DATA AUDIT & STRATEGY
Some, usually smaller to mid-sized, firms, are new to analytics. They have heard it might benefit them, but they don’t know where to start. If you recognize this as your firm, a good place to start is a data audit. The Kwantum team will meet with and make a data inventory: a list of all the data you have access to. These could be customer transactions, sales records, customer service records, etc. We review the data quality of these key data assets and we will write a review and we outline what can be done analytically. We keep our report focused on those analytics that the firm can concretely and immediately benefit from.
Do you have questions about what approach to use, or if more can be extracted from the data? Do you just have a dataset and need someone to help you re-analyze? We can do this for you. We are trained and experienced in a wide variety of statistical and machine learning techniques. Anything from regression, to segmentation analyses, to marketing mix modeling, and much more.
We are trained and experienced in a wide variety of statistical and machine learning techniques. Anything from regression, to segmentation analyses, to marketing mix modeling, and much more. We have applied regression, logit models, cluster analysis, latent class analysis, Bayesian models, neural nets, decision trees, and much more.
Driver analysis is a wide set of analytical tools to assess what variables are driving a metric of interest. For example, what attributes of our product or service are most important for customer satisfaction, or the likelihood they recommend you. What are the factors that determine whether a consumer prefers your brand versus a competing brand, what determines the appeal of a new product concept, etc.
Do you have a particular hairy problem? Too much data, and no hypotheses. Have standard approaches not worked? Maybe it is time to try out machine learning, neural nets including deep learning.
Market segmentation is hard as anyone who has been involved in a full-scale segmentation project know. It takes many iterations, reviewing many (sometimes hundreds) of possible solutions. It also takes activation efforts. Kwantum uses latent class segmentation methods, as these have been shown to perform best (see our paper on this). We also use the concept of core members. Ask us about why this makes a big difference.
Do you need to interact with your data? Do various people in your firm need a different slice to look at. Consider putting your data in a database with a visual user interface.
DECISION SUPPORT SYSTEMS
Do you need to interact with your data? Do various people in your firm need a different slice to look at. Consider putting your data in a database with a visual user interface. Do you want to look at different market scenarios? We can built a simulator for you that does all that.
Brand analytics can entail anything from creating brand maps using multidimensional scaling, to regression to identify brand drivers, to conjoint to quantify brand value. We also offer a unique approach to analyze brand associations.
We provide customized training for your team. This can range from a half or full day general analytics training to more specific target training.
Find out who your switchable and at risk consumers are. Check out our papers section to find a paper that outlines how we do this.
SWITCHABLE CONSUMER ANALYSIS
Mass marketing is not always a good idea. If the goal is to get consumers to switch to your brand it is very worthwhile to recognize that there are some consumers who are very unlikely to switch to your brand. They have just a very strong preference for a competing brand. However, there also consumers, who though they may prefer a competing brand, they may like your brand too. It doesn’t take very much to get to switch. We call these the switchable consumers. Learn how to identify them by reading our paper on this topic. This approach is one of the most successful approaches we have applied.
Text analytics can be done manually. However, when you have 20,000 survey respondents who all have responded to an open-ended question, manual analysis becomes impractical. Kwantum has a proven approach to that combines manual interpretation with machine learning approaches (topic modeling). We also have integrated the results of a text analysis into a quantitative driver analysis.
BRAND ASSOCIATION DENSITY
At Kwantum we are able to do all sorts of brand analytics from brand equity measurement (using discrete conjoint choice techniques), to brand mapping using multi-dimensional scaling techniques. We also have a unique brand equity measurement approach, which we have coined brand association density. See our paper on this topic.