masterzphotofo - Fotolia


Salesforce AI touts intelligence for everyone

Salesforce AI has come to the fore with Einstein, its intelligence application. The CRM provider attempts to round out its portfolio with this accessible AI addition.

Recently Salesforce announced Salesforce Einstein, an artificial intelligence and analytics technology, is now part of the core of Salesforce's portfolio, the Customer Success Platform.

While elements of Einstein are new, it's also an umbrella for new and existing artificial Intelligence (AI) capabilities within each of Salesforce's offerings, including Marketing Cloud and a recent development, Commerce Cloud (formerly Demandware). Capabilities include Salesforce AI built into products as well as advanced reporting tools.

Like other vendors, Salesforce is making a serious play to bring AI capabilities natively to its platform. When companies can bring intelligence into their applications, they can forge stronger relationships with customers, provide more meaningful customer service experiences, boost revenue and make operations more efficient. Salesforce is banking on Einstein to enhance the data and value of its existing sales, service and marketing applications, and position it well against competitors in the marketplace.

Artificial intelligence

Artificial Intelligence simulates human intelligence and includes learning, self-correction, reasoning and understanding of language. Many marketers and others are overwhelmed with the amount of data coming in. So, rather than replace humans, Salesforce AI can help marketers better understand data so they can predict what will happen in the future and, ultimately, better perform their jobs. AI can also help to automatically manage prospect interactions with little or no human intervention.

AI is now possible thanks to the convergence of big data (massive amounts of collected and available data), inexpensive and plentiful processing power often delivered through software as a service or online subscription models, and powerful data models. During the past few years, AI has moved closer to the systems that run the business. Today's advances have enabled AI to go beyond reporting to drive real-time interactions within a company's operational systems.

Salesforce Einstein

According to Salesforce, Einstein is powered by machine learning, deep learning, predictive analytics, natural language processing and smart data discovery. The company also says that it includes models that will be automatically customized for every customer and that it will learn and get smarter as new interactions and data come in. Unlike third-party add-on products, Einstein capabilities are embedded within Salesforce products.

Einstein includes new marketing capabilities, including the following:

  • Predictive scoring to gauge a lead's likelihood of email engagement, unsubscribing or making a purchase.
  • Predictive audiences to build segments of people with common predicted behavior.
  • Automated send-time optimization to deliver messages when prospects are most likely to engage with them.
  • Adaptive customer journey to automatically adapt the journey for each customer.
  • Determining the next best action such as the next product, content or offer.
  • Wave analytics to uncover new patterns and insights.

Salesforce AI benefits

Salesforce Einstein promises to bring new analytics and AI capabilities to marketers and retailers using the Salesforce platform.

As Salesforce AI morphs from a stand-alone tool to a core element of operational systems, it has the potential to transform the role of traditional enterprise systems like CRM. Rather than serving as static automation systems and systems of record that deliver improved tracking and efficiency gains, new AI-driven systems like Einstein may drive greater marketing effectiveness and revenue growth for companies. Einstein should help marketers in other ways as well:

Fewer data scientists required. It is not easy to hire data scientists who are in heavy demand today. Instead of hiring data scientists to develop models for custom apps, brands can rely on Salesforce AI and other vendors' software with native AI capabilities -- that is, artificial intelligence that is built into platforms. It can take at least several weeks to develop a model for third-party predictive analytics software, so this makes it that much easier to incorporate AI into marketing efforts, and do it in less time.

No integration needed. Einstein AI capabilities are available without integration, which should save time, expense and complexity for Salesforce customers. What's more, since Einstein is native to Salesforce, it uses the same data models that customers already use, which in many cases, will eliminate the need to build the custom data models used by many analytics programs.

Less work for marketers. As Quentin Hardy of The New York Times has said, "A technology truly matures when it disappears." More and more, that is exactly what is happening with analytics. Stand-alone and/or visible analytics are disappearing into the bowels of modern systems. So, more and more of the massive processing and low-level, repetitive "thinking" is performed by systems on the back end, like Einstein. This may free marketers to focus on high-level issues and strategies instead.

Fewer dots to connect. Running a sophisticated campaign can require connecting a lot of dots -- such as pulling in data from other systems, setting up special behaviors and user flows, or creating specific segments. When AI is built in, more and more of those dots are already connected, freeing marketers to focus on other issues and to run more sophisticated campaigns.

Shorter time to marketing at scale. You will know that a new technology like Einstein is successful when marketers have successfully tested it and are actively using it to power marketing programs at scale within large swaths of their customer and prospect bases. If Einstein requires less time to integrate and less work to develop campaigns, it should also require less time to market, regardless of campaign size. And that would speed payback as well.

A more personalized user experience. Einstein should help marketers deliver more personalized and relevant experiences based on sophisticated models that incorporate past behavior, data from other similar customers and more.

Better marketing results. If Einstein is successful, all of this can reduce the cost of standing up, running, and using marketing and commerce systems, which will improve ROI. But the big win could come in the form of more effective marketing and merchandising, leading to profitable, new revenue growth.


While Salesforce AI has tremendous potential for marketers and online retailers, you should also dig deeper into a few areas before buying into the Salesforce Einstein hype.

Branding or innovation? Salesforce Einstein brings Salesforce's vast new and existing AI and analytics capabilities together under a common brand. For example, some Commerce Cloud capabilities already existed in Demandware. So, while Einstein is partly about innovation, it's also about branding. Don't let the vastness of the Einstein brand fool you into thinking that all of this is new. That said, Einstein does signal new approaches that should also spread across more Salesforce products over time.

Integration? Many of these technologies are the results of acquisitions still being integrated into the Salesforce platform. Check to see how well these new capabilities are integrated into the platform, and also when integrated capabilities will be available. The big win comes when these capabilities are truly and natively built right into the platform. But for the time being, these integrations are still coming together, so Einstein may create some rough patches or snafus for marketers trying to gain intelligence about customers and prospects.

Predictive lead scoring for MQLs. predictive lead scoring should go beyond evaluating how well a prospect will respond to an email; it should be about how likely they are to make a purchase and become a marketing qualified lead (MQL). And only once a lead is determined to be an MQL should it be passed to sales. As of now, it appears that Salesforce offers this capability in Sales, not Marketing, Cloud. While several Salesforce AppExchange -- the online Salesforce app store -- partners offer predictive lead scoring, this capability should be built into Salesforce Marketing Cloud if it's not there already.

Account-based marketing? Account-based marketing, a marketing approach to target more appropriate prospects more efficiently, is gaining steam in many sales and marketing organizations. As such, it represents a major opportunity for Salesforce AI to use Einstein to connect people to the right accounts, identify trends within accounts, enable account-based ad retargeting and more.

Top-of-the-line AI? While Salesforce Einstein includes vast capabilities, no marketing platform has ever been able to deliver the deep capabilities of top analytics and AI platforms. That said, Salesforce may well deliver far more capabilities than other marketing clouds.


Salesforce Einstein promises to bring new analytics and AI capabilities to marketers and retailers using the Salesforce platform. Einstein's deep integration and always-on and advanced capabilities have the potential to vastly improve marketing results.

As Salesforce AI delivers more and more capabilities and gets closer to parity with the top vendors, marketers will need to determine if those vendors' capabilities are worth the integration, time-to-marketing-scale and data model challenges that they bring. Increasingly, the winner in that decision may be Salesforce Einstein.

Next Steps

Analysts speculate over Einstein features

Einstein brings AI to Salesforce

Microsoft AI is more than just Office 365

Quip ups its collaboration game at Dreamforce

Dig Deeper on Customer service and contact center

Content Management
Unified Communications
Data Management
Enterprise AI