Blog

Will they show up? Fighting pandemic summer melt by engaging attitudes

By Keith P. O’Brien

6 minutes estimated time to read

First, the point of this blog series isn’t to confirm the extraordinary challenges universities face in enrolling the Fall 2020 class. This series shares the ongoing experience of one regional private university in the Mid Atlantic, and their approach to forming an incoming class in these unprecedented times. The university’s challenges—drive deposits, battle summer melt—are commonplace today but the approach to engaging admitted students is rare: understanding their attitudes or mindsets towards matriculation. And then personalizing communications to those attitudes.

COVID-19 Undercuts all Predictive Modelling: New Data Needed

Is there an admissions or enrollment management department that isn’t awash in data on admits? Yet COVID-19 has blown away all the conventional wisdom on student behavior and class projections as students and families grapple with economic contraction, public health concerns, and disparate reopening timelines. Traditional predictive analytics aren’t helpful when decision making is upended by a pandemic and a rapid recession.

People’s decisions are driven by their attitudes, which are rooted in values, not zip codes, income level, age, or race. If you are or know a procrastinator, this mindset shapes how choices are viewed and selected. If Admissions knew in advance that a student’s mindset around college selection was procrastination, then tailoring messages to that mindset would likely be more effective. Attitudes are dynamic and can be influenced by messages that nudge people to make different choices.  Procrastinators can be motivated to act by presenting choices and support appropriately.

Earlier this year and before the pandemic had erupted, the regional private (let’s call it Private U) asked Discourse Analytics (DA) to profile the attitudes of their admitted students using existing behavioral data in Slate and other systems of record. Put simply, a student’s attitudinal profile illuminates the reasons behind an enrollment choice—depositing, registering for orientation, going somewhere else, or taking a gap year.

Admitted Students’ Enrollment Mindsets 

DA’s artificial intelligence-platform took behavioral data on 3,000 admits (no personally identifiable information was used) creating a profile of each student’s mindset. While every student has a unique attitudinal profile, there are likeminded people and the algorithm identified four “Think-alike” clusters based on shared attitudinal attributes. Here’s an overview of the four enrollment mindsets among Private U’s admits:

These four mindsets aren’t specific to Private U. DA’s AI and machine-learning platform use a set of attitudinal dimensions specific to higher ed to profile each student’s attitudes. Therefore, it’s very likely that these four mindsets exist among your own institution’s admitted student population.

The “Think-alike”clusters gave Private U an unprecedented view into their admitted student population:

1. Predicting how students will behave as their attitudes provide the “why” for their actions;
2. Identifying message strategies to engage the attitudes that motivate students to take the desired action; and
3. Enabling personalization at scale through messages tied to the cluster the individual admit belongs to.

Student attitudes aren’t static: the admitted students’ mindsets react to the stimuli provided by Private U’s attitudinal messaging.  DA’s algorithms use the student response to communications (open rates, click throughs, follow ups) as additional inputs into that student’s attitudinal profile. Behaviors can be seen but attitudes must be revealed. As these behavioral data get richer, the AI rescores the attitudinal profiles and updates the clusters regularly.

The insights into the student mindsets helped Private U to begin to understand their admitted population differently.  But what will Admissions do with this information?  How will they convince these students to deposit, to not withdraw and ultimately to matriculate in the fall? 

In the next blog, we’ll share how the mindset clusters continue to evolve as student (and parental) attitudes shift with the ongoing tumultuous changes in the world and their effects on higher education. Additionally, we’ll look at how Private U uses the DA outreach recommendations to radically shift their messaging approach to students.

Download our free whitepaper,  Why Attitudes Drive Decision Making: Reimagining Personalization  to understand how student attitudes and “think-alike” mindsets explain student behaviors. 

To learn more contact: [email protected]

Will they Deposit? Mindsets, May 1st and Melt

By Keith P. O’Brien

7 minutes estimated time to read

Blog two in a series

As mentioned in our previous blog, Private U, a regional private university in the Mid Atlantic, is navigating the challenges in recruiting its undergraduate class for fall 2020. Challenges that are exacerbated by the public health crisis and attendant recession. Private U is employing a new approach to get admitted students to deposit by personalizing communications to each student’s mindset or attitudes, not their demographic characteristics. Next we’ll share some of the initial results.

Nudging Student Mindsets in a Pandemic: Early Findings

Private U deployed its first attitudinal messaging in April with calls to action focusing on the traditional May 1^st deadline; like many privates, the deadline was extended given the circumstances.

Pandemic Affects Student Mindsets

COVID-19, shelter-in-place regulations and the economic downturn clearly affected admitted students and their families. Consequently, Private U received updated financial data for parental income, Pell-grant status, reported EFC and work status. These new data are ingested by Discourse Analytics’ (DA) artificial intelligence (AI) model and the algorithms refine the scoring of each student’s attitudinal profile. The model is updated with the student response to the attitudinal messages.

Large shifts occurred in the “Think-alike” clusters as societal and economic volatility reshapes attitudes towards enrollment. Two of the larger changes involved:  

• The number of “Laggards” grows as uncertainty increases while motivation and financial comfort weakens
• The number of “Strivers” contracts as confidence wanes and fall 2020 plans become tentative

An entirely new mindset cluster emerges—“Ambivalent Searchers.”  Students who are indecisive and risk averse and try to accumulate college-going options to manage uncertainty. Despite being admitted, they tend to have low affinity for many of these colleges and so manage their options quite passively. Private U must gain these students attention, build a connection, and motivate actions.

Most Viable Enrollment Outreach Strategies Emerge

Private U’s messaging encouraging depositing is guided by outreach strategies tied to attitudes. DA’s AI-model assigns one of four outreach strategies to each admit. The two most relevant strategies are:

1. Career Outcomes (about 60% of students get these messages)
2. Fit for Me (about 25% of students receive these messages)

The career focus is unsurprising as COVID-19 stokes uncertainty about job security and unemployment. This prominence also reflects issues around affordability—the cost of private education must deliver well-paid job opportunities. The fit between Private U and the student’s desired campus experience is always a major factor in determining college choice. 

Outcomes: Deposits Trending Positively

Private U’s attitudinal communications boosted deposits through April and the May 1^st deadline. The deposit numbers are impressive given:

• The irrelevance of traditional predictive modelling given pandemic induced volatility; and 
• The withdrawals created by COVID-19’s impact on families and finances.

From Mindsets to Messaging: How Private U Does It

DA helps Private U develop communications targeting admitted students in stages. First, DA’s AI-model evaluated each student’s individual attitudinal profile and grouped likeminded students (sharing attitudes) in these four think-alike clusters 

The first blog describes the four clusters.

Next, the AI assigned an outreach strategy—categorical messages that nudge students to enroll—to every admitted student. The following graphic summarizes the AI model’s steps and outputs.

Four Enrollment Outreach Strategies

Based on the student attitudes shaping enrollment outcomes, DA developed four outreach strategies, which are applicable to nearly all 4-year institutions:

1. Career Outcomes targets students prioritizing the ability of college’s graduates to obtain good jobs. The outreach underscores Private U’s portfolio of career exploration, career services and experiential opportunities that enable students to acquire real-world work experience. 
2. Invest/Affordability engages students wanting to ensure the value of their degree justifies the cost of a private college. The outreach underlines Private U’s high-quality private education and its affordability. 
3. Core Academics targets students focused on academic reputation. The outreach highlights the range of liberal arts and professionally focused programs enabling each student to pursue the major that matches their skills and interests. 
4. Fit for Me targets students assessing the fit between Private U and their desired college experience. The outreach highlights how the environment, learning opportunities and campus services enable a student to find the right place for their learning style, social character, intellectual interests and talents.

The next stage involves developing messages activating the different outreach strategies.

Creating Personalized Messaging aligned to Student Attitudes

Private U’s Admissions and marketing teams develop messages to nudge admitted students to deposit. The messages target student attitudes by:

Step One—Understanding the attitudes characterizing the four mindset clusters; and

Step Two—Developing a set of four outreach strategy messages for each mindset cluster. This entails shaping the subject lines, the language, the calls to action, and positioning of Private U’s value proposition to the mindset clusters.

Private U’s communication plan for admitted students is summarized in the following attitudinal message portfolio:

The message portfolio is updated weekly based on DA reports identifying the number of students in each mindset cluster and their assigned outreach strategy. Additionally, DA provides a weekly payload document specifying the outreach strategy message for each non-deposited student in Private U’s CRM system, Slate.

DA’s ongoing reporting enables Admissions to prioritize the students (clusters) and the messaging (outreach strategies) within its communication plan.

In the next blog, we’ll discuss the (on-going) lessons from Private’s U summer melt campaign, including:

• Using nudging (behavioral economics) in crafting the outreach messages
• Crafting deposit calls to action that tie to other enrollment steps

To learn more contact: [email protected] 

Private U’s summer melt campaign exemplifies Nudge Tech, using AI to personalize nudges that improve student outcomes.  Nudge Tech and Discourse Analytics feature in Gartner’s Top 10 Strategic Technologies Impacting Higher Education in 2020.

Gartner stresses value of Higher Ed Nudge Tech: 6 recommendations

By Keith P. O’Brien

6 minutes estimated time to read

Gartner, the research and advisory firm, recently published its Top 10 Strategic Technologies Impacting Higher Education in 2020.

Nudge Tech is prominent on the list and defined as: “a collection of technologies that work together to achieve timely, personalized interaction with students, staff and faculty, such as a just-in-time text (SMS) reminder for class. Technologies used include chatbot, texting, algorithmic analytics, machine learning and AI CI [Conversational Interfaces].”

Nudge Tech represents a remarkable fusion of advanced technologies and behavioral economics: AI and data analytics activating insights on human decision making—nudging. In Nudge: Improving Decisions about Health, Wealth, and Happiness, Thaler and Sunstein define a nudge as “any small feature of the environment that attracts people’s attention and alters their behavior but does so in a way that doesn’t compel.”

All but the most selective Higher ed institutions face numerous challenges to their continued operations from the immediate effects of COVID-19 to the long-term demographic declines in college-going populations. Nudge Tech must be a component in student engagement strategies seeking to boost enrollment and success outcomes. 

“Above all, nudge tech is a concrete example of how to achieve personalization at scale, which is becoming a key competitive advantage in an increasingly global digital education ecosystem.”

Lowendahl, Jan-Martin and Morgan, Glenda. “Nudge Tech.” Top 10 Strategic Technologies Impacting Higher Education in 2020. (March 2, 2020)

Uniquely within nudge tech providers, Discourse Analytics delivers attitudinal personalization at scale. Our key insight is that humans make choices based on how they think, not their demographics. We harness artificial intelligence to deliver attitudinal personalization: nudges tailored to each person’s attitudes towards an issue. 

For the second year in a row, we are honored to be referenced by Gartner in their analysis of Nudge Tech.  

Gartner’s Recommendations on Using Nudge Tech in Higher Ed

Gartner provides a set of recommendations for higher ed CIO to capitalize on nudge tech, which are applicable to all institutional leaders seeking to improve enrollment management, advancement and student success. These six recommendations align with DA’s core operating principles:

1. Make sure to understand nudge theory, as well as pros and cons of nudging. Look for a solid understanding of nudge theory from the vendors you invite. DA’s work is predicated on the research of leading theorists, including Amos Tversky, Daniel Kahneman and Richard Thaler. 
2. Ensure users’ trust by implementing transparency policies and interfaces, as well as opt-in and opt-out procedures for personal data. DA’s AI-platform only uses existing behavioral data on a university’s systems of record. Moreover, DA’s model does not use any Personally identifiable information (PII) such as name, age, gender, race for any student. 
3. Build a nudge tech business case by identifying narrow use cases that have clearly measurable outcomes on which to judge value. DA’s attitudinal nudging has helped colleges and universities enhance outcomes across a range of challenges—summer melt and yield management, FAFSA verification and financial aid processing, student persistence and retention, and alumni relations and advancement. 
4. Assess suitable data quality by identifying data sources, such as Q&A forums, SISs and LMSs that can act as the foundation for machine learning. DA ingests available student behavioral data from systems of record including the LMS, SIS, the CRM, and card swipe data from libraries and gymnasiums. 
5. Design for a virtuous learning cycle by capturing all interactions in a machine-readable format. DA’s AI-model uses machine learning to map new behavioral data, including responses to nudges, to the student’s individual attitudinal profiles throughout the engagement. 
6. Design the implementation so that statistically valid data can be collected for the “test group” as well as control group. The gold standard is to do a randomized controlled trial. DA benchmarks the results of every nudge campaign against a randomized control group from the target population.

Now more than ever, higher ed institutions must use AI to unlock the student insight buried in their systems to personalize communications. As Gartner emphasizes, “AI is a particularly good case for nudging, as it allows analyzing increasingly complex data for opportunities to impact behaviors.”

To successfully remain relevant and financially viable, institutions should deliver more student-centric personalization–motivating mindsets to attain desired outcomes.  It’s time to embrace Nudge Tech. 

To learn more about DA’s approach to Nudge Tech, contact: [email protected]

Download our free whitepaper,  Why Attitudes Drive Decision Making: Reimagining Personalization  to understand how student attitudes and “think-alike” mindsets explain student behaviors. 

The New At-risk Student: 4-steps to Nudge Attitudes towards COVID-19 Compliance

By Keith P. O’Brien

7 minutes estimated time to read

64% of colleges are planning for an in-person fall semester [as of June 22, 2020]

The Chronicle of Higher Ed

Higher ed institutions are moving towards on-campus instruction this fall through a hybrid of face-to-face and on-line offerings. Institutional leadership must navigate their duty of care towards students and staff while crafting a reopening plan that is deliberative, comprehensive and “normal.”

 Student Compliance is the Achilles Heel of Reopening

“[y]oung people – aged between 18 to 31-years-old – had the lowest compliance rate at 52.4 percent… young people, the group least at risk for COVID-19, displayed more anxiety in their survey answers than other age groups, using words like “anxious,” “disturb” and “nervous,” more frequently than other age demographics.” 

Melissa DeWitte, Stanford News,  April 14, 2020, discussing a Stanford University-led study on social distancing. 

A quick review of university websites turns up similar messages about social distancing, testing, face coverings, protective equipment, etc. All sensible and necessary steps which should lead to “herd compliance.” Student compliance with these restrictions and procedures is critical to the viability of on-campus education. The issue remains, however, what will the impact be if some percentage: 20%, 30% or 4% do not comply. How will institutions spur new and returning students to adhere to what will be dramatic changes to traditional campus behaviors and attitudes? 

Here are four steps colleges and universities can take to boost compliance: 

1. Embrace Attitudes as the Why of Student Behaviors
   Kelly McGonigal in The Upside of Stress writes:  “Your mindset creates your reality….Mindsets are your beliefs about how the world works that shape the reality you experience. Think of them as the operating system for you mind — operating below the surface but controlling everything that happens.”
   Students will assess choices and make decisions based on their mindsets towards COVID-19 compliance. Two students with the exact same socio-economic and demographic characteristics can be very different in how they make decisions: one may be individually focused and the other community focused. Almost every university would communicate to them the same way and the results are likely to be mixed. What if the messages were tailored for activating that student’s mindset? 
2. Pinpoint Attitudes towards COVID-19 Compliance
   Discourse Analytics’ (DA) Artificial Intelligence model uses existing student behavioral data to identify the attitudes driving a student’s behavior. The algorithms deliver a student attitudinal profile that identifies those whose attitudes make them a “compliance risk” with new health and safety rules.
    
3. Message Students based on Shared At-risk Attitudes
   DA applies a patented “Think-alike” engine to form clusters of students based on shared attitudinal attributes. The “Think-alike” clusters predict how likeminded students will behave towards health measures and deliver prescribed messages engaging the attitudes that motivate each individual to minimize risky behaviors.
    
4. Nudge Students to Comply
   Gartner, the research and advisory firm, recently wrote, “Nudge tech is used to impact key institutional priorities… and it enables key business trends such as personalization at a cost-effective scale.”
   Nudging solves for the reality that student decisions are not always optimal and how the choices are framed strongly influences the outcomes. Nudge tech serves to leverage existing digital channels to improve adherence through personalized and scaled calls to action. 

As fall approaches, ensuring campuses provide a meaningful college experience for staff, faculty and students, while minimizing health and safety risks is fraught with challenges. Depending upon traditional approaches to student engagement for unfamiliar and rigorous public health requirements seems unlikely to shift behaviors and reduce the risks.

To learn more about Discourse Analytics and activating student mindsets, contact: [email protected]

Click here to read about Gartner’s research on Nudge Tech in their Top 10 Strategic Technologies Impacting Higher Education in 2020.