Category Archives: Data and Analytics

Integrating Human and Machine Advice: Current State and Future Requirements

Several recent articles and pieces of news pertinent to robos and advisors create an interesting mosaic of the current state of human and machine advice:

The image created by these items depicts the struggles in the advisory business to settle on a clear, promising strategy for integrating advice channels.

The Limits of Disruption

When robos appeared on the scene several years ago, they were heralded as the future of wealth management, a democratizing blow for the industry, and a mortal assault on traditional financial advice. Any who have seen the hype machine movie before won’t be surprised that none of those things turned out to be true. In the real world, the biggest “robos” in terms of assets are those of Vanguard and Schwab that operate as hybrids while the “pure play” B2C robos have struggled to accumulate assets and breakeven on customer acquisition costs.

The reason for this discrepancy between reality and hype is simple: Irrational as it may sometimes be, most people want humans involved in their financial planning. A 2016 survey conducted by EMI and Boston Research Technologies showed not only that most want human involvement, but also that those who were more open to algorithm-driven investing didn’t neatly map to pre-conceived demographic categories. The bottom line is that you can’t will customers and prospects into following your vision for a service offering. Moreover, making assumptions about their behavior based on intuition and truism doesn’t create a strong foundation for success.

Changing Perspectives

The truth is that the majority of customers want a hybrid model. Many of the leading wealth managers understand this and have implemented or will implement various forms of hybrid offerings. In fact, as I mentioned earlier, the largest robos are actually those launched by existing wealth managers Vanguard and Schwab.

But any business heading down the hybrid path needs to recognize that their old models of and assumptions about client management and messaging will likely need to change. Specifically:

  • If portfolio management is outsourced to machines, it becomes a commodity and value must be defined in terms of relationships and communication—an idea that has been around for some time but which has not gained universal acceptance because it is hard to execute.
  • If you are advocating for clients to use your automated platform, you need to recognize that you are now responsible for their adoption of and satisfaction with the investment management software. Firms and their advisors need to be ready to assist clients onboard, answer their questions, and help them realize the full value of the software.
  • Pushing the wrong clients towards a robo solution is a lose-lose situation that will cost time and assets. Firms and their advisors need to have ways of identifying where clients are likely to fall on the spectrum of interest in and comfort with automated portfolio management, recognizing that age and net worth will likely not be great proxies.

Making Use of Data: 5 Questions Marketers Need to Answer to Succeed

A recent Forrester/Dun and Bradstreet survey revealed that only about half of sales and marketing decisions at B2B companies with over 500 employees are made based on data. The following summary from MarketingProfs shows that challenges to the effective use of data abound: From gathering to application to measurement, all are seen as either “extremely” or “very” challenging by a large portion of the respondents.

The paradox here is that businesses are awash in data today. Companies are spending vast amounts on software, hardware, and services related to data acquisition, management, and display. Moreover, businesses are constantly talking about data and the need to make better use of it. Unfortunately, neither these solutions nor the organizational desire solves the problem. In fact, the solutions often make the problem worse because they are seen (and sold) as a silver bullet and enable managers to check a box saying they are pursuing a “data initiative.”

The reality is that the problem lies not in a lack of will or technology. It lies in a lack of focus—a lack of strategy around data management and analytics. When you can measure anything, you do…but more measures mean more data and more difficulty making it all align. More data also mean more potential interpretations—and a decreased likelihood of consensus around what the data mean and the implications.

To construct an effective data and analytics strategy, marketers need to answer the five questions below. You’ll see that several are much more about strategy than data. Why? Because to capture and leverage the right data, we must first be clear on strategy.

  • What are the key stages of the customer journey? The point of marketing and sales is to influence behavior towards revenue generation. If you don’t know the path customers will follow that will generate revenue, what good is your data?
  • What activities are we undertaking today to move people through those stages? Marketing and sales should be targeting points on the customer journey with activities designed to influence behavior. Organizing activities along the journey ensures that you will use data about those activities and customers in the right way.
  • What should be the measure(s) of success for the impact these activities are having? Once you have aligned activities with the journey, identify ways of assessing whether those activities are successful, specifically with respect to the goal of moving from one stage to the next.
  • How can we align data we have or could plausibly get with these measures? Effectively using data to make decisions depends on being able to capture the right data consistently and with confidence. If it’s hard to acquire data or the data is subject to skepticism in the organization, it will never be used.
  • What decisions will we make based on the data? Imagine if, for a given measure or data point, one month it’s high and the next month it’s low. What decision will that trigger? If you can’t articulate the decision that will flow from the data, why capture the data in the first place?

To be clear, answering these questions is NOT easy for most companies. Data strategy isn’t easy. It’s actually much harder than buying software or hiring consultants to deliver more data and more measures through more systems. But without that hard work to decide what data and what measurements can really impact business outcomes, it’s unlikely that surveys like the one above will show any better results.