icon-arrow-back-perspectiveicon-arrow-forward-perspectiveicon-arrow-nexticon-arrow-next-nofillicon-arrow-previousicon-arrow-previousicon-closegoogle-plusinstagramlinkedinnavigate-nextnavigate-previousicon-social-closeicon-social-expandtwitter Skip to main content
    Brain with a superimposed circuit board and the copy, "AI and the Meeker Report"

    Hans Kaspersetz

    President

    July 20, 2018

    The Meeker Report Sees Opportunity in Artificial Intelligence. We See Results.

    When Mary Meeker discusses her annual Internet Trends Report, Arteric listens. 

    This year, we were especially interested to learn what the 2018 edition had to say about the adoption of artificial intelligence (AI) by marketers. For 2 years, we’ve been using AI analysis of customer search data to enhance our customers’ content strategy, website development, organic search optimization, and paid search (PPC) campaigns. The artificial intelligence application that Arteric developed uses natural language processing to detect signals of customer behavior that escape human data analysis. Our analysts, writers, and strategists leverage these insights to drive traffic growth and engagement in professional Rx and consumer OTC campaigns. We also use them to improve the efficiency of paid search campaigns.

    Ironically, as we’ve integrated AI technology into many of the services and assets that we deliver, survey data suggest that many marketers remain skeptical about the value of AI. According to a November 2017 Resulticks study of more than 300 marketers across multiple industries1

    • 47% consider AI to be overhyped
    • 40% use “skeptical” to define the emotions they felt when they saw or read about AI 
    • 42% have no plans to employ AI in their marketing strategy  
     

    Our experience is that marketers are skeptical until we demonstrate a narrow use case that they can conceptualize and visualize in their mix of marketing tools. We started educating on and socializing the use of AI with our clients well in advance of delivering the AI-influenced insights and strategies.

    As we dug into the Meeker report, we wondered:

    • Does the Meeker Report see AI as help or hype for organizations?
    • Is there evidence that AI is being adopted as a tool for marketers?
    • Are organizations aggregating and managing the data that are required for effectively applying AI and machine learning?

    We sought these and other insights that carry ramifications for the value of AI to healthcare marketers. Here’s what we observed. We’ve included the slide number where we thought it would be helpful. It’s a big report.

    It’s Early, but Organizations Are Accelerating Investment

    Meeker reports that across enterprises, the spend priority for AI is relatively small but is increasing rapidly [See Slide 201]. In January 2018, approximately 2.5% of chief information officers in the United States and the European Union reported that their largest spend increase will be for artificial AI solutions. In April 2018, that rate doubled. 

    Bar chart that describes corporate spending on artificial intelligence

    Image reproduced with permission from Mary Meeker, Kleiner Perkins.

    Marketers Are in the Game

    Although AI investment is modest now, this masks the active role that AI plays for marketers. In 2018, Blueshift, an AI solutions provider, surveyed 200 marketers from 198 B2C companies across a wide range of industries.2 Forty-three percent of respondents use AI technology to expand their audiences. Large percentages of marketers reported that they were using AI for audience targeting, product recommendations, and campaign optimization.  

    Bar chart that describes the applications for which marketers use artificial intelligence

    Image reproduced with permission from Blueshift.com from Activating Customer Data for AI Powered Marketing.

    And interest by marketers is growing. Sixty-four percent of respondents plan to increase their use of AI to support their campaigns. 

    It is worth noting something that Meeker and Blueshift did not address. All marketers use AI every day when they use Google and Bing or a voice assistant to search for information on their competitors or their own brand. These artificial intelligence generated search results affect the marketing decisions and strategies that marketers choose, even if marketers are not cognizant of that influence. Marketers are also using AI when they review data from Google Analytics, Google AdWords, and programmatic advertising platforms. AI is pervasive and in daily use, even if the brand’s direct investment in AI is low.

    Healthcare marketers have been leveraging AI since at least 2013, and are succeeding

    In 2013, GSK used AI to identify parental concerns about vaccines and the vaccination process. GSK’s analysis of thousands of conversations from online discussion boards uncovered multiple issues that influenced the decision of whether or not to vaccinate. GSK strategists then created content that addressed these issues.

    In 2016, Arteric began to use AI to analyze PPC data from an Rx brand that competes with billion-dollar blockbusters. The data that ensued helped our analysts increase organic search traffic by 145% in 2018 (year over year). In 2017, we used AI-generated data to discover a new target demographic for an OTC brand.

    Timeline arrow showing the years 2013, 2016, 2017 and 2019. Under 2013 is an icon for a GSK AI project. Under 2016 is the icon for an Arteric Rx project. Under 2017 is an icon for an Arteric OTC project. Under 2019 is the word "Next."

    The Essential Role of Data

    Large sets of data such as Wikipedia and journal articles are required to train an AI solution to perform its analysis effectively. Paired with the training data, large sets of structured and unstructured brand data — paid search queries, organic search queries, Google Search Console data, customer service transcripts and other voluminous sources — are required to analyze and effectively draw insights and conclusions from digital marketing campaigns.

    The Meeker report recognizes data’s critical importance. More than 10% of the report (31 of 294 slides) is dedicated to the gathering, sharing, and optimization of data; its importance to AI; and data privacy issues. 

    Data management correlates with satisfied customers

    Meeker points out that Amazon and Google, which aggressively aggregate and analyze data, have high scores in surveys of customer satisfaction. Both of those companies, along with Facebook, Netflix, and Booking.com, rely heavily on AI.3,4,5,6,7

    Arteric’s original goal in implementing AI for our customers was to enhance brands’ connections with customers by analyzing how customers talk about brands and by using insights about lexical variance to build more engaging content. A by-product of this effort was that we discovered many gaps in our customers’ content and content strategies; these gaps were evident in the search patterns we found in the data. We were then able to work with our customers to fill those gaps with the content that customers were seeking. Filling those gaps led to greater engagement and customer satisfaction. In one case, AI-derived content enhancements led to a 145% year-over-year increase in traffic to an HCP brand site.

    Bar chart that correlates customer satisfaction with high use of Internet data

    Image reproduced with permission from Mary Meeker, Kleiner Perkins.

    Data use correlates with revenue

    The Blueshift survey reports that 70% of organizations exceeded revenue goals when they used 75% or more of their data. Only 50% of organizations exceeded goals when they used 75% or less of their data.

    Bar chart that correlates achieving revenue goals with using more than 75% of brand data

    Image reproduced with permission from Blueshift.com from Activating Customer Data for AI Powered Marketing.

    Own Your Data or Go Home 

    The linkage between brand data and success with AI is unequivocal. For years we’ve strongly encouraged clients to aggregate and DIRECTLY manage their data. Brand teams, not their vendors, must control and have immediate access to their data. 

    Research supports the value of aggressive and direct data management by brand teams. Marketers who access brand data without the involvement of their information technology (IT) teams are 1.6 times more likely to transform their data into practical marketing activities. 

    Bar chart that compares the percentage of companies where marketers control the data versus where the IT group controls the data.

    Image reproduced with permission from Blueshift.com from Activating Customer Data for AI Powered Marketing.

    Our experience also points to the essential role of data management. Because the marketing team provided us with a rich collection of PPC data, Arteric’s AI solution was able to generate insights that enabled identification of a previously unknown target audience for a well-known OTC cold, cough, and allergy brand. 

    Here’s how the discovery unfolded. Arteric’s AI solution analyzes 250,000 rows of PPC data in 2.5 minutes. Without AI, 400 person-hours would have been required. This efficiency allowed us to repeatedly interrogate the data and uncover foreign-language inquiries in an English-only campaign. We integrated language translation and processing algorithms into the AI solution and re-analyzed the data. The language classifier and translation algorithm identified 147 foreign-language inquiries, nearly all relevant to the brand. The client incorporated these insights into marketing strategies that are currently under study. 

    The Arteric Viewpoint 

    Properly applied to an appropriate business problem, AI accelerates data analysis and the discovery of insights from digital marketing campaigns. However, people are still the magic. Arteric’s success in applying AI to Rx and OTC marketing challenges and Glaxo’s use of AI to enhance the content strategy of their vaccination program align with the findings reported by Meeker. All of the data confirm our belief that AI is well within the reach of most pharmaceutical and biotechnology brand teams.

    There are several criteria to evaluate when determining the value and applicability of AI for brand campaigns. Call me at 201.558.7929 to schedule a complementary 1-hour audit of your business challenge to assess the impact of AI on your marketing program.  

     

    *The Meeker report made no mention as to whether the Cambridge Analytica scandal contributed to Facebook’s below-average rank. 

    References

    1. Half of marketing pros consider artificial intelligence an overhyped buzzword. Cision website. http://www.prweb.com/releases/2017/10/prweb14842770.htm. Published October 26, 2017. Accessed July 13, 2017.
    2. Activating customer data for AI powered marketing: from aspirations to reality. Blueshift website. https://downloads.blueshift.com/report-customer-data-for-ai. Accessed July 13, 2018.
    3. Machine learning on AWS. Amazon.com web site. https://aws.amazon.com/machine-learning. July 13, 2018.
    4. Advancing AI for everyone. Google AI web site. https://ai.google/. Accessed July 13, 2018.
    5. Tools for advancing the world's AI. Facebook Artificial Intelligence website. https://facebook.ai/developers. Accessed July 13, 2018.
    6. Bandyopadhyay A. How Netflix deploys open source AI to reveal your favorites. https://itsfoss.com/netflix-open-source-ai/. Updated April 11, 2018. Accessed July 13 2018.
    7. Booking.com about data science, machine learning and artificial intelligence. Quint website. https://www.quintgroup.com/en-us/insights/booking-com-about-data-science-machine-learning-and-artificial-intelligence. Accessed July 13, 2018.
    8. Meyer R. The Cambridge Analytica scandal, in 3 paragraphs. The Atlantic website. https://www.theatlantic.com/technology/archive/2018/03/the-cambridge-analytica-scandal-in-three-paragraphs/556046. Accessed July 13, 2018.