When I was a kid, the idea of what “data” was evoked a vague image of a computer (think Commodore 64). Later on in school I learned about relational databases (hint: it’s a bunch of tables!). Now, further along in my career, with the rise of Big Data and accompanying technological advances, I’m fortunate to be in a position to operate on large datasets in a variety of formats. And help people get out of “spreadsheet hell” by digging in to data.

Data is King

Content used to be king. But content creation is most effective when informed by data. So now, data is king. Perhaps in the not too distant future, AI will be king, but we’re not there yet. Simply put, data is empowering. Data is instructive. Therefore, data can be lucrative. And while the term Big Data certainly implies a lot of volume (and probably some eye rolls) it’s really more about the actions taken on that data as opposed to the size of the dataset itself. That brings us to analytics, which describes the process of drawing conclusions out of all that big data.

Smart organizations of all sizes use analytics to act upon their data, and their customers’ data, to discover trends and patterns that may not be obvious. This is expressed with charts, visualizations, and infographics that tell the story of the data and make it digestible in smaller chunks. When a business can tell the story of the data efficiently and effectively, it becomes much easier to demonstrate ROI. What follows is a direct relationship between how well a business knows their data, and how well they can serve their customers. This means that organizations need to design with data in mind. What does that look like? It can take many forms, from third-party companion analytics applications, to “bolt-on” analytics integrated via single sign-on, all the way to fully embedded analytics capable of responding to users’ unanticipated actions immediately at the point of thought.

How Does Wicket Gather Data?

Every day throughout the country Wicket’s suite of products generate raw data from our networks of connected smart screens and IoT sensors. In some cases, we augment our internal data with supplemental data from our data partners. In all cases, we transform this data stream into actionable insights for our customers as well as meaningful Business Intelligence (BI) for internal stakeholders. This translates into smarter business decisions.

How do we do it? With a dedicated team that includes data engineers and data scientists who have set up a flexible analytics pipeline that runs data workloads to extract, transform and load (ETL) pertinent data. The data makes a few stops along the way, such as an in-memory, relational, or document database. It may be semi-structured data, sitting in our data lake. And it likely gets “transformed” and ends up in our data warehouse to be used for later analysis. In total, ISM’s analytics pipeline culls the results of the advertising and other sensor data on our networks to produce BI for brands and venue owners/operators, for targeted advertising or perhaps venue management purposes.

For example our ​Audience product uses facial recognition to anonymously gather information about the people in a given area, maybe a stadium concourse or outdoor retail mall. It can measure the demographics of the group including age and gender, count the number of people in the area and determine who is looking in the direction of the sensor, as well as how much time they spent in front of the sensor. When the sensor is paired with a smart screen running a content management system (CMS) such as ​Publish​, we can better understand how that audience is engaging with the media running on the screen. This is because we match up the CMS’s proof-of-performance data with the sensor’s audience composition data by correlating timestamps and other distinct fields within the data. The combination of these datasets reveals insights about a brand campaign’s efficacy, such as what type of content may have attracted a particular demographic, as well as guide the brand’s decision on future campaigns.

How Does ISM Visualize Data?

I like to say analytics is the deliverable. If a customer is paying us to do something, we can use analytics as the proof that we did what we set out to do. And while our products are designed with data in mind, we also have a dedicated ​ISM Analytics​ product capable of packaging and providing that proof by visualizing data with key performance indicators (KPI), charts, graphs, and other metrics. This is data at your fingertips, via a web application or mobile device, wherever you are.

The “proof” is descriptive, and sometimes diagnostic, analytics. We’re always monitoring incoming data. What happened, where, and why. We continuously search for and report on anomalies in the data, alerting interested parties if something is missing or doesn’t look quite right, and may warrant a deeper dive. But ​Audience also provides predictive analytics, i.e. what will happen the next time? This is data science at work, using historical data to predict future trends.

Where Are We Going?

Software applications, platforms, and devices throughout the world are all generating huge amounts of data every day. ISM-generated data is of course a small fraction of the total. But as our concentration of data grows, our machine learning algorithms will improve, and we’ll be able to tell better and better stories with the data, in a fully automated way. This is prescriptive analytics, which will answer our customers’ question “how can we make it happen?” without human intervention. For example, a sensor measuring the person in front of it to be in a certain demographic group, and triggering the CMS to serve a targeted ad in real time. A small step towards a future where AI is king.