Making Data Work – Part II
This past year our team released our first edition of the “Making Data Work Newsletter”. In the newsletter we explore hot topics in the tech and MarTech space, exciting developments at Softcrylic and other nuggets of content that hasn’t been released in other places. Initially, we release this to our newsletter subscribers, so they get the information first! Now in the new year, we want to share it with you all! In Part Two we will go into more detail around what we have been working on in 2022 and the products and services highlighted based on the articles in Part One. Don’t miss out on the next edition, subscribe to our newsletter here. Enjoy!
What are we working on
Client Propensity Model
The Data Science & Analytics team at Softcrylic has had many opportunities to assist our clients with turning their data dreams into reality. One such opportunity that came about this year began as a request to support one of our clients in the construction and deployment of a new propensity model. The scope of this new project required our knowledge and expertise every step of the way from inception to post-deployment testing and validation. In less than six months’ time, we were able to help our client stand up their own internal propensity model to begin targeting the audience output with real marketing campaigns and analyzing the results of a predictive model.
The main goal of the project was to generate learnings about customer performance data and enable automated deployment of target audiences in support of achieving key marketing business objectives. Our team’s expertise was critical for the development of the hypothesis testing structure and incrementality testing requirements necessary for measuring success of this new propensity model. The main hypothesis we needed to test for the client was if the propensity model produced a better audience for activation on premium marketing campaigns. The audience identified by the propensity model in fact did outperform all other traditional audiences that were deployed during the marketing campaign across all major KPIs like conversion rate, click-through rate, and unsubscribe rate. The results showed that we were able to use machine learning to produce an audience that would be more receptive to premium marketing messaging and who would have a higher likelihood of completing the conversion funnel towards the ultimate marketing goal of making a purchase of a premium product.
Overall, the client was thrilled with the results of the first iteration of the propensity model and has engaged our Data Science & Analytics team to continue consulting on future enhancements and improvements to the propensity model. We are currently in the process of reviewing and analyzing the results of the second deployment of the audience in a different marketing channel with additional paid media channel deployments tentatively scheduled in Q4. Our team looks forward to sharing updates with you (possibly even in the next newsletter!) and feel free to reach out to us if you have any questions about propensity modeling or how we can assist you in the development of your own machine learning models.
Note: We will be releasing an extended, more detailed write-up of our Propensity Modeling work as a blog post in the near future. If this interests you, stay tuned!
Email Survey Sentiment & NLP
One of the key problems we solve at Softcrylic is advising clients how to leverage “unused data” in effective and insightful ways. For the past year, we have been leading a project for one of our clients that took previously unused email survey data and transformed it into a pillar of their customer listening initiative. By utilizing natural language processing (NLP) and sentiment scoring algorithms, we were able to take a burdensome amount of seemingly random free-response feedback and leverage that text data into meaningful insights that identified the most common discussion topics regarding various areas of the business and measuring how customers felt about those topics. This allowed the client to identify pain points customers were experiencing with the business and make corrections in a timely manner before small issues had the opportunity to develop into larger, persistent headaches dragging down their company NPS scores.
When the problem was first presented to us, the client was collecting tens of thousands of unstructured text feedback responses each month from the surveys included in their marketing and transactional email deployments. At the time, it was prohibitively inefficient to have someone try to individually read through such a high volume of text data – it would take too much time and the effort required to try classifying each response manually would be mind-numbing to say the least. After consulting internally across disciplines on the Data Science & Analytics team, we were able to develop a machine learning and AI solution that leveraged developments in natural language processing for the client’s unused data problem. According to IBM:
NLP combines computational linguistics—rule-based modeling of human language—with statistical, machine learning, and deep learning models. Together, these technologies enable computers to process human language in the form of text or voice data and to ‘understand’ its full meaning, complete with the speaker or writer’s intent and sentiment.
The implementation of NLP for this unused data problem was key to streamlining one of the clients’ marketing analytics operations, increasing analyst productivity, and simplifying the client’s business objective of improving their “customer listening” abilities. The solution our team developed for the client has now been part of their regular reporting structure and delivered to business leadership every month for the past year. The benefit of being able to analyze complex customer feedback data quickly and efficiently has given our client a new tool in their effort to remain an industry leader in customer service and continually delivering a world-class customer experience.
Note: We will be releasing an extended, more detailed write-up of our Email Survey NLP work as a blog post in the near future. If this interests you, stay tuned!
Product and Service Highlights
We highlight new and continuous developments to Softcrylic’s services and offerings.
Introducing Softcrylic’s New Incrementality Workshop
Research regarding the benefit of using randomized control tests to evaluate marketing success have shown that they can assist marketers in making improved decisions between 25-50% of the time. This is one reason why leading companies like Facebook, AirBnB, and eBay have already made the switch to incrementality measurement as their key tool when evaluating the success of their marketing efforts. Softcrylic’s Incrementality Workshop is designed to help you understand and implement incrementality measurement for your business. To learn more about our workshop and the benefits we can provide for your marketing team, click here.
Dashboard In A Day
In partnership with Microsoft, Softcrylic is pleased to offer the Dashboard in a Day (DIAD) Workshop – A complimentary one-day workshop that covers the dynamic capabilities of Power BI.
Our Data Science & Analytics team will guide you through the Power BI platform and accelerate your understanding of how to bring life and value to your data in a rich and compelling way.
The skills you develop in this workshop will allow you to…
DIAD is a great opportunity to gain practical hands-on experience. You will walk away with foundational knowledge and skills that will make you more efficient and will greatly increase the impact of the work you do every day.
Keep an eye on upcoming DIADs on our Event Calendar here.
Never miss another newsletter again, subscribe to our monthly newsletter to stay up to date on industry news, Softcrylic events, content and more! Subscribe to our newsletter here.
