The National Retail Federation’s Big Show attracted tens of thousands of attendees . Far from resigning themselves to conquest by market places they are determined to make the right investments to thrive and survive.
Three years ago we found retail CIOs most concerned about innovation and security
https://www.forrester.com/report/The+Retail+CIO+Agenda+2015+Secure+And+Innovate/-/E-RES121188
At the NRF show in 2018 it was evident that they are still looking to innovation to drive customer engagement and a healthy bottom line.

As store and online shopping converge, retailers need innovative applications to deliver distinctive customer intimacy and a choice of exciting new merchandise.  The NRF 2018 show convinced Forrester that:

1) Myriad point solutions will coalesce into retail applications ecosystems. Retailers, especially those selling branded merchandise, differentiate through process and experience, so they are reluctant to consign all their applications to a single vendor. They have to experiment with point solutions and in-store gadgets, but face challenges to scale and integrate the winners into their legacy enterprise stack. Emerging retail applications ecosystems help retailers to capitalize on opportunities more quickly by delivering preconfigured integration between leading solutions in each domain.
For example, Dunnhumby and JDA demonstrated ‘transformational’ assortment planning, applying Dunnhumby’s powerful customer segmentation to JDA’s proven assortment planning and execution platform. Salesforce Commerce Cloud demonstrated its ability to integrate Instagram and the Salesforce Marketing Cloud.

2) (Almost) every process now includes embedded “AI”.
For years retailers lacked the infrastructure to store large volumes of granular data such as basket analysis, click stream, or social data. They used weighted moving averages to predict sales at the SKU and store level – a bit like driving while looking only in the rear view mirror. They lacked the computing power to develop, test, and run predictive models. They lacked trust in the predictive power of data generated by millions of “natural experiments”, recording the outcome in terms of sales uplift from combinations of causal factors. But widespread adoption of cloud computing in retail opens a new world in which applications can recommend pricing, promotion, assortment, and space plans, based on systematic understanding of multiple independent causal factors. Somewhat confusingly, vendors and retailers lump into the term “AI” multiple capabilities, driving a transition from descriptive to prescriptive retail management systems.

3) Stores are becoming logistical hubs as well as showrooms. Retailers like Rent the Runway are transforming stores from showrooms for item sale or rental to logistical hubs for unlimited subscribers to drop off items and check out new ones, all through self-service tablets.