Dive Brief:
- After investing in artificial intelligence (AI), decision-makers say they expect to see that technology impact their revenue, improve customer experience or innovate products, according to a survey of 1,001 executives from PwC. The survey did not specify executive's ideal time frame for return on investment.
- Deal value involving AI companies have steadily climbed for the past five years, hitting $1.6 billion in transactions — including acquisitions, IPOs and venture capital funding — during 2018. That's up from just $44 million in 2013.
- Though buying companies that offer AI solutions can save time when compared to in-house development cycles, the report warns acquisitions could also mean difficulties in customizing systems to specific use cases, or exposing companies to "vaporware," technology that's been advertised despite not being available yet.
Dive Insight:
Millions of dollars in venture capital or acquisition costs come at a price: Executives need artificial intelligence to begin delivering on its promises of increased worker efficiency and product innovation.
But if checks are signed too quickly, companies may not find the ROI they would expect waiting for them at the end of the process.
PwC analysts recommend AI technologies must first be evaluated by a diverse team of experts who can attest to the technical backbone of the company, asses if the needed data sources can be improved over time, and determine if AI systems can applied to a company's unique business model.
"Moreover, integration plans should consider responsible AI principles and practices to extract a deal's value well after it closes," the report states.
A laser-tight focus on business objectives can help maximize ROI for companies when it comes to AI initiatives.
"Companies need to define goals or outcomes for its AI initiatives," said Soumendra Mohanty, EVP at India-based L&T Infotech. "And, companies need to be careful of expanding AI's scope to achieve incremental benefits, as this comes at expense of increased costs and complexities."
