Streak’s AI roadmap
AI will be a key component of Streak's product roadmap moving forward. Learn how we've already begun building AI into Streak and what's coming next.
ChatGPT is a language model that has been used in various applications from writing assistants to customer support chatbots. While it has some obvious applications, individuals and organizations are still experimenting with it to see if it will remain a novelty or transform their workflows.
Here at Streak, we believe AI can redefine the customer onboarding process by teaching customers how to use your product or service for you.
<a href="#customer-onboarding" class="anchor-link">Customer Onboarding</a>
<a href="#challenges-in-onboarding" class="anchor-link">Challenges in the Onboarding Process</a>
<a href="#considering-ai-for-onboarding" class="anchor-link">Considering AI for Customer Onboarding</a>
<a href="#ai-in-onboarding-example" class="anchor-link">AI in Onboarding Example</a>
<a href="#ai-use-cases-in-onboarding" class="anchor-link">Use Cases for AI in Onboarding</a>
<div class="anchor-wrapper"><div id="customer-onboarding" class="anchor-target"></div></div>
Customer onboarding is the process of welcoming new customers to a business and guiding them through the steps needed to start using a product or service. It's an important part of building strong relationships with customers, increasing their satisfaction, and reducing churn.
Effective onboarding typically involves providing customers with introductory materials, such as tutorials, user guides, or online courses, and offering ongoing support and resources to help them succeed.
The onboarding process can vary depending on the product or service being offered, but it generally involves a series of steps designed to introduce customers to the key features and benefits of the product, help them set up their accounts or profiles, and provide them with the tools and resources they need to get started.
By helping customers get up to speed quickly and efficiently, businesses can ensure that they have a positive experience and are more likely to continue using their products or services over time.
<div class="anchor-wrapper"><div id="challenges-in-onboarding" class="anchor-target"></div></div>
As important as customer onboarding is, it can be tough to get right. Businesses need to be able to scale the onboarding process, and yet each customer has unique needs, timelines, and learning styles.
A few common challenges in the customer onboarding process include:
Currently, many businesses create onboarding resources that are separate from their product or service, like video courses and written articles. While these resources can be helpful, they require the customer to spend time outside of the product or service.
After watching a video or reading an article, the customer then has to go back to the product or service and try to apply what they learned. The specific task they’re trying to achieve likely differs from the example in the video or article, so they’ll need the instructions to be broad and thorough enough to help them figure out their specific use case.
One alternative is one-on-one onboarding, which is preferable to many customers but often requires a large time commitment from both customer and business. Many times, the onboarding calls and training sessions fall victim to scheduling snafus and communication issues.
While customers will always likely always derive some benefit from videos, articles, and direct support, we think there’s a new – and better – way to manage certain aspects of the onboarding process for many customers: AI
<div class="anchor-wrapper"><div id="considering-ai-for-onboarding" class="anchor-target"></div></div>
AI is a great tool to onboard customers with a variety of use cases and needs since it can analyze data, recognize patterns, and make predictions based on past experiences without being explicitly programmed to do so.
AI has the opportunity to teach customers to use a product or service by showing them how something works, rather than just doing it for them.
Building AI into your product or service allows customers to interact with it by giving instructions or prompts in natural language, and then watching how the AI solves a problem or completes a task within the product or service. The key is to have the AI show the customer what the non-natural language input is in your product and how the problem is solved, rather than just doing it for them
The first few times, a customer will take the time to type out the instructions and see how the AI solves the problem for them. Soon enough, they’ll be able to remove the AI training wheels and save time by working directly with your UI.
<div class="anchor-wrapper"><div id="ai-in-onboarding-example" class="anchor-target"></div></div>
Enough abstractions, let’s see how this would work in real life. We’ll use a hypothetical example to illustrate how AI can teach customers to use the saved view feature in Streak.
One of the biggest challenges of onboarding is complexity. Software and other types of products often require new vocabulary as a customer learns new features and systems. Integrations, automation, and setup can quickly become too technical for many end-users, leaving businesses with low adoption rates.
Our saved view editor has a series of drop-down menus that let you filter, sort, and group your pipelines into custom segments that update in realtime. In the beginning, some customers don’t know how to create the exact combination of filters, sorting, and grouping that results in the segmentation they want.
With AI, you can allow the customer to type out a description of which boxes they want the saved view to return.
Here are a few examples of what this might look like:
“Show me all my boxes in the Midwest region that have deal sizes over $150,000 and sort them in order of projected close date.”
“Show me deals assigned to me that haven’t had any email outreach in the last two weeks. Exclude deals in the ‘Closed - Won’ and ‘Closed - Lost’ stages and sort them by the date of the last interaction.”
“Show me all of the candidates who have submitted a resume and completed a phone screen for the Account Executive role and who still need to complete the take home assignment.”
The key to AI teaching a customer how to use your product or service is that it doesn’t just do the action in the background, but it takes their natural-language prompt and shows them how they should be performing the action in your UI.
Instead of just returning the correct saved view for the customer, we’d program the AI to update the UI to reflect how each drop-down menu should be configured for filters, sorting, and grouping.
As a result, the customer can see how the task should be completed and learn how the UI works from the demonstration.
The thing is, typing out natural-language instructions is slow.
With some repetition, the customer can switch to creating their own saved views using the UI. This saves them time typing out specific instructions and allows them to hone in on the exact pipeline segment that needs their attention.
<div class="anchor-wrapper"><div id="ai-use-cases-in-onboarding" class="anchor-target"></div></div>
AI is especially well-suited for software products, where the customer is using an app or web-based interface.
However, any business with a website or app-based portion of their product or service could leverage AI in the customer onboarding process.
For example, a restaurant or food service that allows online or app-based ordering could help customers build their order by suggesting dishes based on previous order history, the weather, or time of day. You could imagine a scenario where it says “last time you ordered takeout when it was raining, you liked <this dish> with a side of <side dish>, do you want to order that again or modify your order?”
Although many businesses and organizations are still experimenting with AI, we’re sure to see more and more applications of it to help with the customer onboarding process and beyond.
Learn how we’re incorporating AI into our product roadmap at Streak.