Birdbox, a thriller starring Sandra Bullock, sucked in 45 million viewers in its first week of release. Bandersnatch, part of the Black Mirror franchise, set new standards for interaction.

Netflix, which reports fourth-quarter earnings on Thursday, is quietly transforming the entertainment industry with data.

In fairness, the Los Gatos, Calif., company was a data business long before it turned to content creation. Founded in 1997, Netflix began as a subscription mail order DVD company. Back then, distributing the shiny discs in prepaid mailing pouches was so disruptive it brought Blockbuster Video, a nationwide Goliath, to its knees.

But the strength of the company was predictive analytics.

Netflix software engineers developed algorithms to steer customers away from high demand blockbusters, toward its plentiful, lesser known library titles. The strategy was a huge success. Through the years, mail order logistics have been replaced with online streaming, then content creation. Data analytics played a key role during every phase.

House of Cards, a Netflix original series, debuted in 2013. The company shocked the industry by committing $100 million to the edgy political thriller. The best creative talent was secured, but success was never really in doubt. Executives used big data analytics to inform all of the most important decisions.

According to a New York Timesreport, company executives offered actor Kevin Spacey and director David Fincher contracts without even seeing a pilot. Netflix algorithms showed the end product would be successful based on the subject matter, remnant fans of the original British television series and the appeal of Spacey and Fincher to the larger database.

The same factors were at work with Birdbox. Based on the Josh Malerman novel, the apocalyptic film got Bullock as the lead, a holiday release date and plenty of free advertising inside the Netflix application. The result was huge viewership. The official Netflix Film account tweeted that the film was watched by 45,037,125 accounts during its first seven days.

The casting seems to break all of the traditional rules. Bullock is 54 years old. Malorie, her character in the novel, is described by Malerman as young, athletic and reluctantly pregnant. Tom, her interracial love interest, is played by Trevante Rhodes, a 28-year-old rising star.

Critical and online user reviews have been mixed. The film is successful because it plays well to a carefully targeted audience. It shows the true power of data analytics.

Bandersnatch is important in a bunch of other ways. It takes the unusual route of asking viewers to choose the plot lines. Selections are made using any standard remote control. Each choice brings a unique narrative branch, where new choices later appear.

Netflix began experimenting with these interactive stories in 2017. The first two shows, Puss in Book: Trapped in an Epic Tale, and Buddy Thunderstruck were aimed at younger audiences. In a corporate press release, Netflix promised users would get to choose their own adventure experience.

The real winner is Netflix. With its wealth of Easter eggs and clever symbolism, Bandersnatch is the type of dark drama that screams for a cult following. Its puzzle-like structure opens up potential new business models.

Writing at the Verge, Jesse Damiani speculates that gathering user interaction data could become the basis for an internal programmatic marketing infrastructure. In addition to building data profiles to understand what plots work well with specific audiences, there are engagement and product placement opportunities.

It's easy to imagine marketing deals with Spotify  (SPOT) - Get Report  for music, or a consumer packaged goods conglomerate like Procter and Gamble (PG) - Get Report .

These new businesses are possible because of digital transformation. When media moved from analog to digital files that could be streamed to any screen, it created a continuous flow of metadata. Netflix managers have tirelessly collected and analyzed this information to build lucrative markets where none previously existed.

Those insights allowed it grow into the only worldwide streaming network. Today, the service has 137 million subscribers and is on pace to do $15 billion in sales through fiscal 2018.

At 65x forward earnings, Netflix shares are not cheap. However, it's rarely been cheap. Investors have paid up for the stock because of its dominant position in media, and the growth potential.

This is a stock longer-term investors should continue to buy into panic declines.

My new book is out, just in time for the holidays! It's called, "Fast Forward Investing: How to Profit from AI, Driverless Vehicles, Gene Editing, Robotics, and Other Technologies Reshaping Our Lives". Click here to buy a copy at Amazon, where it is listed as the #1 New Release in Business. And if you like it, please leave a review and rating.

To learn more about my recommendations at the crossroads of culture and technology, check out my daily newsletter Strategic Advantage.

To learn about my practical research in the short-term timing of market indexes and commodities, check out my daily newsletter Invariant Futures.

The author does not own any stocks mentioned in this column.