Beyond Efficiencies: How Asset Managers Can Use AI and Machine Learning to Improve the Impact of Their Marketing Collateral

  • 1709

    Beyond Efficiencies: How Asset Managers Can Use AI and Machine Learning to Improve the Impact of Their Marketing Collateral

    Volumes have been written and discussed about the application of artificial intelligence (AI) technologies, such as machine learning and natural language processing, for sales and marketing functions, particularly in the consumer retail space. Retailers are leveraging AI and big data to identify daily buying patterns, target ads, recommend products, generate customized content and communications, predict needs and questions, personalize experiences, optimize pricing dynamically and engage customers intelligently via chat bots.

    But what about industries that are focused on businesses or high net-worth individuals, rather than the day-to-day consumer? How can AI and machine learning be applied to the world of financial services, and in particular, the asset management sector?

    As AI technologies and advanced concepts such as deep learning continue to evolve, retailers’ ability to improve the effectiveness of sales and marketing will increase exponentially. With mountains of data at their disposal collected from millions of consumers across numerous online and offline platforms and channels, big retailers like Amazon, Netflix, Apple and Wal-Mart are poised to capitalize on AI. Retail marketing organizations understand this continually expanding potential, which is why 84% of them are planning to implement or develop their AI and machine learning initiatives in 2018.1

    Granted, some of the giant, retail-oriented banks and asset managers such as Fidelity, Vanguard and American Funds can stand toe to toe with the major consumer goods retailers when it comes to generating large volumes of data, client interactions and transactions to analyze. In fact, in a recent webinar — AI for Asset Managers: Ignore the Hype, Focus on the Value — Publicis.Sapient’s Rashed Haq and David Poole indicated that larger retail asset managers are benchmarking their AI programs against those of the top Silicon Valley retailers such as Amazon and Google.

    However, many more asset managers—even the largest and most retail-oriented asset managers—do not generate the same kind of data volumes as the consumer goods retailer giants for a variety of reasons. For example, they may just be smaller, perhaps by preference. They may have a lower profile, are more exclusively focused on higher net worth customers, or their client base may be entirely or mostly institutional. In short, they draw less traffic, and thus do not generate massive data volumes and transactions on which to unleash AI to produce powerful insights and take intuitive actions.

    Yet, for asset managers, the actual and potential monetary value of each prospect or client interaction may be much higher than that of a typical consumer retail-oriented organization. A single interaction could lead to decisions involving thousands if not millions or hundreds of millions of dollars, making it critical to draw more from less. But how?


    Analyzing Marketing Collateral Use

    One possible solution is for asset managers to dissect how prospects and clients use their marketing and client service collateral. Asset managers have one key advantage over consumer product retailers: the buying decisions their customers make about investments are typically much bigger and more important than deciding which book, movie or shoes to buy. They are faced with more significant financial decisions with far-reaching impacts, and, as such, the decision process requires much more data and analysis. And, of course, the data must be 100% accurate and current.

    Fortunately, many asset managers have already automated production and digitized the marketing collateral customers require to support this analysis and decision making. Fintech companies, such as Kurtosys, Xinn and Seismic, offer platforms and solutions that automate the production of fund fact sheets, pitch books, client reports and other key customer-facing documents, ensuring they are accurate, current and compliant.

    In implementing these solutions, the initial goal of asset managers may be to reduce the cost and improve the quality and accuracy of marketing artifacts. But by digitizing these assets, they become more versatile, enabling asset managers to track and analyze how a customer engages with them.

    For example, a retail customer or prospect may decide to conduct some investment research and, in the process, may review their client statements, as well as fund/ETF fact sheets from an asset manager’s website. Some potentially useful data may be tracked from this online interaction, such as how long each document was open, scrolling patterns, which portions of which document were viewed the most and longest, drill-down patterns, the order in which documents in which were viewed, and so on.

    This information—in combination with actions taken just before and after reviewing these documents, recent transactions, customer profile data, current investments, asset allocation, and even external data such as current market movements, economic and political news, and world events—could form the basis for some compelling insights about that customer or prospect.

    For instance, this data may reveal that investors in their 40s employed in cyclical industries with large families and aggressive asset allocations will typically make a decision to invest in a fund after an average of 2.4 fact sheet views during a bull market. These insights could lead to much more effective customer engagements, such as cross-selling or upselling the right products and services, providing the next best customer service action, and ensuring better outcomes such as increasing assets or acquiring new customers and retaining existing ones.

    AI Insights

    Delivering Greater Value

    Taking it a step further, more retail-minded asset managers may want to compare behaviors and patterns of their customers’ interactions with those of non-financial retailers by leveraging powerful data analytics platforms. Deeper and more powerful insights may then emerge where retail customer behaviors and buying patterns are similar across multiple industry sectors.

    Perhaps less obvious but more intriguing is the case of institutional asset managers who use pitch books and investment reviews to engage current and potential clients. As with fact sheets, fintech solutions are capable of automating the production of these documents and reducing the overall effort and cost of the process as much as 50% while significantly increasing quality and accuracy. 

    This automated and digitized collateral can be presented and distributed to customers in such a way that it actually collects feedback and data about what the customer finds most useful, relevant and valuable. Over time, this data — in conjunction with other data sources such as investment profiles and goals, industry news and earnings reports — can be collectively mined to reveal patterns and draw insights from specific customer actions and behaviors.

    Next-generation content automation platforms based on core AI principles, from such providers as Xinn, generate hyper-personalized client communications. They have the ability to track basic metrics such as opens, views, time on page, interactions (e.g., likes, questions, bookmarks), content sharing, conversations around content, return reading patterns, and longevity and popularity of content — providing significantly more valuable insights than those produced using passive reading patterns.

    The continuous data gathering, combined with internal data such as customer relationship management (CRM) and external data such as current and historical market information, builds a large library of data mined by machine-learning algorithms to discover patterns and predict future results. These patterns can relate to specific behaviors among asset management customers, such as:

    Predicting the outcome of a set of content for a given audience

    Identifying sequences that end in success, failure or another result

    Discovering high-engagement/low-engagement content across client teams

    Grading the effectiveness of content collections and sequences

    With these insights, asset managers can benefit from smart, hyper-informed suggestions about which content to use for which audience. For example, a relationship manager at an institutional asset management firm who wants to build a pitch book or report for an upcoming meeting will not be limited to just a list of assets previously used or that fit her search criteria. She will also be able to view entirely new content based on analytics focused on the effectiveness of the content.

    More determined asset managers may further improve these insights and recommendations by:

    1. Extending the data set to include information and events about the client’s industry, markets, customer base, supplier, competitors and financial performance
    2. Applying, as appropriate, ever-evolving machine learning and data science methods and technologies
    3. Engaging the support of subject matter experts with a combination of machine learning and in-depth knowledge of the asset management industry

    Asset managers equipped with these capabilities will find themselves on par with some of the most technologically advanced online retailers­, which may be helpful should the likes of Google or Amazon decide to enter the asset management business.


    The Author

    Thomas C. Kracz, Vice President, Practice Lead

    Vice President and Strategy and Solutions Practice Lead for Publicis.Sapient’s artificial intelligence practice and directs the company’s Client Connect asset management solution and services. Based in Boston, Tom has more than 25 years of experience in technology services for the global capital markets, having held executive positions with established software and consulting firms in the sector.

    Leave a Comment