7.3 Smart Marketing and the Marketing Strategy Framework

DataToBiz, LeewayHertz, Markovate, and Systango are just four of the many companies that offer AI product development solutions. Systango alone has completed 1,000 projects for clients spanning more than 10 countries.10 Experts see AI playing a bigger and bigger role in product development as companies learn how to integrate it into their business teams.

Black Swan Data is a London-based technology and data science company that looks into the future of consumer behavior using predictive AI. It pulls data from social media conversations and identifies growth opportunities for companies by reading trend signals from the underlying patterns in consumer data. The cofounder, Steve King, says, “Black Swan is akin to the world’s largest focus group. It continuously analyzes this data to map growth opportunities and identify emerging trend signals earlier, and more accurately, than traditional market research approaches. This capability is bringing a more scientific and comprehensive approach to the new product innovation process, helping brands to de-risk decision making in uncertain times when consumer behavior is rapidly shifting.” Black Swan is doing well and has grown its customer base quickly since the COVID-19 pandemic. One of those companies is PepsiCo, which is quite open about its partnership with Black Swan Data.11

PepsiCo asked Black Swan Data to analyze social media conversations to help determine their next generation of sparkling water product offerings. King describes the work saying, “We ingested 157 million real-time beverage conversations from online data sources and filtered out any irrelevant content, e.g. duplication, spam, advertising, bots, leaving only useful, high-value consumer verbatim. . . .  We identified several thousand trend manifestations that were shaping the category in each area. Our TPV (Trend Prediction Value) metric then ranked each trend based on its maturity and future growth potential.”

PepsiCo used three consumer insights coming from the data to drive product development within the sparkling water category. These insights were, (1) I like the sensation of carbonation, but it needs to be in a “better for you” form, (2) I want the product to be clean, that is, free from preservatives, artificial sweeteners, and any other “nasties,” and (3) I want the product to taste good and be refreshing, but it must also deliver functional benefits beyond just hydration. PepsiCo then introduced a new range of natural-flavored sparkling water products named Bubly. Bubly has no calories, no artificial sweetener, no added sugars, no artificial flavors. After the product launch, Todd Kaplan, VP of the water portfolio at PepsiCo North America Beverages, commented to AdAge, “When we looked at the sparkling water category, we saw an opportunity to innovate from within by building a new brand and product from the ground up to meet consumer needs. . . . We created Bubly to provide consumers with a great-tasting, flavorful, unsweetened sparkling water in a fun, playful, and relevant manner that is unlike anything we’ve seen in the sparkling water category today.”12 Within 12 months of product launch, sales of Bubly exceeded $100 million.13

AI’s contribution to brand management is still developing. What’s the secret for improving brand management with AI? Marc Pritchard, Procter & Gamble’s chief brand officer, says “The best way, we found, to do it: Don’t talk about the algorithms, don’t talk about the technology, don’t talk about AI. . . . Talk about the outcome you want. What are we trying to achieve?” P&G wants to reach 100% of their target audience without redundancy, develop creative content that is individualized for each market and each need set, and improve how they spend media dollars.14

Tide is the biggest brand in P&G’s fabric and household care division. AI algorithms are now driving its planning and scheduling process, which places ads across dozens of networks and thousands of program combinations while cutting down on redundancy, something particularly difficult to do when advertising on broadcast TV. In Pritchard’s words, “We want to reach, mass reach—it’s close to 100%, with precision. We don’t want excess frequency. We want to get to as close to 100% as we possibly can, and we’re doing that on some brands. We’re close to more than 90% reach on Tide. We’re close to 95% reach on Pampers.” 15

Amazon is known for its AI-driven Dynamic Pricing Strategy. Amazon AI algorithms analyze competitor prices, changes in product demand, amount of stock, and customer browsing history to adjust their prices in real-time. Amazon AI also predicts changes in demand for its millions of products and, based on those predictions, changes its pricing strategy on-the-fly.16 Amazon uses AI and machine learning extensively to optimize its pricing strategies.

Here are some key ways AI is employed in Amazon’s pricing:

• Amazon AI tailors pricing premiums and discounts for individual customers as well as customer segments based on their personal browsing and purchasing history. Amazon also uses automated price matching, inventory overstock, shipping and warehousing costs, customer conversion rates, and customer post-purchase perceptions to ensure their products earn an optimal profit while still keeping their prices competitive.17

• Amazon’s AI systems will help the retailer challenge new competition from low-price rivals Temu and Shein. Now, through a discount section, Amazon will offer competitively priced products to US shoppers willing to wait 9 to 11 days for delivery, rather than the usual one or two days. This new part of Amazon’s business will target price-sensitive, time-insensitive shoppers with goods airshiped directly from warehouses in China. In its official statement, Amazon said, “We are always exploring new ways to work with our selling partners to delight our customers with more selection, lower prices and greater convenience.”18

Distribution is key element in the marketing mix because the product needs to be in the right place and at the right time before a consumer can make a purchase. AI improves product distribution by automating many of the repetitive and tedious tasks associated with supply chain management, providing demand forecasts based on more and better data than in the past, and improving inventory management.

UPS is an innovator in using AI to improve distribution. In their business, even small changes in delivery routes can either save or cost tens of millions of dollars. UPS chatbot, UPS My Choice, and ORION are three AI platforms that have had a significant impact on their business. UPS chatbot uses NLP and NLG to mimic human conversation when answering customer service questions about shipping locations, package tracking, shipping rates, and other frequently asked questions. The more conversations the chatbot has, the better it will answer future questions.

UPS My Choice is another AI tool. It allows millions of residential customers to choose how, where, and when to get their home deliveries. ORION is an acronym for On-road Integrated Optimization and Navigation. Using data provided by UPS drivers, delivery trucks, and customers, ORION suggests optimal delivery routes as each driver works through their deliveries. ORION adjusts to route changes due to weather and traffic conditions in real-time.19

AI is particularly useful for targeting specific customer groups and personalizing the messaging and buying incentives. Sales promotion is another area in which AI can play a successful role. Responding to the demand for greater personalization from young consumers, Starbucks has developed an AI platform called Deep Brew. Deep Brew, in addition to managing employee scheduling and product inventory, also makes menu recommendations, personalizes the customer journey through the Starbucks mobile app, and streamlines the Starbucks Rewards program to improve customer engagement.20

Following Starbucks’ chart-topping first-quarter fiscal 2024 earnings report, CEO Laxman Narasimhan explained how AI is generating greater value for Starbucks Rewards members. Narasimhan said, “In the U.S., we implemented targeted offers aimed at bringing our occasional customers into our loyalty program. . . . As we’ve seen over time, Starbucks Rewards members develop a routinized long-term relationship with our brand that increases both tickets and transactions. Additionally, we activated new capabilities within our propriety Deep Brew data analytics and AI tool to identify and incentivize specific rewards members cohorts.” Because of Deep Brew, said Narasimhan, “we saw our mobile order and pay surpass a record-high 30% of all transactions of the quarter, and we reduced downtime of mobile order and pay by half.”21

Procter & Gamble is also using AI to improve the images, music, copy or text (i.e., the creative execution) of their advertising. As an example, Pritchard says, “Pampers has 140 different pieces of creative optimized through this program called AI Studios, which allows us to be able to test ads versus a big database of reactions and figure out how we can make it better.” AI Studio is a “neural data network” exclusive to P&G. The network is built on decades of advertising research with consumers. Consumer reactions to new creative executions are uploaded to AI Studio and measured against previous ads to predict future performance. 22

A retail media network is a “retailer-owned advertising service that allows marketers to purchase advertising space across all digital assets owned by a retail business, using the retailer’s first-party data to connect with shoppers.” Amazon advertising is an example of a retail media network. Do a search on Amazon for Tide Pods and you will see sponsored ads for Tide the brand as well as individual Tide products. You will also see display ads for Tide Pods and Tide Pod competitors. Marketers purchase these advertising placements from Amazon based on Amazon’s own customer data, also called first-party data. Many retailers offer retail media networks to advertisers. Three more examples of extensive retail media networks are Walmart Connect, Kroger Precision Marketing, and Target’s Roundel. Retail media networks are one of the largest and fastest-growing spend categories of a CPG company’s marketing budget, according to Pritchard. And P&G is using AI to make their media dollars spent on retail media networks more effective. In cooperation with retailers, the company is creating what Pritchard calls an “auto-bidder.” The auto-bidder taps into a retailer’s first-party search results which, in turn, makes it possible for P&G to adjust and optimize its ad buys and content every 15 minutes automatically. Using auto-bidder, P&G has captured four times higher returns on brand sales.23

Using historical sales data to predict future sales is nothing new. But it has been said that “Telling the future by looking at the past assumes that conditions remain constant. This is like driving a car by looking in the rear view mirror.” (Herb Brody, MIT Technology Review, May 19, 2010) Predictive AI, however, integrates multiple sources and types of data and will update sales forecasts as new data become available. AI changes the sales forecasting paradigm. For example, when online shopping exploded during the COVID-19 pandemic, retailers using AI could quickly update their sales forecast to reflect this huge shift in consumer behavior.24

Predictive AI can do more than forecast sales, it can shape future sales by improving upon traditional methods for finding sales-qualified leads. Quoting Harvard Business Review, “A salesperson with a rich pipeline of qualified potential clients has to make decisions on a daily, or even hourly, basis as to where to focus their time when it comes to closing deals to hit their monthly or quarterly quota. Often, this decision-making process is based on gut instinct and incomplete information.”25 The main problem with the traditional approach to qualifying sales leads is that the salesperson may assign scores in any way they see fit without clear guidelines. Using predictive AI to qualify sales leads helps businesses to identify the best prospects faster. AI helps salespeople reach the right prospects at the right time with the right messages. All this adds up to more qualified leads and a higher success rate in closing deals. Integrating data from hundreds of sources, including social networks, predictive AI assigns a number to each potential customer to determine which leads qualify for moving forward in the sales process. Also, as in other AI applications, the process adjusts its predictions as business conditions change and new customers are acquired.

WeightWatchers for Business is the B2B side of the WeightWatchers organization. It contracts with small, medium, and enterprise businesses across North America to provide weight-related healthcare offerings, including behavioral health, medication management and virtual care. Its programs are designed to change behavior associated with diabetes, high blood sugar, and obesity. Looking to boost enrollments, WeightWatchers partnered with HubSpot to better focus on the most qualified sales leads and deliver a seamless customer experience. WeightWatchers Health Solutions Director for Commercial Growth, Traci Shoemaker, wanted a solution to help her sales team be more efficient and effective. She says, “I wanted to make sure that my team was getting leads when they were warm and could respond quickly when somebody was thinking about our product.” Describing the current state of lead generation and qualification, she continues, “The process was very manual. We had to copy-paste message templates from a Word document. To reach a large batch of people, we had to either recreate an email for every single prospect, or use blind carbon copy. . . . We weren’t able to track leads through robust deal stages,” Traci says. “And there was no customization in how we filtered or viewed our data. . . . HubSpot shone for us by visualizing our pipeline. It allowed us to identify leads with a viable path to revenue, share that visibility with everyone in our organization, and connect with them seamlessly.”26

With HubSpot’s predictive AI, Traci’s sales reps were able to rethink the way they qualify leads. Rather than wrangling an enormous pool of low-quality leads, they only generate a WeightWatchers “deal” for leads who have a clear path to revenue. “Before, our pipeline seemed inflated with unlikely deals,” Traci says. “Now, we have a sense of the true opportunity for revenue.” Some of the adjustments suggested by HubSpot’s AI have been surprising. WeightWatchers B2B has shifted its enterprise target upmarket, and focused on corporations that want to offer their program as a fully-covered benefit rather than a temporary perk. AI is particularly good at performing repetitive, mundane tasks. This is also true for HubSpot’s AI. Traci says, “You don’t want your sales folks spending their time doing administrative work. With HubSpot, our contact details feed automatically into the company record, and our nurture messages are created right inside the system using templates stored inside the sequences. . . . The biggest thing to me is that the salespeople want to use the system. They see the benefit it’s having within their book of business”.

AI has paid off for WeightWatchers. Traci says, “Our close-to-win ratio is near and dear to my heart, because we’ve worked hard to tighten up that process. On our large and jumbo accounts, it went from around 1 in 50 to 1 in 7. HubSpot made that easy for us to achieve. . . . Being able to shorten the sales cycle has allowed us to prospect the type of leads that have the strongest close ratio. There’s a better match now between what we offer and what they’re looking for.”27

A key to customer retention is personalizing brand experiences. L’Oréal’s Skin Genius uses AI to analyze skin types and then generates a science-based, personalized skincare program. Discussing the benefits of Skin Genius, Elisabeth Bouhadana, global scientific director at L’Oréal Paris International, says, “It quickly answers the question every single woman asks herself when selecting a product for themselves: based on the visible aging signs on my face, which product would really be best for me? . . . Women trust professional advice, and Skin Genius is like having a professional advisor on demand.” Skin Genius is backed by 50 years of research augmented by the analysis of 10,000 clinical images of women and 18,000 smartphone selfies taken in all manner of different lights. Bouhadana explains, “It’s highly reliable and matches 95 per cent of dermatologists’ evaluation on all skin types. [It’s] easy to use; pedagogical, since it explains what skin problems need treating first; and works in a few seconds.” Unlike other skin analyzers, Skin Genius recommends specific ingredients that will be the most effective at treating individual skin needs.28

“Now, our marketing is augmented through generative AI. We started our digital transformation journey more than a decade ago, and back in 2018 we introduced beauty tech to harness a new kind of relationship with our consumers, one that is based on data, technology, and AI. Our goal here is ultra-personalization; to bring beauty for each—powered by beauty tech,” says Asmita Dubey, L’Oréal’s Chief Digital and Marketing Officer. “We are democratizing beauty services using gen AI. We have just launched L’Oréal Paris Beauty Genius, a generative AI-powered personal beauty advisor that is available 24/7 in our customers’ pockets.” In 2023, innovations such as virtual try-ons for hair and makeup, as well as the Skin Genius app have attracted more than 40 million users.29

In our opinion, better marketing data usually leads to better marketing decisions. AI is making it easier than ever before to sift through millions of data points and extract meaningful insights. By automating the data analysis process, AI has the potential to give every marketer a window into the behaviors, preferences, and needs of their customers. AI is always gathering and churning through new data. Consequently, AI can be a powerful tool for increasing sales per customer.

Customer journey mapping is an area where AI can make a big impact on sales per customer. A customer journey is the series of steps that a customer takes in their journey from discovering a brand to purchasing and advocating a brand. A simple customer journey map is usually built around five stages and mirrors the consumer decision-making process first proposed by Cox et al.30

Each step in the customer journey provides an opportunity for marketers to influence customer attitudes and behavior. Because the customer journey is often online, AI can inform what marketers should do at each stage of the journey for individual customers. AI can (1) improve awareness by optimizing email subject lines and content, (2) guide consideration by shaping organic and sponsored search results, (3) influence decision by managing social media reviews, blogs, and online articles as well as setting optimal prices and/or promotional incentives, (4) improve retention by monitoring personal experiences with the product, and (5) increase advocacy by identifying optimal messaging and incentives to encourage review writing and positive social media posts.