Waste Not, Want Not: How AI is Helping Clean Up and Reduce Marketing Waste

Analyzing consumer preferences and behavior to understand their wants and needs – and then being able to actually meet those needs exactly where and when they need them – is considered ‘the promised land’ for marketers and customer engagement practitioners.  This kind of marketing effectiveness promotes customer loyalty, reduces churn, and attracts new customers.  In the not-so-distant past, many of us would have said that this concept of true omnichannel customer engagement was within reach.  But the rush to close the gaps left by data deprecation has laid bare just how far we are from that goal.

Instead, our MarTech stacks are bloated and underutilized, meaningful measurement remains elusive, and paid media and digital marketing spend is a black box, collectively adding up to wasted marketing dollars.  The Association of National Advertisers released an explosive report this past December highlighting just how much money brands are hemorrhaging in the programmatic advertising ecosystem: 64 cents on the dollar.  That’s right – 36 cents of every advertising dollar spent in the programmatic advertising effectively reaches the consumer., suggesting brands are wasting approximately 64 percent of their paid media budget. And that’s just paid media.

Additionally, a recent Gartner statistic suggests that CMOs report using just 42% of the total capabilities available in their MarTech stacks, down from 58% in 2020, as well as allocating a quarter of their entire marketing expense budgets to marketing technologies in 2022. These numbers don’t include duplicative impressions delivered on owned channels.

These inefficiencies are draining marketing budgets with no return, but we can reverse these effects to dramatically reduce waste thanks to new technologies like AI, as well as a proper implementation strategy.

Optimize Technology and Create Value with a Centralized Approach

Optimizing marketing programs isn’t that big of a stretch. We know this because we’re seeing brands who have leaned in hard to marketing and customer engagement technologies with centralized AI reaping the benefits. These companies are ensuring their AI is analyzing customer data across all channels to get a more holistic view of their customers with significantly less room for error.

Centralized AI can unify inbound and outbound channels, effectively breaking down data silos that disrupt customer experiences and waste dollars.  For example, a communications service provider with data silos in their stack risks creating duplicate customer profiles for the same user, potentially resulting   in a customer receiving differently priced offers for the same streaming service in different channels. At the very best, interactions like this prompts the customer to call into the contact center to understand why this is happening – a much more costly interaction than an online cross-sell.  At the worst, the consumer distrusts the brand, feels taken advantage of, and takes their business to a competitor.

Because centralized AI can activate data in real time across connected channels, engagements happen with consistency and only in the channel that consumer is in.  The customer journey isn’t driven by the product a brand wants to sell, it’s driven by the customer’s needs and preferences.

Marketing Technology News: MarTech Interview with Matt Hertig, CEO at ChannelMix

Understand the Difference Between Personalization and Hyper-Personalization

While marketers currently have the capability to offer some level of personalization (in part with third-party cookies), it typically materializes as more of a personalized sales pitch.  An example of this is a customer logging into their mobile app to pay their cell phone bill and immediately being confronted with the opportunity to upgrade to the newest phone model. The brand knows the customer doesn’t have it, so they try to sell it to them.  But what if that customer is already struggling to pay the bill? What if historical data on that customer indicates that their payments are late each month?  A data silo would prevent marketing from understanding that this is not a solid prospect for the offer. This also indicate to the customer that a brand doesn’t have their best interest in mind.

In this particular case, a truly hyper-personalized interaction would let the customer know that there is a discount available they haven’t yet claimed, as well as payment plan options with no interest available, or that they are paying for services they haven’t been using and have the option to downgrade their plan. Once the customer is back on their feet and can pay their bills, the brand can share that option to upgrade if appropriate. This helps the customer understand that they’re not just a number – the brand also cares about their financial wellbeing.

A hyper-personalized interaction (or interactions) like this extends the lifetime value of a customer by keeping them current, saving the brand money on collections, as well as setting the stage for a larger purchase in the future – net new revenue.  This also injects empathy into the customer’s experience. While the short-term sale isn’t there, neither is the possibility for default and, with service like that, less of a churn risk – both of which are extremely costly to the brand. When you have AI on your side constantly analyzing and evaluating customer information, the technology itself can communicate to marketers where and when is the right time to reach out to a customer with certain information. Bolstering your marketing with a centralized AI strategy can practically guarantee customer interactions will be more hyper-personalized than trying to toggle between siloed data to find an answer.

Measure Marketing Outcomes that Connect to Revenue

Marketers have historically struggled to connect their program spend to actual revenue because there is a wide variety of metrics used to assess business outcomes – many don’t offer a straight line between a customer seeing a marketing engagement and making a purchase.  At times, reporting can consist of a patchwork of contextual measurements like click-through rate, acquisition cost, email open rate, bounce rates, and cost-per-lead to directionally understand some level of impact. There is also a significant transparency gap on reporting for owned versus paid channels. With owned channels like web, mobile, email, and chat, the brand has access to detailed analytics and user interaction data (first-party data), which offers immediate insights into performance and engagement. In contrast, paid channels, like social, print, television, and audio advertising, are often controlled by agency or technology partners, and results also are not provided until a campaign has ended. A marketer won’t know if their message is resonating until after the campaign has concluded and the money has been spent.

On the contrary, adaptive AI is constantly learning from customer signals and cues as they travel through a brand’s channels and can adapt. This enables brands to understand if there is a mismatch between the message and the customer and immediately pivot with a more relevant engagement – such as stopping a customer from getting a specific ad via email if they’ve already declined the offer on their mobile app.

A 2023 study by Publicis and Yahoo reported that repetitive ads damage brand sentiment and that 67% of surveyed views were annoyed by seeing the same ad more than once in a short period of time. What’s worse is seeing the ad even after a customer has purchased an item the is still advertising to them. Data silos and fragmentation prevent the brand from understanding the connection between their ad and purchase behavior.  But with AI-enabled engagement, a brand can understand quickly if a customer purchased something based on an offer they were served, close the loop, and not serve that same customer with that same offer again.

Protect Important Assets

Amidst all this, we must protect one of our most important assets as marketers: brand equity.  AI technology is only as good as the data powering it – if bias (intentional or not) exists within that data, AI-powered decisions will also be biased.  While this can result in something relatively harmless like an irrelevant ad, it can also result in offensive offers that can seriously damage a brand’s reputation.

Running afoul of regulations and social norms can come at the cost of a financial impact as well as consumer trust. To avoid this, brands must be transparent about how they collect and use consumer data, and make sure the data they’re training their AI with is fresh and bias-free. AI algorithms must be tested with and exposed to real-world scenarios and its decisions must be explainable. It’s also imperative to have human oversight auditing outcomes with the ability to approve or override certain decisions.

The use of AI in marketing can deliver a massive value to brands and consumers alike. It will enable productivity and efficiency, help reduce waste and increase sustainability, and help deliver better customer experiences. We must approach AI like any other technology and treat it as augmentative to human creativity, integrate empathy into processes, and understand that while we’re trying to drive revenue, we’re ultimately trying to build and nurture human connections.

Marketing Technology News: Find your muse: How an inspiring figurehead can transform your brand

Also catch: Episode 179 Of The SalesStar Podcast: The Impact of Al in Sales and Marketing with Ketan Karkhanis, EVP & GM, Sales Cloud, Salesforce