Most companies that aspire to disrupt their sectors need either large economies of scale (like Walmart Inc.) or a powerful network effect (like LinkedIn), but the rise of artificial intelligence (AI) offers a third option—data analytics—a speaker said Wednesday at a conference of freight brokers and third party logistics (3PL) providers.

Data analytics can reveal valuable patterns and predictions that provide a competitive advantage for firms as they enter "the mysterious world of a thinking business," said Paul Zikopoulos, a technology book author and vice president of cognitive big data systems for IBM Corp.

However, even if they can collect the data, most companies face a major challenge in figuring out how to leverage it, Zikopoulos told attendees at 3PL Technovations, a conference held in Tucson, Ariz., by the Transportation Intermediaries Association (TIA). "Most companies have an A+ in data collection, but only a C- in data decisioning," which is the process of using the data to make better choices, he said.

Without a sufficient strategy for applying AI and algorithms to that data, businesses will simply obscure their own results by drowning them in data. "Did we really think that adding more hay to the haystack would help us find the needle?" he asked, in reference to industry's failure to produce promised results of generating significant conclusions from "big data lakes."

To avoid that problem, users need either better algorithms or more creative sources of data. "You have to both use the data you have and understand that data will come from places you never imagined," Zikopoulos said. He cited examples of valuable data sprouting from unexpected sources, including: the rising use of What's App text messaging files as a tool to prove infidelity in Italian divorce courts; the case of a dermatologist who diagnosed an actor's thyroid cancer simply by seeing her appearance in a reality TV show; and rail line BNSF's practice of detecting wheel cracks by analyzing the sounds of trains passing over tracks.

According to Zikopoulos, companies that have leveraged that unexpected data for profit include the Canadian cell phone network Rogers, which uses in-car wifi nodes to offer additional services such as geofencing for tracking subscribers' teen drivers or analyzing telematics data to predict maintenance issues. Another example he cited is the hair products vendor Pantene, which has boosted its sales of certain shampoos and conditioners by analyzing weather patterns and prompting individual customers to buy products to help them handle common challenges like the "frizzy" hair caused by humidity or the "fried" hair caused by dry weather.

Retail experts acknowledge that some of these applications may strike current consumers as an overly aggressive use of private data, but technology may also have an answer in the form of blockchain applications for secure information sharing. "What blockchain will do for trust is what the internet did for search," Zikopoulos said.

Because AI is still a young, fast-growing technology, most users will have to progress toward these goals by taking a series of small steps, he advised. That patient approach can help companies start with AI based on "supervised learning"—where humans "teach" computers to build algorithms for situations like distinguishing between alphabet letters in different fonts or between facial recognition images of similar-looking celebrities like comedian Will Ferrell and Red Hot Chile Peppers drummer Chad Smith, he said.

More advanced AI applications subscribe to "unsupervised learning," where computers themselves apply machine learning techniques to determine for themselves which variables produce the best results, and generate mathematical models and algorithms to track them, said Zikopoulos.

Whoever applies them first, the principles of AI will soon make an indelible mark on the business landscape, delivering profit to some users and leaving others on the sidelineshe said. "This is 'lift, shift, rift, or cliff' territory; there will be winners and there will be losers," Zikopoulos said.