Blog

Kicksaw’s Crawl, Walk, Run Approach to Einstein for Service: A Right-Fit Roadmap for Success

If you want actionable information about using AI to improve the service and customer support aspects of your organization, we’re here to help.

Kicksaw has taken our knowledge about Salesforce’s Einstein for Service (a flexible set of AI features that pairs with Salesforce Service Cloud), combined it with our experience and observations about the growing need for intelligent service across all industries, and created a guide that will help your organization scale.

In this blog we will walk you through:

The role that AI plays in improving service
Our recommended 3-phase crawl, walk, run approach for evaluating and adopting Einstein for Service
Information about organizational suitability for each phase (i.e., recommended solution maturity, use cases, functionality details, and ROI tips)
Suggestions to help maximize solution value

‍

Service With an Impact

When it comes to service, customer care is still king, and AI’s role in providing intelligent service is key. Intelligent service is a term that refers to using AI, automation, and data to address operational tasks in a way that increases efficiency — in tandem with building customer sentiment and loyalty — to achieve business-wide benefits. Basically, a win-win.

Solutions that incorporate intelligent service can be used to help organizations:

Combine high quality and consistent service
Create a customer (patient, student, etc.) experience that is personalized, convenient, and hassle-free
‍

When done right, these solutions are the golden ticket to balancing customer care and profitability — the promise of which is why AI’s role in service has risen dramatically across all industries. Let’s talk numbers. A Forbes study recently found:

86% of consumers approve of brands using artificial intelligence,
48% of consumers are willing to pay more for quality customer service
46% of customers will buy more when given a personalized experience
‍

These numbers reflect a positive change in attitude about AI from consumers that is directly linked to profitability, and businesses are having to follow suit — but getting it right is the sticking point for most organizations. A McKinsey article with a focus on customer care and the role that AI plays stated:

Among the respondents who report that their operations are delivering better-than-expected performance, more than half have high levels of digital integration (advanced digital technologies, including AI and intelligent service)

Among responding organizations reporting that performance was in line with or lower than expected, more than 80 percent say their levels of digital integration are partial or low

Gen AI is top of mind for customer care leaders, but scaling AI is their biggest challenge

‍

Getting AI Right: Crawl, Walk, and Run With Einstein for Service

So, using AI for service is important for customer care and sustainability. It’s also important to get right and scale correctly in order to have a positive impact on your customers and business — which is where most organizations struggle.

Cue Kicksaw and Salesforce’s Einstein for Service — with a clear path forward for AI adoption and scalability.

Einstein for Service is a flexible set of AI-powered features that pairs with Salesforce Service Cloud. We’ll be categorizing those features below into starting (crawl), intermediate (walk), and advanced (run) adoption phases.

No matter what stage your organization may be in you can expect:

Improved process efficiency
Faster case triage and resolution
Better agent user experience
Increased support capacity (self-service, after-hours support)

‍

Crawl: Intelligent Recommendations

These features are a great place to start because they’re included with Service Cloud and involve less of a learning curve for users. They can serve as a base for users to build their confidence while gaining the benefits of AI functionality.

If you’re new to Salesforce, this phase is a great way to gauge your users’ acceptance of, and confidence in, AI tools.

‍

Feature: Einstein Case Classification

What it Does: Predicts Case field values based on underlying data and historical patterns

Impact: Speeds up case resolution times and resolves customer concerns more quickly

Recommendation: This feature has various levels of automation; work your way up through levels, from Recommended Top Values (less automated), to Select Best Value (more automated), and then Automate Value (fully automated)

‍

Feature: Einstein Article Recommendations

What it Does: Suggests recommended Knowledge Articles to agents based on relevance to a Case

Impact: Improves the agent experience with real-time recommendations

Recommendation: Solution maturity and data impact this feature’s benefits; a good baseline is having 3+ Knowledge Articles, 1000+ Cases, and 500+ Case-Article Attaches (less important, but helpful)

‍

Walk: Intelligent Automation

In terms of availability, some features in this phase will be dependent on your edition of Service Cloud, or may need add-on licenses.

This phase is best suited to organizations with a slightly more mature Salesforce org and/or users who are more comfortable and familiar with the use of both Salesforce and AI. These features delve more deeply into predictive and automated functionality — and allow you to adjust your digital strategy as you go.

‍

Feature: Einstein Case Wrap Up

What it Does: Predicts final Case field values based on historical cases and chat transcripts

Impact: Agents can more efficiently tackle live case and chat volume with the automatic population of recommended fields

Recommendation: Again, solution maturity will impact readiness for this feature; make sure you have enough data (e.g. 400+ closed Cases from the past 6 months) — but you won't need Chat transcripts to start, the model will start learning as they’re created

‍

Feature: Einstein Case Routing

What it Does: Further automates Case Classification by saving field predictions and running assignment rules to route cases to agents after predicting field values

Impact: Great for reporting and getting cases in front of an agent with the skill set needed to resolve them quickly

Recommendation: Implement this once you have a strong handle on your Einstein Case Classification app and its performance/accuracy

‍

Feature: Einstein Bots

What it Does: Enables chat-based interactions with bots across customer service channels

Impact: Gives customers more options for communication and lowers call center volume

Recommendation: Start with Enhanced Bots, as it’s the standard moving forward, and consider your overall omnichannel strategy before implementing — as this feature can serve as a key component for digital engagement

‍

Feature: Einstein Copilot for Service

What it Does: Empowers agents with a trusted, natural language; conversational AI assistant that can boost productivity

Impact: Completes tasks based on prompts, like generating suggestions to expedite processes, which adds up to major time savings and a better customer experience

Recommendation: This features delves into Generative AI, so it may be better suited for a ‘jog’ phase to your approach, but it’s a more beginner-friendly version of generative AI; start small and enable event logs for optimal testing and troubleshooting to scale with confidence
*Copilot uses your Salesforce solution’s own unique data and metadata, so the longer it has to process and leverage that information the better it gets

‍

Run: Intelligent Content

If your end users are experienced with Salesforce and AI, and you want to do even more with your Salesforce org’s data — you’re ready to run with advanced and Generative AI functionality. These features will require Einstein for Service or other additional licensing, but offer more opportunities for growth and benefits for the investment.

Pay attention to the order of adoption in this section, as it can impact ROI.

‍

Feature: Service AI Grounding

What it Does: Indexes fields from Cases and Knowledge Articles so that Einstein knows which information to base recommendations on

Impact: Allows you to do even more with data, in terms of speeding up processes and meeting a higher volume of customer demands

Recommendation: Get a Salesforce Admin on board to Review your Case and Article sharing settings and field-level security before enabling

*Prepares you for more accurate intelligent content generation with other features of Einstein for Service

‍

Feature: Einstein Service Replies (Email)

What it Does: Creates AI-generated email responses, grounded in your knowledge base

Impact: Dynamic, personalized responses that go further with customers and save agents the time and effort of manual input

Recommendation: Start with the easiest generative service response use case to start with, then work your way to Live Chat Service Replies & Recommendations (next features listed)

‍

Feature: Einstein Service Replies (Live Chat)

What it Does: Provides service agents with generated responses during a customer chat based on a generic model

Impact: Lets customers get back to their day faster and agents to rapidly clear case volume

‍Recommendation: Start with Service Replies using a generic model before creating your own model for Reply Recommendations

‍

Feature: Einstein Reply Recommendations

What it Does: Recommends relevant replies to support agents in the console during chat and messaging sessions, based on a custom model

Impact: Similar to Service Replies, agents have access to template responses that allow them to respond quickly in an accurate way, get through cases, and make customer happy

Recommendation: You’re ready for this feature when you have around 1000+ closed chat transcripts in a given language with 4+ chat turns; pilot this in production with select users instead of a sandbox to avoid heavy data migration

‍

Feature: Einstein Work Summaries

What it Does: Predicts and fills summary, issue, and resolution at the end of agent and customer conversations.

Impact: Provides information at a glance, allowing agents to quickly get up to speed during future interactions

Recommendation: Plan for which channels and customer interaction scenarios you want to help summarize; deployment and configuration will vary based on how/when you want users to summarize work

‍

Feature: Einstein Knowledge Creation

What it Does: Automatically generates draft Knowledge articles

Impact: Puts your data to work by automatically creating referenceable information based on your own operations, customers, service trends, etc. — leading to more accurate responses and cases resolution for customers

Recommendation: Establish a process for reviewing and approving articles drafted by Einstein and consider a phased rollout with selective permission set assignment

‍

Tips for Success

Could your organization be ready to start at the run phase of this approach? Absolutely.

Does it make sense for some businesses to stick around at the walk phase for longer than others? Of course.

Will some organizations need to pick and choose which features they adopt from each phase? Yes.

This approach isn’t meant to be an all-or-nothing strategy, it’s meant to provide basic and scalable guidelines
Salesforce is incredibly flexible in terms of features and functionality, so don’t be afraid to go feature by feature in terms of adoption when you progress to a new phase

The most important thing is adapting this approach to your organization’s maturity, capacity for change, user readiness, budget, and goals — and sticking to best practices, like using pilot groups, sticking to sandbox environments when appropriate, and remembering to account for change fatigue to improve adoption rates and ROI.

We’re great at making sure organizations get more value out of their solutions and making them more approachable, so if your organization is looking to adopt AI successfully — contact us.

‍

Additional Resources:

‍

Back to Knowledge Center