Customer Experience

AI, ML Not Yet a Plug-and-Play Proposition for Marketers: Study

Personalization and automation, two of the hottest buzzwords in the lexicon of CRM practitioners, are all the rage for marketers these days. But only 14 percent of organizations are using artificial intelligence and/or machine learning to automate their marketing campaigns.

A global survey Rackspace Technology conducted in January reveals that the majority of organizations worldwide lack the internal resources to support these critical high-powered CRM initiatives.

Rackspace, a multicloud technology solutions company, sought the status of AI/ML use in marketing in its survey “Are Organizations Succeeding at AI and ML?” conducted in the Americas, APJ, and EMEA regions of the world.

Responses indicated that while many organizations eagerly want to incorporate AI and ML tactics into operations, they typically lack the expertise and existing infrastructure needed to implement mature and successful AI/ML programs.

This study shines a light on the struggle to balance the potential benefits of AI and ML against the ongoing challenges of getting these initiatives off the ground.

While some early adopters are already seeing the benefits of these technologies, others are still trying to navigate common pain points such as lack of internal knowledge, outdated technology stacks, poor data quality, or the inability to measure ROI, according to researchers.

The most successful uses for AI/ML almost always deal with intelligence that helps companies understand their customers and their behavior, noted Jeff DeVerter, CTO of Rackspace Technology. The results from these projects are the easiest to directly tie back to revenue — either gaining new customers or better serving and retaining existing ones.

“The highest benefits our participants saw from AI/ML adoption is increased productivity (33 percent) and improved customer satisfaction (32 percent), he told CRM Buyer.

Some Survey Surprises

Three most significant findings about AI/ML adoption surfaced from the survey results. Researchers were actually surprised by a few things, noted DeVerter.

One is that more organizations seem to have an established plan for AI/ML and are actively developing AI/ML projects. The average spend per year is US$1.06 million. That, by comparison to total IT budgets, is small.

“It is not insignificant,” said DeVerter.

Another interesting finding is the percentage of companies who recognize that while they may not have established AI/ML practices today, they definitely view the tech as part of their future.

Only 17 percent of the participants stated they were approaching or had factory of model production. The survey found that 51 percent of the participants are exploring what AI/ML is and how to put it into production.

Meanwhile, 31 percent of the participating organizations are moving from pilot to an AI/ML solution in production.

Smart Goals Fall Short

Once the slow adoption rate is resolved, a more glowing vision for AI and ML awaits organizations. The success factor will not be realized until after the data has been curated.

“The best way to gauge success is how deeply business employees (not IT) start asking questions of the findings and then request more results,” said DeVerter.

This shows their trust in the results and adoption/acceptance of the technology as a real tool to help them do their job better, he added.

The survey results show less than stellar interest and efforts among many organizations regarding use of AI/ML technologies to enhance marketing efforts.

For instance, 30 percent of organizations are using AI/ML to create personalized customer journeys. Of current plans to use AI/ML, 36 percent of respondents want to understand customers better.

Current plans to use AI/ML target being able to deliver personalized content for customers for 33 percent of the responding organizations. Another 29 percent of respondents want to understand the effectiveness of marketing channels and content.

Failure Causes Common

One of the initial stumbling blocks to moving into adopting AI/ML strategies for organizations is getting out of the exploration phase. Potential adopters are still exploring how to implement mature AI/ML capabilities, the researchers found.

A mere 17 percent of respondents reported mature AI and ML capabilities with a model factory framework in place. In addition, the majority of respondents (82 percent) said they are still exploring how to implement AI or struggling to operationalize AI and ML models.

AI/ML implementation often fails from a lack of internal resources. More than one-third (34 percent) of respondents reported artificial intelligence R&D initiatives that have been tested and abandoned or failed.

The failures underscore the complexities of building and running a productive AI and ML program. The top causes for failure were nearly evenly divided among four categories:

  • Lack of data quality (34 percent)
  • Lack of expertise within the organization (34 percent)
  • Lack of production-ready data (31 percent)
  • Poorly conceived strategy (31 percent).

Untapped Smart Benefits

Successful AI/ML implementation has clear benefits for early adopters, according to the report. As organizations look to the future, IT and operations are the leading areas where companies plan on adding AI and ML capabilities.

The data reveals that organizations see AI and ML potential in a variety of business units. Among them are IT (43 percent), operations (33 percent), customer service (32 percent), and finance (32 percent).

Further, organizations that have successfully implemented AI and ML programs report increased productivity (33 percent) and improved customer satisfaction (32 percent) as the top benefits.

Successfully reaching those benefits requires organizations to carefully define their key performance indicators (KPIs). That is critical to measuring AI/ML return on investment, noted Rackspace.

Along with the difficulty of deploying AI and ML projects, comes the difficulty of measurement. The top key performance indicators used to measure AI/ML success include profit margins (52 percent), revenue growth (51 percent), data analysis (46 percent), and customer satisfaction/net promoter scores (46 percent).

Hopping Over Adoption Hurdles

The hurdles to smooth adoption of AI/ML technology are fairly consistent. Organizations need to do one essential thing to more quickly get beyond all the barriers, according to DeVerter.

That one key thing is to establish a data office to oversee the validity of the data used. AI/ML is absolutely beholden to the source data to which they apply their machine learning models.

“As such, AI/ML projects can become suspect if the source data is not cleaned and validated by a data office,” said DeVerter.

He noted that 34 percent of participants stated their R&D projects failed due to lack of data quality. Also, 31 percent said it was due to lack of production ready data.

“Unfortunately, not all companies have this office or its equivalent whose mission is to validate and curate approved corporate datasets. With the success of early AI/ML project, companies must establish the data office role in tandem to their AI/ML projects,” DeVerter explained.

In-House or Outsource Efforts?

Many organizations are still determining whether they will build internal AI/ML support or outsource it to a trusted partner, according to the survey results. But given the high risk of implementation failure, the majority of organizations (62 percent) are, to some degree, working with an experienced provider to navigate the complexities of AI and ML development.

“In nearly every industry, we are seeing IT decision-makers turn to artificial intelligence and machine learning to improve efficiency and customer satisfaction,” said Tolga Tarhan, chief technology officer at Rackspace Technology.

Before diving headfirst into an AI/ML initiative, organizations should clean their data and data processes, he reiterated. In other words, get the right data into the right systems in a reliable and cost-effective manner, he explained.

To address adoption obstacles, the majority of organizations (62 percent) work with an experienced provider to navigate the complexities of AI and machine learning development. This solution gives organizations access to expertise and technology that can accelerate development and increase the overall success of a project, according to the report.

The Cost of Customer Knowledge

As noted earlier, organizations adopting AI/ML strategies spend an average of $1.06 million per year on initiatives. That spend is spread across current and planned projects to grow revenue, drive innovation, increase productivity, and enhance user experience.

The most common ways that businesses reported using AI and machine learning functionality are as a component of data analytics (40 percent), a driver of innovation (38 percent), and through its application to embedded systems (35 percent). These point to the need for businesses to innovate and spur differentiation, and illustrate how AI and ML technologies can be used to drive an innovation engine.

That spend also supports upcoming AI and machine learning initiatives, the report noted. AI and machine learning projects currently in the planning phase lean more toward customer experience enhancements, with four of the top ten ranked areas specifically focused on improving these customer relationships:

  • Offering new services (38 percent)
  • Understanding customers better (36 percent)
  • Delivering personalized content for customers (33 percent)
  • Understanding the effectiveness of content marketing channels and content (29 percent)

Automation to the Rescue

Citing one key element in the marketing battle — constantly changing supply and demand — Mark William Lewis, CTO of Netalico Commerce explained how automation can aid marketers and retailers.

“With the ever-changing buyer environment, retail marketers have to adjust strategies on the fly. The best way to stand out against competitors is to use intent data to inform the message, medium, and timing of marketing touchpoints,” he told CRM Buyer.

For example, after a consumer views a backpack on a retailer’s website, intent data can automate an email reminder with comparable backpack options. If there is no response to the email, an automated piece of direct mail can be triggered that highlights the backpacks and a 25 percent off coupon, Lewis explained.

Survey Dynamics

The survey occurred between December 2020 and January 2021. It is based on the responses of 1,870 IT decision-makers across manufacturing, digital native, financial services, retail, government/public sector, and healthcare sectors in the Americas, Europe, Asia and the Middle East.

A copy of the full report is available here. Before downloading the report, you must fill in a form with your name, email address, and company affiliation. No promotional consideration or transmission of data from Rackspace is received by this publication, or its parent company ECT News Network, when our readers download the report.

Jack M. Germain

Jack M. Germain has been an ECT News Network reporter since 2003. His main areas of focus are enterprise IT, Linux and open-source technologies. He is an esteemed reviewer of Linux distros and other open-source software. In addition, Jack extensively covers business technology and privacy issues, as well as developments in e-commerce and consumer electronics. Email Jack.

Leave a Comment

Please sign in to post or reply to a comment. New users create a free account.

More by Jack M. Germain
More in Customer Experience

CRM Buyer Channels