With customer experience extending across so many parts of an organization, CX tech stacks are understandably complex. For marketing leaders in charge of CX, these entangled systems make measuring ROI across the CX tech stack all the more difficult.
Complicating things further is the potential integration of generative AI. Generative AI offers ways to make real improvements in CX, driving new levels of personalization and efficiency, but it also raises the stakes for delivering results.
There's a broad willingness to embrace generative AI. Nearly two-thirds of marketing leaders intend to invest in generative AI in the next 24 months, according to a Gartner survey.
However, the emergence of this new, innovative technology has created a trade-off for CX leaders between investing more in the existing tech stack to drive utilization on current tools and reallocating existing budget towards new generative AI applications.
“CMOs recognize both the promise and challenges of generative AI,” said Benjamin Bloom, VP analyst with Gartner, in a company statement.
But CX and marketing tech utilization continue to decline, driven by the complexity of the ecosystem, customer data challenges and inflexible governance — and there’s a pressing need to simplify.
What has become an unwieldy stack in many cases is ripe for paring back. The good news is that it also creates headroom for investing in new and emerging technologies like generative AI, according to the research firm.
“Cutting underused technology within the current stack can preserve some ‘dry powder’ for transformative applications that aren’t yet generally available,” Bloom said.
Components critical for delivering CX stack ROI
For CX leaders, assessing ROI is first and foremost about keeping the focus on customers and business use cases, according to Chief Martec’s Martech for 2024 report.
The question needs to be, which features are going to drive value?
The heart of the CX tech stack is the customer relationship management or customer success platform for centralizing customer data, along with tools for analyzing metrics and analytics to reduce churn, according to Shannon Nishi, director of customer success at Customer.io.
The goal is a unified view of the customer, personalized communications and customer support. Then, depending on the unique needs of the organization, it usually includes tools for customer service, reporting on churn, and providing analytics for CX metrics to improve.
“It can also include a tool for scheduling and recording calls, and one for building community,” Nishi said.
Data is critical in identifying patterns and trends across customer segments that drive personalization and help with predictive capabilities to identify issues before they have too great an impact, said Nishi.
“CX leaders can utilize behavioral analysis, predictive analysis, customer journey mapping, feedback analysis, A/B testing and churn prediction to gain insights into customer behavior and the actions to supplement CX,” she said.
However, aligning data cross-departmentally can be a challenge for CX leaders because there are so many moving parts, according to Nishi. “The intricate nature of these systems can make implementation, integration and overall management time-consuming,” Nishi told CX Dive.
Guiding all of the budget decisions is ROI, which can be challenging to prove with some solutions implemented to lift subjective measures such as improving customer satisfaction.
To help address this challenge, there needs to be agreement on key metrics. “It’s important to align stakeholders on concrete definitions for measures of success and minimize the number of factors that could influence any given indicator,” she said.
CX tech stack complexity
With the ranks of martech solutions swelling to more than 11,000 applications in 2023, according to ChiefMartech’s latest snapshot, it’s no wonder many CX stacks are suffering from bloat.
We’re at an inflection point, driven by the sudden rise in generative AI that could impact code generation, software, apps and content, according to Scott Brinker, editor at Chief Martec and HubSpot VP platform ecosystem.
While there’s some essential complexity because CX links to most parts of the organization, there’s also unnecessary complexity in these environments, said Brinker.
“There are a lot of moving parts to CX, and almost by necessity that means having different technologies that are related or contribute a component of it,” Brinker told CX Dive.
“A lot of CX leaders also want to rationalize the stack, but for a lot of companies at scale, it’s hard to simplify it down because one product won’t do everything they want,” he said.
It’s vital to have consistency across all the different customer touchpoints and monitoring tools for evaluating CX, but this is a constantly moving target, according to Brinker. “CX is driven by customer expectations, and as they evolve, new technologies will enter into the consumer environment, which naturally impacts the CX tech stack.”
CX tech must deliver real value to the business
In the face of this growing push for AI, CX leaders and their marketing peers face a paradoxical tug of war: squeezed between demands for efficiencies and ROI driving stack consolidation on the one hand, and disruptive innovation triggered by new tech like AI and ever-evolving customer behaviors demanding investment on the other hand, according to Chief Martec’s report.
To prioritize competing demands, business case decisions need to be framed around customer experience value. This will increase the need for streamlined, efficient access to customer data, although it could also pose some hurdles for organizations that may not have fully desiloed data.
“A cloud data warehouse is crucial for consolidating data and enabling analysis across various customer touchpoints, but there must be a mechanism to pull all of it together,” Brinker said.
With chatbots, for example, conversational technology is available, but they need to be connected to the right data to have the right information to help address customer questions. “It’s a matter of feeding the appropriate information to the chatbot, but you may have to do a better job of managing the data,” he said.