Onboarding to Machine Learning: Mapping Sales Processes

Improving a sales process with machine learning starts with a straightforward assessment of the existing content, including video, audio, text, graphics, and training, a company uses to onboard a new distributor to its policies and practices. These first steps, which set the stage for confident selling by new distributors, are essential to improving sales success during the first two weeks with a new direct selling company. People who close their first sales within 14 days earn an average of 71 percent more than a distributor who takes just four weeks to complete a sale.

Sales and marketing leadership tackling machine learning for the first time need to break their existing onboarding practices and initial selling activities into steps, then organize those steps into collections that are expected to produce a specific result that can be measured. We recommend assembling a map of the onboarding, training, and sales support experience for new distributors, as their immediate success will produce immediate improvement in revenue and profitability results. Tier your product content in terms of 1.) Company overview and welcome programs and content; 2.) Selling materials and programming for distributor use; 3.) Deep product information, such as sales sheets or detailed product knowledge videos.

Break down the first month of distributor experience into:

• Onboarding: Introduction to the company, its mission, and selling process at the overview level — what you most want your new enrollees to know on Day One and to have internalized by the end of Week One.
• Prospect Development: This is the first, most important step for a successful sales enablement tool. Rather than explain how to use the contact management tools, get the distributor to work immediately on adding prospects and following up.
• Product Knowledge Development: Ongoing and frequently updated, product knowledge and product-specific training.
• Sales Skills Improvement: If there is sales training content that is not product-specific, such as coaching on how to follow up or present at a meeting, these programs will be useful throughout the entire distributor lifetime, not just as they become familiar with the company.

We suggest beginning with a list of all existing content. Write the title of each asset on a sticky note and, on a second note, the goal for the asset, such as “Create a sense of welcoming support” or “Establish product- and lifestyle-claims policy.” Place the two sticky notes, asset and goal for the asset side by side. Examine all the content related to onboarding to see if there are multiple assets seeking to achieve the same outcome.

As common goals are identified, cluster the content assets by the expected outcome. It is likely there will be several assets that drive to the same distributor goal, and these variations are natural places for a machine learning content system to start testing to see which content assets are most effective.

Introductory content, such as a generic welcome message and overviews of the company, should be separated from practical how-to content related to using tools and services offered by the company to refine distributor sales skills. The latter training content will distract distributors from mission-centric learning. For example, most direct selling systems begin with a series of introductory videos about the company, its products, and how the distributor can start to work its selling process. These videos set the stage for future training, but they have a narrow set of goals: To build confidence in the distributor that they’ve made the right choice of product or service to sell, that the company is reliable and supportive of their success. This is essential for winning younger distributors’ loyalty.

With mission- and policy-centric content organized into the first category, the next step is to organize each of your sales task workflows for use by the machine learning platform.

Each days’ distributor training activities during the first two weeks must have a goal, such as confirmation that the new distributor understands the basic value proposition and mission of the company or that they enter and start communicating with prospects. And each day’s activities should contribute to the next day’s goals — if on Day One, the distributor enters five contacts, Day Two should include follow-up activities and content that help convert those leads to a call, presentation, or online meeting.

Look for multi-day processes, such as prospect development and determine whether multiple assets address the same steps and issues. These are convenient reference points when thinking about how to shorten and improve onboarding programming, which can produce immediate improvements in distributor success. Sales process steps in a “What’s Next” machine learning tool allow the distributor to focus on doing sales work instead of learning how to use tools.

Once the Welcome and Onboarding workflows are complete and redundant content identified for testing, the organization of product knowledge and sales skills coaching content if there is any in the current asset library. These are content categories that can be populated over time, as well as licensed from training providers for integration with sales coaching machine learners, which can target sales training based on the distributor’s sales challenges. For instance, if they consistently add contacts, get meetings, but don’t close, the tool can direct the distributor to training videos about closing, getting commitments, and handling objections.

With a smart platform in place, a variety of training programs can be added to address your network’s training needs and to address individual distributor challenges. In the next installment, we’ll explore attribution modeling for machine optimization of each step in the sales process.