Machine learning techniques such as our ACETM methodology provide a detailed and comprehensive evaluation of customer needs and insights at a particular point in time.
The conclusion of an engagement begs the natural question – how often should I conduct a machine learning study to gather consumer insights?
The answer depends on several factors.
- New Products/Services: In rapidly changing industries, machine learning engagements should be conducted more frequently. For example, virtual reality, biotechnology, renewable energy and cyber security are all examples of industries with rapid growth and advances. As these industries shift, so do consumer expectations, and new customer needs emerge.
Tip: Industries may be stable for a while and then suddenly experience rapid growth. Therefore, re-evaluating the pace of change in your industry on an ongoing basis is valuable.
- Emergence of New Players: As new players enter the market, industry dynamics shift. Consumer expectations change. Sometimes one entrant alone can transform an entire industry. For example, with the advent of Uber, expectations related to private transportation options evolved. Consumers suddenly formed opinions about on-demand booking, automatic payment, live tracking updates, driver reviews, and other factors that weren’t available before the new entrant. These changes altered consumer needs and expectations related to transportation and, as a result, new insights emerged.
Tip: Brainstorm with your team the best way to identify new emerging players in your field. There are a myriad of ways to find new emerging players including industry conferences and bloggers.
- Changes in Adjacent Categories: Sometimes it’s not your industry that’s changing rapidly, but adjacent industries with rippling effects on your industry. For example, when the iPhone was first introduced it not only impacted mobile phones but also the personal camera industry, the music industry, and others. When adjacent industries are rapidly evolving, keeping a pulse on user generated content becomes even more critical so you don’t miss important blind spots.
Tip: Machine Learning scales well for exploring adjacent categories – you can cost-effectively explore multiple adjacent industries as part of a single study.
- New Major Launches: Launching new products? For each major launch, plan to conduct a machine learning study pre-launch to get a baseline and post-launch to get an early read on customer feedback. It’s important to strategically select how long to wait post-launch. Depending on the product use cycle and the speed of adoption, you might wait as short as one month or as long as 6-12 months post-launch.
Tip: If the new product is a true game changer, wait longer to account for the adjustment period to get a true read on customer feedback. Remember, it’s human nature not to like changes, initially. If you evaluate the change too soon, you may get a misleading pessimistic read on the change.
In summary, if your company falls into any of the categories above, you’ll want to conduct machine learning more frequently.
How frequently? That’s as much of an art as a science.
Generally, in a climate of rapid change we’d recommend leveraging the power of machine learning every 3-6 months. If your industry is slower changing, keeping a pulse on user generated content every few years may be sufficient. Even if at first glance your industry is rather static, conduct machine learning periodically and you may be surprised by some unexpected industry changes that you don’t want to miss.
Contact us and we’ll help you assess the right cadence for your machine learning initiatives so you can keep a pulse on the consumer.