Posts tagged Key Driver Analysis
Key Driver Analysis: Top-down & Bottom-up Approach

Text Analytics Tips - Branding Get a complete picture of your data: The ‘Top-Down and Bottom-Up Approach’

At OdinText we’ve found that the best way to identify all key drivers in any analysis really, especially in customer experience management (including but not limited to KPI’s such as OSAT, Net Promoter Score, Likelihood to Return or other real behavior) is through a dual process combining a theory-driven (aka “top-down”) and a data-exploratory or data-driven approach (aka “bottom-up”):

Top-Down

This approach requires you to identify important concepts or themes before even starting to explore and analyze your data. In customer satisfaction or brand equity research you can often start by identifying these key concepts by reviewing the strengths and weaknesses associated with your brand or product, or by listing the advantages and challenges that you believe may be prevalent (e.g., good customer service, poor management, professionalism etc.). This is an a priori approach where the user/analyst identifies a few things that they believe may be important.

Bottom-Up

This approach requires you to use a more advanced text analytics software, like OdinText, to mark and extract concepts or themes that are most frequently mentioned in customers’ text comments found in your dataset and that are relevant to your brand or product evaluation (e.g., high cost, unresponsiveness, love). Better analytics software should be able to automatically identify important things that the user/analyst didn’t know to look for.

Top-down vs. Bottom-up

The top-down approach does not reflect the content of your data, whereas the bottom-up approach while being purely based on the data can fail to include important concepts or themes that occur in your data less frequently or is abstracted in some way. For instance, in a recent customer satisfaction analysis, very few customer comments explicitly mentioned problems associated with management of the local branches (therefore, “management” was not mentioned frequently enough to be identified as a key driver by the software using the bottom-up approach).

However as the analyst had hypothesized that management might be an important issue, more subtle mentions associated with the concept of management were included in the analysis. Subsequently predictive analytics revealed that “poor management” was in fact a major driver of customer dissatisfaction. This key driver was only “discovered” due to the fact that the analyst had also used a top-down approach in their text analysis.

It may be that some of the concepts or themes identified using the two approaches overlap but this will only ensure that the most important concepts are included.

Remember, that only when combining these two very different approaches can you confidently identify a complete range of key drivers of satisfaction or other important metrics.

I hope you found today’s Text Analytics Tip useful.

Please check back in the next few days as we plan to post a new interesting analysis similar to, but even more exciting than last week’s Brand Analysis.

-Gosia

Text Analytics Tips with Gosi

 

[NOTE: Gosia is a Data Scientist at OdinText Inc. Experienced in text mining and predictive analytics, she is a Ph.D. with extensive research experience in mass media’s influence on cognition, emotions, and behavior.  Please feel free to request additional information or an OdinText demo here.]

New Text Analytics Process Patented

Anderson Analytics’ OdinText Announces Patent for Powerful New Text Analytics Process  

Text analytics Patent Anderson Analtyics OdinText

Anderson Analytics today announced the granting of Patent No. 8,473,498 by the United States Patent and Trademark Office (USTPO) for the new powerful Natural Language Text Analytics process utilized in their OdinText software.

The approach leverages contextual data and provides a process for filtering out the noise which is so common in unstructured data. Both of these important benefits have been deficient in text analytics software until now.

"The problem with most approaches of text analytics out there is that they are focused on trying to do what humans do best rather than on what computers do best. They’re also focused on the individual document level, thus completely missing some of the greatest benefits that come from the increases in computing power, consideration of contextual information, and statistical techniques." explained founder Tom H. C. Anderson, adding "In our approach, it’s not just about unstructured (text) data any more. It’s often about ‘mixed data’, both structured and unstructured, and this approach takes advantage of that whenever possible".

OdinText can read and analyze millions of customer comments and other text data in a matter of minutes. The process is extremely fast and leverages the 100% consistency of coding inherent in text analytics. The approach works with non-English language text as well. An international patent application has been filed, designating a large number of countries, including the European Patent Office.

Marketing researchers can use OdinText to monitor and improve their customer satisfaction programs or understand key drivers from various other survey data. The software is also being used by customer service departments to analyze customer complaints, praise and suggestions received via telephone or email. Finally, OdinText is also extremely powerful when analyzing various social media and web based data regardless of whether foreign languages, slang or other acronyms are used.

In 2005, Anderson Analytics became the first marketing research firm to leverage text analytics and has been honored with several awards for innovation from trade organizations such as the American Marketing Association, the Advertising Research Foundation, and the European Society for Opinion and Market Research. The firm’s expertise ranges across several industries and includes traditional fortune 500 companies from Disney to Unilever as well as new social media giants such as Facebook and LinkedIn.

About OdinText - Text Analytics Applied!(TM)

OdinText is the first text analytics software platform developed by market research professionals for market research professionals. OdinText leverages a decade of Anderson Analytics’ experience in actual Applied Text Analytics(TM). Those interested in finding out more about OdinText including how to request a demo may do so at OdinText Info Request.

 

 

Leveraging Social Media with Text Analytics

Interesting Q&A about Social Media and Text Analytics Text Analytics News just published an interesting panel interview on the Social Media Summit website with myself, Dana Jacob who is Sr. Manager of Social Media Insights & Analytics of Yahoo!, Judy Pastor who is Principal Operations Research Manager of American Airlines, Usher Lieberman Director of Corporate Communications at TheFind, and Marshall Sponder of WebMetricsGuru.

I think it's great to hear how fellow market researchers on both the supplier and client side are leveraging social media with text analytics.

The four of us participated in the Q&A which is available on the Social Analytics Summit website (note: requires registration), but you can read part of the interview below. I understand this is the first part in a series of interviews Text Analytics News is doing. I strongly believe that we're just skimming the tip of the social media iceberg when it comes to more serious analytics and insights and it's important for those of us who have started taking deeper dives analytically to share some of what we have found, both good and bad.

Text Analytics News: What are the primary ways in which you leverage Social Media for your clients?

Pastor (American Airlines): American Airlines and our loyalty program, AAdvantage, have both Twitter accounts and Facebook pages to communicate with customers. Twitter is our first line of problem identification and is quickly becoming a major portal to our Customer Relations and Reservations departments for passengers on the "day of departure" (i.e. in the process of traveling).

Currently, we answer 700 tweets per day and that will increase as we ramp up our twitter agents to 24/7 by the end of first quarter, 2012. Inquiries and problems are also addressed on Facebook, though these tend to involve future travel plans. We also monitor the frequent flyer forums such as www.flyertalk.com and www.milepoint.com and reply to questions that may come up about our policies and procedures. Our Facebook pages also highlight fare sales, and special travel destinations and deals.

Anderson (OdinText): Anderson Analytics has been leveraging Text Analytics (now OdinText) to help our clients understand customer comments since 2005. Early on our focus was primarily on customer comments in survey open ends; this is still a key area for most of our clients. However since 2007 our clients have asked us to incorporate analysis of social media including discussions taking place on open forums such as Flyertalk.com. Twitter and other social networks are also a growing area of interest, however we try to help our clients to make educated decisions and prioritize all the unstructured data sources now available rather than trying to "boil the ocean". We've found a deeper analysis and understanding of the relative importance of these sources makes for a better analytics ROI.

Sponder (WebMetricsGuru): I'm an analyst, so the lens I use to leverage social media is through an understanding of online conversations and how they can be best categorized. I don't consider myself a marketer, and let other others do that for me, so I am not focused on marketing messages or how to best respond, but I do look at platforms that handle those functions for marketers and public relations.

As far as the particulars of how I leverage social media for clients, I look at the following categories

• Is it brand messaging? • Is it non specific, but pertaining to the overall industry? • Is it messaging pertain to a specific topic related to the client? • Is the messaging online representative of offline sentiment (often, we've found, it's not).

Lieberman (TheFind): TheFind is a shopping search engine with perhaps the only index of every online store and available product; roughly 500k stores and 500 million items for sale. We are increasingly using social media, specifically the social signals broadcast by the open graph, to influence our search results. Already, the most Liked stores and brands in your network are ranked higher and the products people Like are more visible. We expect the influence of social signals to accelerate as more stores and brands recognize the obvious SEO benefits of driving social signals such as Like and +1 down to the product level.

Text Analytics News: What kinds of social media analytics have you found most useful?

Jacob (Yahoo!): For our needs, text analytics is the foundation of social media analytics. Sentiment analysis is not sufficient as the bulk of social conversations convey a spectrum of emotions, therefore can't really be categorized as positive or negative. In order to extract actionable insights, it is critical to sort, filter, analyze and code the conversations into quantifiable categories that are meaningful to specific business context.

Even for engagement metrics, text analytics are key to understanding the key drivers behind the engagement.

Anderson (OdinText): We've looked at and worked with many of the providers out there from social media monitoring to pure play text analytics vendors. What we have found is that there is no one tool out there for every situation. The needs differ tremendously by both data source as well as use case. Many of the social media monitoring tools out there have focused more on Twitter data and on the public relations use case, while the pure text analytics tools have taken an almost too broad of an approach. This is the reason we developed OdinText specifically for market research managers, however we make no claims about our own tools being the best tool for every use case and every type of data.

We've found a discussion about what client objectives are, what data is most suitable to answer these is the best approach. Sometimes this means recommending a provider like SalesForce's Radian6 or other vendor, sometimes it means designing a custom ad-hoc study that may incorporate survey data, and other times OdinText is used. Regardless, in many cases we do find that further and deeper analysis is beneficial; these are usually conducted using one of the common statistical analysis packages out there whether it's something as basic as Excel or something more advanced like IBM's PASW Modeler, SAS, Latent Gold etc. It really depends on the need and the data.

Pastor (American Airlines): We track click through rates and referrals from our Facebook pages to our booking engine, www.AA.com, and attribute revenue to each. Number of issues solved via Twitter is also tracked. We are still looking for ways to quantify the softer side of SM - the good and the bad stories that our customers tell. We are currently working on tracking Likes, re-tweets, and shares...[Full Q&A on Social Analytics Summit site]

Thank you to Text Analytics News and Social Media Summit for organizing these Q&A panels. I'd love to hear how others are using social media in their insights programs.

@TomHCAnderson