I love data. And I love visualizing data.
And as someone who has a tendency to treat external expectations as an obligation, I especially love being able to respond quickly to those one-off, C-level questions with the number or the visual from which they can simply… decide. Tessitura Analytics, the analytical business intelligence and dashboarding features of the Tessitura platform, has been a powerful tool for my peer obligers in the Tessitura community (either by tendency or job description), because it enables them to produce amazing data stories for their organizations.
After an exhaustive review of the BI platform market in 2016, Tessitura Network found and partnered with Sisense, finding it the most closely aligned with the needs of the Tessitura community. In 2018, we released Tessitura Analytics, powered by Sisense. Since that time, Sisense has been repeatedly recognized as a Visionary on Gartner’s Magic Quadrant for Analytics and Business Intelligence Platforms. And as Sisense vigorously pursues Artificial Intelligence (AI) and Machine Learning (ML) solutions in their products, the Tessitura community benefits from that growth.
I’m going to tell you about a couple of AI and ML features you may already be using in Tessitura Analytics, and give you a preview of others we’re working on for the next version of Tessitura.
Tessitura Analytics alerts notifying of anomalous changes in data
Already, in Tessitura Analytics, we can set up data alerts that notify us of anomalous changes in the data — not just how many sold yesterday, but that yesterday we sold more or less than usual. In this case, what is “usual” is determined by machine learning algorithms, identifying anomalies, and learning over time how to make those identifications more and more accurately for a given metric. We then benefit from noise reduction, receiving notifications not just when the data changes, but when it changes in a way that might legitimately warrant our attention.
Tessitura Analytics already offers simple trendline analysis. Following the behavior of our patrons over time, we can extend that trendline several periods past our dataset, and predict the next few periods. In our Tessitura v16 release, however, we’ll introduce truly predictive forecasting. The advanced artificial intelligence uses a blend of algorithms to produce the most accurate results, and if desired, allows you to adjust the respective weights of the analytical models used in the blend. You’ll be able to override and fine-tune the forecasts, compare them against your actuals, and even specify anomalous time periods to ignore (like a holiday or, ahem, COVID).
Exploration paths: trends over time
Also in Tessitura v16, we’ll introduce machine learning that anticipates your questions about a given KPI. The artificial intelligence generates a set of break-by and trend-over-time analysis visuals of that KPI. The visuals are ranked by the AI’s observation of your organization’s behavior in Tessitura Analytics. This will reduce the need for dashboard designers to build more visualizations manually when business users ask questions about why a number is what it is.
And finally, we’re exploring a new machine learning feature to explain why a trend has changed in some way. Looking at revenue over time, let’s say, you notice a bump in the trend. As of now, if you want to determine what caused that bump, a business analyst needs to make some assumptions and hypotheses (using their gut and savvy), to determine what may be the contributing factors. They follow their intuition, looking for a correlation between the revenue bump and other noticeable changes in data-trends. Tessitura Analytics in the future could skip the assumptions, the hypotheses, the intuitive research, and instead present you with real data that corresponds to your bump in revenue. A selection of visualizations, ranked by the AI’s perceived impact on your revenue, will be served up for you to confirm as impactful or dismiss as coincidence, and ultimately, allow you to make the best response to change more quickly.
I’m so eager to bring these artificial intelligence and machine learning solutions into the Tessitura community, and to see how our creative members leverage these features to improve their businesses and their relationships with their patrons.
Top photo by Michael Dziedzic on Unsplash