In 2013 Salesforce acquired a company called Edgespring, which was the start of the Analytics Cloud. A year later at Dreamforce we came to know Edgespring as Wave Analytics, and that later changed name to what is now Salesforce Einstein Analytics. Needless to say that regardless the names Einstein Analytics is an analytics solutions that allow you to quickly drill down into any data since data can be sliced and diced to explore the answers to your questions.
When Wave was launched in Europe in 2015 I was lucky enough to attend the Brown Belt training delivered in Paris and I’ve been in the space since implementing analytics solutions, done public speaking and delivered training, so I’ve had the chance to follow how Wave has evolved and become Salesforce Einstein Analytics. Interestingly there are a few questions that keeps coming up that dates back to the first days of Einstein Analytics.
Two main questions are “is it still expensive?” and “do you still have to manually write your own instructions?”
To answer the question if it’s still expensive, well that will be relative. To get the prices you should refer to the list prices on the Salesforce website or contact your Salesforce AE. But here’s the main thing to remember between now and then. There is no minimum licenses to be bought and the license include a lot more features than back in 2014. In fact since the Winter ’19 releas,e Einstein Analytics include Einstein Discovery and with your license you can get predictive intel – something we didn’t have back in 2014.
The second question is not that straightforward to answer. Do you still have to write your own code or instructions in form of SAQL (Salesforce Analytics Query Language) or mess around with the code (JSON) that makes up the dashboard? That depends on what you want to do. The honest answer is that you can write your own SAQL and customize any of the dashboard instructions and we shouldn’t remove that, that’s the power of the platform. But the key thing is the product team has listened to the feedback from the community and has worked very hard to bring more power to the user interface. You do not have to sit a navigate through code to modify how you are bringing data into Einstein Analytics in the Dataflow neither would you have to manually write the instructions for the mobile-optimized dashboards. It is now possible with clicks and not code to create datasets and dashboards and it’s been that way for a while. Of course, there will always be edge cases where we need to go deeper, but that is why Einstein Analytics is a platform and not just a tool.
If you look beyond the questions that dates back to the beginning of Wave, Einstein is a force to be reckoned with that gives you a lot more than you get in standard reporting. How about creating custom maps to illustrate any data, enable Salesforce Global Actions to act on the insight you get, create complex calculations with compare tables (no code needed), bring data from external sources into your dashboard or get automatic intel on what’s going on in your business? Listing all the features may sound a little abstract if you haven’t worked with Einstein Analytics before. But imagine you are selling tickets at a stadium and you want to be able to see where in the stadium tickets are selling. Well, this is where custom maps comes in. Instead of viewing a table you will see the actual stadium that is clickable to drill down into different sections of the stadium. If you want to see the quantity of the sales for this month compared to last month then compare tables can make this easy for you and when your users are exploring the dashboard they have all the Salesforce global actions available with one click directly from the dashboard.
The fact is with all this data companies are gathering across systems they are not able to get insight out of all of it without an analytics platform. While standard reports in Salesforce are great they are not geared to handle data the way Einstein Analytics is. So if you are struggling with creating derived measures in your report, maybe your report times out or you need to combine data from more than three objects I would encourage you to start looking at Einstein Analytics.
As with most of the Salesforce products, there are a vast amount of opportunities to learn on Trailhead. You can get your own org with Analytics enabled to try out the whole suite of analytics tools. To make your learning even easier there is an official “one stop to all you need to know about Einstein Analytics” website the Einstein Analytics Learning Map (www.einsteinanalyticslearningmap.com) plus if you are a customer you can ask your AE to nominate you to a two-day free academy learning the A to Z of Einstein Analytics. Also if you want a little more guidance on how to get started I wrote a blog to help people that are embarking on their Einstein Analytics Journey (http://www.salesforceblogger.com/2018/02/05/embarking-on-your-einstein-analytics-journey-start-here/).
Service Cloud is being bolstered yet again to further ingrain Einstein AI capabilities into the product – a next generation of Service Cloud Einstein, if you like. Salesforce have the figures to prove their move: the Salesforce “State of Service” report surveyed over 1,900 industry leaders, with a high proportion (88%) of ‘high-performing’ organisations committing significant investment towards service over the following year.
What’s next for AI-powered Service Cloud? Salesforce have coined the term: “Proactive Service”, with the pumped-up Einstein for Service offering positioned to deliver this transformation.
Einstein for Service will bring a number of new features, that can be added to existing Service Cloud deployments. The move makes intelligent customer service evermore accessible to enterprises, the new features include:
- Einstein Case Routing
- Einstein Article Recommendations
- Einstein Reply
- Einstein Next Best Action
Einstein Case Routing
The Case Comes In
Directs cases to the best-suited queue or agent, able to cater to more complex routing than previously possible; factors such as agent skill-set and past resolution success rates are considered for precise case assignment – including factors not visible to the ‘naked eye’.
Benefit from optimal agent-case match, leading to faster resolution time. Consider it a weight off everyone’s shoulders knowing that routing is continuously optimised despite your diversifying product portfolio, team changes, or simply increasing customer expectations – all without admin intervention.
The Case Gets Actioned
We shouldn’t side-step the human element in Customer Service, especially when it comes to receiving customer requests written in free-form text; after all, shouldn’t customers be able to communicate naturally with your brand?
Einstein uses Machine Learning to constantly learn from every interaction your service teams have. What did they do resolve the case faster? What did they say in response to specific enquiries?
Crunching all of these interactions, Einstein can recommended replies by using natural language processing (NLP) to decode the content of the customer request. Einstein also has the ability to be very specific, tailoring suggestions based on the conversation and customer context.
Einstein Article Recommendations
The Case Gets Explained
The Einstein learnings aren’t just limited to replies. What Einstein’s Machine Learning discovers can be applied to other service scenarios: take knowledge article recommendations, for instance.
You may have seen this in play on Salesforce’s own customer support hub, where ‘Blaze’ appears with a number of suitable knowledge articles. By identifying the best knowledge article, customers can self-service more effectively, being able to locate the information they need faster, without trawling through your knowledge base or searching multiple keywords.
Einstein Next Best Action
The Case Goes Further
Using a magic combination of business rules (set by a human) and predictive intelligence (the workings of Salesforce Einstein), a next best action, when recommended to agents in ‘the heat of the moment’, can deliver maximum impact when the case reaches critical checkpoints. An example could be at the point of case resolution, to suggest a relevant cross-sell opportunity, such as an extended warranty.
Quip for Service
The Case Needs Backup
A bonus – in addition to the 4 Einstein-branded features, Salesforce also announced that Quip is now embedded within Service Cloud. Quip for Service will place the Quip collaboration platform directly into the case object, for when a case needs backup from multiple agents, also known as ‘swarming’. Quip templates can be accessed by agents to communicate efficiently with customers. In less high-pressure moments, Quip for Service can also be effectively used for discussing and editing knowledge articles.
Here’s a recap of previously existing Service Cloud Einstein features:
- Einstein Bots: AI-powered chatbots, configured on-platform, that learn from agent feedback to the bot if it has, or hasn’t, got it right.
- Einstein Case Classification: populate picklist and checkbox field values for new cases based on past data.
- Lightning Flow for Service: a tool to design and activate automated processes.
- Einstein Supervisor: predictive analytics for Service Agent Managers.
The 4 new Einstein for Service features mean intelligent customer service is increasingly accessible to service teams; helping the move from case-centric Agents, to customer-centric ambassadors. I predict the benefits would particularly resonate with organisations that have complex or rapidly diversifying product portfolios; for teams with high-turnover or niche-skill sets; or simply to keep up with the raising benchmark in customer service expectations.