Marketers are beginning to amplify their efforts around building optimized customer lifetime value (CLV). What was humanly impossible for marketers to achieve for the most part of the decade is now within their grasp, thanks to the proliferation of Artificial Intelligence (AI). We chatted with Andrew Toner, Chief Technology Officer at Amplero, to know how marketers could derive better ROI from their data and analytics if they are AI-enabled and not AI-driven.
MTS: Why do marketers feel that they are dealing with a black box while working with AI-driven platforms?
Andrew Toner: The perception of the artificial intelligence “black box” within the marketing technology stack is that you deploy it, add creative inputs and KPIs, and—magic—the machine does the rest.
While this is actually a simplified version of how an AIM platform should work, it’s absolutely critical that these tools operate with a high-level of transparency—both in terms of campaign results and delivering machine-intuited insights in a way the marketer understands, but also in terms of what are the underlying machine learning algorithms driving decisions for the machine.
In our latest commissioned Forrester research report that came out this week, researchers found that 78% of respondents believe human involvement is critical for providing guidance alongside AI-enhanced decisions and insights.
The industry suspicion of black box technology stems from two main areas, and rightly so.
First, programmatic buying platforms operating as “black boxes” have created an advertising environment where companies have little visibility into how their vast digital budgets are being automatically allocated, resulting in overcharging for placements or mischaracterizing results.
Second, the marketing technology vendor landscape has often relied on human capital to continuously deliver new technology features. In this case, it’s common for AI capabilities to be born on the backs of large data science teams building custom models for each customer. Look at the acquisition histories of the large marketing clouds. You see a massive scale-up of data science teams.
MTS: Does AI pose a disruptive challenge to companies that are still evolving digitally?
Andrew: Absolutely. The companies investing in artificial intelligence technologies aligned with their customer experience goals now are poised to compete with Google, Amazon, Netflix, or industry outsider X.
The ones that aren’t exploring artificial intelligence innovation to automate action and insight from their vast customer data stores will be playing catch-up to whichever competitor figured it out first. Many companies consider “digital transformation” as just filling a data lake with all the behaviors they can find, while AI is the next necessary step to make the most of that digital transformation investment.
MTS: Do you think that marketers would gain better ROI from their data and analytics if they are AI-enabled and not AI-driven?
Andrew: Both AI-enabled and AI-driven data and analytics have a place within the global analytics strategy for any major organization.
AI-driven processes effectively close the loop for testing and optimization. Instead of a static A/B or multi-variate testing process where the marketer chooses a hypothesis, runs the test, and then uses the results to inform the next test—artificial intelligence marketing technology is continuously applying behavioral insights to identify high value micro-segments and optimize each customer experience. Often, this results in the ability to run thousands of experiments simultaneously.
Meanwhile, AI-enabled data and analytics allows analysts to identify and curate valuable, deep-dive insights that were either intuited by the machine or enabled because their burden of labor shifted from manually creating reports or running tests to true analysis and data storytelling.
MTS: What is the best roadmap to build an AI-enabled Marketing adoption plan?
Andrew: As with any enterprise martech investment, start with a clear set of objectives and requirements. Which systems do you need to integrate with? What are the available data sources? What does success look like within your organization? Which KPIs are you optimizing toward? For example, launching an AIM platform can truly expose issues in your KPI strategy. If you’re optimizing to the wrong KPI or cannibalizing other channels, the platform will surface these challenges.
As you’re evaluating this emerging technology, start with a commercially available AI engine that integrates with existing technology stacks and offers the capability to ingest and enrich data from a broad range of customer, behavior, and content sources, at scale and real-time speed. Self-learning models should be continuously testing and optimizing, empowering marketers to take action based on real customer insights and cut down on resources needed for analysis and reporting.
One crucial item for marketers to know is that in-house data science teams will have a strong point-of-view on any AI or machine learning technology being evaluated. It’s important to involve them early in the conversation because they will either be your favorite champions or worst roadblocks.
MTS: A large percentage of respondents said that they “don’t believe AI in marketing is currently real”. What are the chances that these businesses would manage to sustain performance beyond 2020?
Andrew: I believe this is more a commentary on the state of the marketing technology industry as opposed to these individual marketing leaders’ AI adoption initiatives. They are responding to an overabundance of hype for any emerging buzzword amongst marketing technology vendors. They’re constantly being oversold new capabilities around big data and customer experience that fail to deliver for their business—typically related to hidden resourcing challenges with enterprise implementation and rules-based execution or because the technology itself is flawed. Despite core algorithms existing since the 1950s, artificial intelligence is the latest tired buzzword being trotted out at every conference and press release.
That being said, there are several concrete and highly valuable use cases for AI within the modern marketing technology stack. On the customer interface side, chatbots and AI-assisted customer service reps are obvious channel-specific solutions where companies are finding value. For enterprises using a core AIM platform like Amplero at the center of their technology stack to deliver multi-channel, 1:1 customer experiences at scale, we’re seeing a strong lift in overall customer experience and relevant KPIs—typically related to revenue and retention.
In the Forrester research report, the majority of respondents (78%) are planning to adopt or expand AI platforms within the next year. It’s rare that we meet an executive with their head in the sand around artificial intelligence, and in that case, they are facing severe business challenges currently and in the future.
MTS: Thanks for chatting with us, Andrew.
Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at email@example.com