experience in formulating stories for several CPG and retail brands through data analytics and market research. She is now pursuing her masters’ degree majoring in IMC at Medill Northwestern, with the concentrations in marketing analytics and digital marketing. She is strongly interested in work opportunities in related areas.

Connect with her on LinkedIn or Twitter @Amy__Zhou.

Saturday, November 8, 2014

Digital Strategic Analyst – 3 ways to boost the visitor engagement time on your websites


As a digital strategic analyst, making sure that the web visitors are fully engage in your websites is important. Being a graduate student in Integrated Marketing Communications program at Medill Northwestern University and specializing in digital analytics, with the experience of being a B2B website planner, there are two articles that I found them insightful in better understanding the mystery of engagement in websites. 


The first article Easy Ways to Keep Visitors on Your Website Longer by Eric Siu who is the CEO of San Francisco-based digital marketing agency introduces some principles to keep visitors on the website longer. If you see a bounce rate higher than 50 percent of your visitors, or an average time-on-site that's less than one minute, he suggests that the website should add more multimedia content, add internal pages, make the content more readable and install a 'Related Posts' plugin to your blog. 
.

    ↑ The heat map of websites provided by ClickTale shows which areas on the websites drive the most attention and provides the detail of interaction with the content.

The second article is Time on Page vs Visitor Engagement Time, written by Sam Green who is the Chief Content Editor for ClickTale. This article is to differentiate the time on page and visitor engagement time. The length of time on page sometimes is not accurate to judge the engagement level for a website because visitors might do something else while keeping the browser open, which is still counted as time on page. The aggregation time of actions like mouse move or scroll reach can better represent the time of engagement which can measured by ClickTale, which is a digital analytic service company. For example, ClickTale provides heat map of websites to detect which part of the website attracts more attention. (picture above)


From these two articles and my graduate studies of social marketing in the Medill IMC program, here are three action items to help you boost the engagement time on the website by focusing on the actual engagement.
  •    Ask For Action- If people don’t know what to do on that website, then they will just leave. Add the action buttons on the sites. A clear navigation of a flow toward to a destination page also can let visitor know what to do.
  •    Site Related Posts- If people like A, they are more likely to like something relevant to A. Therefore, putting a related post beside the content or adding internal links to the sources will make the visitor continue to dig in the topics with deeper engagement with the community.
  •   Make It Readable- Shorten the paragraph and add bold subtitles. Visual graph is always better than a bunch of words. People will leave the page if the first glance of the page is not appealing.

These tips should help you improve your site’s engagement time. Note that evaluation of time on page is not enough to evaluate the website’s performance. Instead, you should look at action-related metrics. If you have enough budgets, it’s worthy to invest an advanced analytic tool which can identify the real behaviors of the visitors when staying on the page. When your websites are in low engagement, try these three tips to increase the engagement time on your page. 


Jean Chang is M.S. Candidate of Integrated Marketing Communications at Medill, Northwestern University. She specializes in digital marketing and marketing analytics. Prior to IMC, she has worked for a candy & gift company as a marketing planner for 2 years. Her passion in digital industry began with the experience of being a B2B website planner, being in charge of providing content of the website and organizing the function and flows. Welcome to contact her by LinkedIn and Twitter@JeanIMC with any questions or comments.

Wednesday, May 7, 2014

In data analytics, you need to evaluate whether your data has value before you start analyzing it.

With the explosion of data that your department and organization has access to, whether it’s generated from daily operation or social media, it’s really intimidating for you as managers to decide whether you should and how you can make sense of them. As a business analyst that has help many fortune 500 companies solved various types of data analytics based business problem and a student that is pursuing my master of science degree in integrated marketing communication with concentrations in marketing analytics and digital marketing at Northwestern University, here are the metric that I develop for managers like you to evaluate whether your data can add value.

Relationship at Scale, Data Velocity & Design for Analytics
In the article “Every Business is a Digital Business”, the Accenture Technology Vision 2013 identifies seven technology trends that are shaping business in the near future. Three of the trends Accenture identified are related to the accessibility and quantity of customer data, which are essential for analytics managers like you to understand and utilize. First, “Relationship at scale”. Rather than “faceless digital transactions or a cookie file”, customers are real people with real differences. This brings to you as a marketer the new responsibility to utilize available date, redesign your relationship with your customers and improve their experiences. The second technology trend progresses with "Data Velocity". It is “the pace at which data can be gathered, sorted and analyzed in order to produce insights that managers can act on quickly.” Needless to say, matching the speed of decisions to that of actions comes to be the key for marketers to win their market. Third, “Design for Analytics”. Accenture argues that despite several relevant functions from the currently available software, the next generation of software will be “designed for analytics” in the first place, giving you the option to outsource real-time analytics while performing customer segmentation analysis internally.


Evaluate your data's worthiness 
With possibilities derived from huge data quantity, two important questions should be raised. Does your company have valuable data? What is the right business model to extract value from worthy data? In the article “What’s your data worth?” , Accenture Interactive Research and Finding provides us useful metrics to understand data. First of all, the data your organization has should have the following characters to be consider valuable: reflecting consumer behavior, customer identification, transaction frequency and being accessible. After ascertaining the value of your data, the next step is to monetize the data through the process of data value chain. In this process, you need to refine the raw data into processed data, to interpret the processed data into insights, to profile the insightful segments  into presentations and to use the presentations to adjust actual transaction. 

After reviewing these two articles, there are three action items you should consider immediately to lead your department or company to move forward.

1.       Evaluate your data’s worthiness
Evaluate the available data of your organization to see if they can reflect consumer behavior, customer identification and transaction frequency.

2.       Understand your data’s velocity
Estimate the time, capability and budget needed for making sense of available data through the five stages, which are raw data, processed data, insights, presentation and transact.

3.       Post hoc or real-time analytics
Decide how you want to analyze the data you have in terms of whether you want to conduct post-hoc analytics such as segmentation and targeting or you want your data to be analyzed in real-time such as web monitoring. I would recommend marketers to do segmentation internally and outsource real-time analytics to design for analytics software providers.

It’s true that the word big data is over used while little people really understand why and how to actually utilize it. But given the fact that every business is a digital business and that the availability of data on your business operation, it’s important for you as analytics and marketing manager to start right away to evaluate the value of the accessible data, estimate the energy needed to make sense of it and to decide whether you should analyze it internally in order to be customer-centric and gain customer growth. If you want a detailed check list of how to evaluate the data worthiness and how to transform raw data to actual business transact, please feel free to reach out.


Joey Bian is a business strategist and data analyst that is passionate about transforming data into actionable business strategies. He's got 2 years' relevant experience across industries and has solved various type of analytics based business problems. He is now pursuing his master of science's degree in integrated marketing communication with concentrations in marketing analytics and digital marketing at Northwestern University. And he wants to continue his career in business/marketing analytics afterwards. 

Contact him on  LinkedIn  or follow him on Twitter @JoeyWantsJoy to continue the conversation.