Skip to Main Content
HBS Home
  • About
  • Academic Programs
  • Alumni
  • Faculty & Research
  • Baker Library
  • Giving
  • Harvard Business Review
  • Initiatives
  • News
  • Recruit
  • Map / Directions
Working Knowledge
Business Research for Business Leaders
  • Browse All Articles
  • Popular Articles
  • Cold Call Podcast
  • Managing the Future of Work Podcast
  • About Us
  • Book
  • Leadership
  • Marketing
  • Finance
  • Management
  • Entrepreneurship
  • All Topics...
  • Topics
    • COVID-19
    • Entrepreneurship
    • Finance
    • Gender
    • Globalization
    • Leadership
    • Management
    • Negotiation
    • Social Enterprise
    • Strategy
  • Sections
    • Book
    • Podcasts
    • HBS Case
    • In Practice
    • Lessons from the Classroom
    • Op-Ed
    • Research & Ideas
    • Research Event
    • Sharpening Your Skills
    • What Do You Think?
    • Working Paper Summaries
  • Browse All
    Use Artificial Intelligence to Set Sales Targets That Motivate
    Research & Ideas
    Use Artificial Intelligence to Set Sales Targets That Motivate
    22 Oct 2019Research & Ideas

    Use Artificial Intelligence to Set Sales Targets That Motivate

    by Michael Blanding
    22 Oct 2019| by Michael Blanding
    Setting sales targets has always been an inexact science, with serious consequences if done poorly. Using AI-based advanced analytics might be the answer, argues Doug Chung.
    LinkedIn
    Email

    Setting the right sales targets for employees is a difficult balancing act, with long-term consequences on growth and morale.

    Setting a target too low, making it easily achievable, might cause an an employee to not put in the effort. Setting a target too high can be equally problematic. “Then there is no chance of meeting it,” says Doug J. Chung, MBA Class of 1962 Associate Professor of Business Administration in the Marketing unit at Harvard Business School. "The salesperson will be discouraged, and just as unlikely to work to their full potential."

    Chung’s prime area of research lies in finding the sweet spot between these two undesirable outcomes, and determining how compensation can motivate salespeople best. In a recent Harvard Business Review article, Chung and several executives from the consulting firm McKinsey & Co explored a new way to thread that needle: using advanced analytics that incorporate artificial intelligence (AI).

    “Chung has seen companies dramatically improve productivity after adopting advanced analytics to guide compensation.”

    In an ideal world, a company would use trial and error to set the best sales targets for employees, experimenting until they hit the right formula. In reality, that’s problematic, says Chung.

    “If you want to know if a compensation plan is working or not, you need to change the compensation plan and observe and measure the change in productivity, using a control group of employees that does not experience the change,” he says. “Firms typically don’t want to do that because once you change it, it’s very difficult to change it back. Also, experimenting with a select group of employees is deemed as unfair.”

    Past performance doesn’t guarantee future returns

    Most companies rely on past performance, setting an employee’s goal slightly higher than their sales for the previous year, a term known as “ratcheting the quotas.” But that, too, can have its drawbacks.

    “Suppose I got over the quota by 20 percent, then my quota next year will be 120 percent of this year’s,” Chung says. That approach encourages employees to do only the minimum necessary to get their bonus, lest the company sets a too-high goal next year.

    “An employee says I know I can do 120 percent, but I am not going to do it,” Chung says. “You are penalizing the high performers.”

    Besides, a compensation plan that ratchets up slowly at a consistent rate, related to the firm’s growth objectives, only works in a stable market, in a stable industry, in a stable region. “That’s a lot of ‘ifs,’” Chung says.

    Putting artificial intelligence to work

    Artificial intelligence—an array of approaches that rely on computer systems and algorithms to handle human tasks—allows companies to use multiple variables to compute the best targets for employees, often in real time. Many companies have started using machine-learning algorithms to construct AI systems.

    Such algorithms can take two forms:

    • Supervised learning. In this approach, using labeled data, the algorithm learns to predict the outcome from the input data.
    • Unsupervised learning. In this case, using unlabeled data, the algorithm learns the inherent structure from the input data.

    Either machine learning approach is only as good as the humans who are setting the objectives and managing the data. Companies that succeed with AI tend to:

    Consider the company’s broader goal. It’s important that the company first determine its overarching goal, Chung says. “It could be that we need to increase our revenue by 10 percent, even if we incur a loss, or it could be our revenues are fine, but we need to improve our profit,” he says.

    For example, a US industrial service company was losing 20 percent of its customers every year. Through research, its leaders realized that when customers stayed with the company for six months, their odds of leaving declined. Reducing that churn became part of that company’s overall goal. “They put in a proxy that they wanted to not only acquire customers, but retain them as well.”

    Identify and collect as much relevant data as possible. Once executives set an overall goal, they can feed data into an AI system that sets an individual goal for each employee. “The more data that goes in, the more accurate a prediction we can have,” Chung says.

    Variables that companies might use include employee data, such as average and past year’s sales; macroeconomic indicators, such as GDP growth and consumer price index; and industry-specific measures, such as growth of competing companies or countries. Firms can also process and use unstructured data, such as text from documents.

    “There is so much data we can use,” Chung says. “For example, average temperature—you may think it has nothing to do with sales, but if the temperature is too high, you don’t go into the field so often, so it does.”

    Develop an algorithm that goes deep. Large firms may be able to design their own in-house machine learning algorithms to employ their AI system. Smaller firms can contract one of the many consulting companies that offers AI services.

    Whatever route a company goes, the right algorithm will use the data to generate a number for each employee. That figure can include an overall target, as well as specific targets for each product line to maximize an employee’s productivity while striving towards the company’s overall goal. While this process is automated, field sales mangers might need to fine tune the final number.

    From resistance to productivity

    The success of using such a system, Chung says, depends on two things: the amount of data a company includes, so as to create the most accurate measures, and management’s confidence, which helps foster buy-in by sales employees.

    “You get a lot of pushback at the initial stages,” says Chung. “You can see people who get a higher quota saying, this is not right.”

    When companies stick with the program, however, Chung has seen such resistance from employees lessen over time.

    About the Author

    Michael Blanding is a writer based in the Boston area.
    [Image: stock_colors]

    Related Reading

    • What's the Ideal Frequency for a Sales Quota?
    • Money and Quotas Motivate the Sales Force Best
    • Why Salespeople Struggle at Leading

    What do you think of this research? Share your insights below.

    Post A Comment
    In order to be published, comments must be on-topic and civil in tone, with no name calling or personal attacks. Your comment may be edited for clarity and length.
      Trending
        • 08 Sep 2022
        • Book

        Gen Xers and Millennials, It’s Time To Lead. Are You Ready?

        • 25 Jan 2022
        • Research & Ideas

        More Proof That Money Can Buy Happiness (or a Life with Less Stress)

        • 25 Feb 2019
        • Research & Ideas

        How Gender Stereotypes Kill a Woman’s Self-Confidence

        • 17 May 2017
        • Research & Ideas

        Minorities Who 'Whiten' Job Resumes Get More Interviews

        • 28 Mar 2023
        • Research & Ideas

        The FDA’s Speedy Drug Approvals Are Safe: A Win-Win for Patients and Pharma Innovation

    Find Related Articles
    • Information Technology
    • Forecasting and Prediction
    • Salesforce Management
    • Sales

    Sign up for our weekly newsletter

    Interested in improving your business? Learn about fresh research and ideas from Harvard Business School faculty.
    This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
    ǁ
    Campus Map
    Harvard Business School Working Knowledge
    Baker Library | Bloomberg Center
    Soldiers Field
    Boston, MA 02163
    Email: Editor-in-Chief
    →Map & Directions
    →More Contact Information
    • Make a Gift
    • Site Map
    • Jobs
    • Harvard University
    • Trademarks
    • Policies
    • Accessibility
    • Digital Accessibility
    Copyright © President & Fellows of Harvard College