AI and machine learning in marketing analytics: A revenue-driven approach

Marketing analytics stands out as an ideal starting point for businesses to integrate machine learning and artificial intelligence. This is due to a confluence of complementary factors.

• First and most importantly, marketing analytics has immediate impact on revenue generation and the ability to build stronger customer relationships. This not only delivers tangible, high-value ROI, but it directly addresses the strategic issues that C-suite executives are most concerned with in their digital transformation journeys. 
• Second, there is a strong automation landscape in martech that already can deliver results at scale and has been heavily invested in over decades. There has never been a better time to see AI and ML not as features but rather as framing principles for the evolution of a martech strategy, and in so doing create better return on investment from martech overall. 
• Thirdly, the latest wave in AI and ML tools, especially in “upstream” data environments, are incredibly well-suited to breaking down the data silos that have unfortunately come to define the customer data landscape. These represent the most strategically critical collection of data assets available to today’s most competitive enterprises. The scaling of customer data with AI and ML, either through data creation or resolution, is therefore a catalyst for better marketing — marketing that depends on deeper, more actionable insights and more transportable, available data sets.

The clear adoption benefits associated with these capabilities have met their moment as businesses signal their willingness to embrace and invest, and as cloud technology companies accelerate the delivery of a new generation of tools designed with the non-technical user in mind.

In an ever-more competitive marketing landscape, the transformation pressure to move rapidly toward the scaled use of AI and ML in marketing will become exponentially higher with each passing quarter, and savvy enterprises will prioritize the cross-functional collaboration necessary to achieve acceleration.

Adopting AI/ML in marketing analytics is aligned with a powerful and durable trend

It’s hard to find a business function leader who isn’t excited about AI and ML’s potential for transformative benefit. Why wouldn’t they be? Recent BCG CEO research places AI as thematically competitive only with changes in overall consumer behavior and climate and sustainability as top opportunities for their businesses.

So, while it’s expected that every aspect of business will be transformed by these technologies in the coming years, marketing analytics holds a unique position in directly impacting revenue generation making it an evergreen top priority among executives. 

The effectiveness of marketing campaigns, customer targeting, and personalized messaging are value pools that have been targeted for years by major corporations, who had become extensively committed to procured data sets and the experts who interpreted them. Despite these challenges, scaled organizations took it upon themselves to set up audience teams, marketing analytics centers of excellence, and in-housing initiatives to build up their data war chests, following hard research that top-performing companies tackled the first-party data challenge with aplomb.

As stated by BCG in 2022 research:

One of the biggest drivers of digital maturity is the use of first-party data—data that companies collect directly from consumers, including browsing behavior, transaction history from a CRM database, and loyalty program activity. Sophisticated companies understand that first-party data is differentiating (because proprietary), relevant (it directly relates to the company and its customers), and consistently high quality (it comes from customers directly).

Almost 50% of leaders use first-party data to derive insights that create better customer value propositions, but just 5% of laggards do.

The shift toward data and analytics as a major focus area has thus been seen, and continues to be seen, as an imperative to meet the personalization and relevance demanded by today’s complex, fragmented consumer attention economy — said another way, the organizational leaders who have emerged to meet the transformation imperatives fully expect to stay committed to data and analytics transformations overall, and want to be at the very forefront of progress. In a recent Ernst & Young survey, 53% of respondents flagged investment in data and analytics as their number one investment priority, up from 35% just two years earlier.

The confluence of these factors makes the observable outcomes of marketing analytics highly attractive and easy to understand entry-points for adopting advanced analytics. The brilliant basics of optimizing marketing spend, identifying high-value customer segments, and tailoring campaigns for maximum impact don’t have complex stories that need telling to explain their value — and in so much as analytics teams are already tasked with delivering against these outcomes, the enablement of AI and ML amounts to little more than prioritizing the introduction of best-in-class capabilities to drive increased sales and revenue. 

Boston Consulting Group

AI and ML are crucial to driving evolutionary value from martech ecosystems

The executive level attention on the development of better organizational ability to create, analyze, and activate on proprietary data sets has generated huge investments in martech stacks, which themselves offer awesome potential for automation and scale.

The very valid perception that operating at scale with data, especially as enabled by ML, is a contingency for personalization is supported by positive proof points from within a variety of channels where automation is now pervasive, and where AI and ML are already transforming approaches to segmentation and delivery, especially as a hedge against volatility.

The downside to this big investment is of course cost and complexity — a decade’s worth of added cost has businesses scrutinizing their tech infrastructure ROI at a time when they are also feeling huge pressure to continue to innovate, which frames a crucial trend in the MarTech ecosystem – the composable approach. 

Simply stated, the platform-dominated martech ecosystem of the past decade has failed to overcome the ROI challenges presented by data poverty issues, data integration issues and diminishing performance returns. Companies are looking for lighter-weight, nimbler approaches that will allow those data challenges to be navigated over time on a lighter tech budget.

Composability, as an idea, addresses this challenge by contemplating data as a currency that is exchanged across a marketing application ecosystem — analysis tools, activation channel tools, etc. The currency is, in essence, stored, managed, exchanged, and distributed by a data warehousing technology, which makes data available to a diverse array of operational systems when and how the systems are best suited to use it, each system playing its role in an efficient harmonious way according to a composed strategy.

Powerful marketing analytics live natively inside these warehouse platforms, and thus a centralized strategy to develop marketing analytics is distributed and amplified to downstream connected applications.

In theory, such an approach has the potential of significantly improving ROI, by reducing systems integration costs, improving performance, and lowering licensing and storage costs — and it is ROI that remains at the top of executive minds even as they continue to invest. A recent Marketing Week survey fingered ROI as the number one effectiveness metric e-teams and boards cared about — even against a backdrop of compelling performance improvements in other effectiveness measures.

However, any savvy trend spotter would notice a crucial assumption of this approach, namely that the data that feeds this approach is “figured out” — accurate, available, and at scale.

This is where AI and ML are game-changers that put this idea of composability into overdrive. For companies who might balk that they are not yet far enough up their maturity curves with data to overcome the challenge of accurate, available, and scaled customer data to feed their marketing programs, enter two critical tech innovations: warehouse-native AI/ML technologies and composable, AI/ML-enabled CDPs

The power in these technologies as ROI drivers is four-fold:

1. They can identify patterns from a diverse data set and create aggregated data in the form of a prediction.
2. They function as a two-way connector to both receive and distribute data, making them highly effective components of a feedback loop that can power improved performance.
3. They circumvent the need to implement explicit data creation and testing strategies that are reverse engineered around the data limitations of a given operational system.
4. As a result of the above points, they lead to improved performance outcomes over a shorter period of time. 

The upshot in the context of marketing analytics is that these tools shorten the time to value from data companies already have, and reduce the pressure on the marketer to attain a deep understanding of the data assets they wish to simply extract value from, letting them get back to the job of implementing exceptional execution plans against data.

The upstream data revolution: warehouse-native AI and ML solutions in marketing analytics

Until recently, a big inhibitor in going composable was ensuring there was a sufficiently strong unified data set (in this case, a single-consumer view) to justify building out a system that leveraged a centralized database.

Getting to a single-consumer view often hit a wall upon the realization that the overall interoperability, quality, quantity and relevance of customer data that is collected across an enterprise was in some way deficient. This resulted in years-long missions among analytics and data teams to analyze, resolve, reanalyze, and recombine data — and ultimately deliver, in the best cases, visualized insights to downstream marketers. 

AI and ML in an upstream-composable setting circumvent so many of these challenges. Use ML to spot the pattern in a vast array of diverse data and create a recommendation, and use AI to contextualize the recommendation in the vernacular demanded by the marketer — text, images, offers, etc. Put another way, the ability of these technologies to create new data is an exceptional expression of silo-busting. 

As such, AI and ML deployed in marketing analytics are poised to deliver on face value against a core set of basic, high-impact use cases demanded by any major enterprise. 

Silo-buster use case 1: Customer-centric focus: Building stronger relationships

Drive long-term growth by tapping into customer behaviors, preferences, and expectations across touchpoints, long time horizons, and micro-interactions. ML algorithms excel at identifying patterns within vast datasets, allowing businesses to understand individual customer journeys and tailor marketing strategies accordingly. Put simply, always be there at the right moment with the right message by letting the data anticipate on your behalf. 

Silo-buster use case 2: Optimize customer acquisition and retention

Effective use of ML and AI in marketing analytics enables businesses to identify and target potential customers with precision, ensuring that marketing efforts are directed toward individuals most likely to convert, and to predict churn events with ease, enabling powerful retention interventions.

To date, the algorithmic fuel for these activities rested largely inside walled gardens or single-channel platforms. But warehouse-native AI and ML can scale first-party data in a way previously unimaginable, and can itself be further enriched by the products of its own implementation. This strategic approach to customer lifecycle management further solidifies the feedback loop itself, creating a flywheel effect of testing, learning, and implementing that would be unattainable by human effort alone.

Silo-buster use case 3: Dynamic and relevant marketing outputs — meeting evolving customer expectations

ML and AI bring a level of dynamism to marketing outputs that is unparalleled. This dynamic responsiveness ensures that marketing messages remain relevant and impactful in a rapidly evolving landscape, and dramatically amplifies the quantity and quality of both creative and targeting.

Additionally, predictive analytics powered by ML can forecast trends and customer preferences, allowing programmatic marketing systems to proactively adjust engagement parameters, like ad spend, email frequency, or call center management. This adaptability is particularly crucial in industries where consumer preferences and market dynamics are subject to frequent changes.

In these environments, the generative features of AI amplify the predictive intelligence further, all but eliminating the operational workflows associated with iterating creative and content as demanded by a dynamic dataset, or by real-world conditions.

The simplest truth is that by resolving the execution strategy to the individual level using advanced martech fed by AI/ML inputs, businesses can run a virtually unlimited number of tests at any moment to determine an optimal strategy on a consumer-by-consumer basis to satisfy the preferences of each one and to make the appeal as unique as they are.

AI and ML in the context of digital transformation: 

Any digital transformation strategy works best when it harnesses the power of what experts in the enterprise are already good at doing. For too long, the demands on marketers themselves to become experts in data and technology to satisfy basic strategic imperatives for data use have been far too high. AI and ML can change this paradigm if those capabilities are delivered to them by experts in technology, data, and analytics, who have the chance to seize upon the exciting trends that are proliferating in advanced marketing analytics. 

But, lest the message be delivered too subtly, the clear perception at the highest levels of corporate responsibility is that the paradigm has already shifted — and if you’re not started up, you’re behind. 

According to McKinsey research expectations for “significant” or “very significant” estimated impact from AI use cases in the realm of various aspects of marketing are held by a solid majority of CMOs.

Furthermore, it’s clear leaders are already more active with AI and ML than might be assumed, given the ongoing discussion of data challenge, with more than 70% of BCG surveyed companies stating they are at least using some capability today.

In the minds of deciders, AI and ML are the keys to gaining competitive advantage through acceleration — making the right strategic choice to embed this capability as a centralized data management function within marketing analytics will further amplify the results that can be delivered to the sector leaders across the enterprise landscape.