BACKGROUND IMAGE: iSTOCK/GETTY IMAGES
Companies want their customer data automated and delivered in real time. In fact, they want more than real time: They want to stay ahead of issues and proactively alert customers about problems or offers.
But collecting this data on the minute -- and making sense of it before it becomes outdated -- is notoriously difficult. In fact, relatively few companies have succeeded.
Why is the prospect of using meaningful data from the Internet of Things (IoT) still so notional? The ability to get an object, such as a thermostat or refrigerator, to generate data is possible, but collecting volumes of this data challenges companies to derive meaning from that and act on it before the moment elapses.
The Salesforce IoT Cloud service -- which companies, including Emerson Climate Technologies, have been experimenting with since its release at Dreamforce 2015 -- addresses some of these long-standing issues.
Dylan Steele, senior director of product marketing, App Cloud and IoT Cloud at Salesforce, sat down to talk about the promise -- and the pitfalls -- of IoT data and how Salesforce hopes to tackle some of these issues in 2016.
SearchSalesforce: Why is IoT data difficult to make sense of?
Dylan Steele: You can think about IoT as having a unique set of challenges: Every customer has a separate set of data, separate set of data standards. It's not like CRM: Everyone has a basic customer format for their customer data. Every customer has a separate set of data standards. Because of that, we're working with a handful of customers through the pilot.
How are customers using the IoT Cloud today?
Steele: Emerson, [which provides heating and cooling technologies], is using IoT Cloud to sort through the massive volume of data generated by in-home thermostats. A customer could have a low battery or an issue being triggered by a thermostat. We can identify that in the IoT Cloud and create a customer case automatically in the Service Cloud, then send out a service rep, or send the customer a text message or an email.
You've mentioned maintenance as a key use case for the Salesforce IoT Cloud. That makes sense, but how could these services morph into new products and services?
Steele: Companies who traditionally were selling products, like a thermostat, they were just saying, 'Here's your thermostat.' But now, we are seeing a whole new world developing around being a product and services company. We're not just solving cases in the Service Cloud scenarios [i.e., customer service issues]; we're providing a service to you that is often powered by a digital experience.
Emerson is not just selling a thermostat, but also an automated digital experience and the mobile app that is connected to that thermostat can predictively support anything customers need.
At the same time, that creates new considerations. As you move from selling physical products to services, you now have a whole slew of data. Emerson not only has to think about how the thermostat data is performing, but also how the application inside the mobile phone is performing, how is the data performing every time you click a button or travel to various locations. It's not just about solving a service case, but about what are you doing with this massive volume of data to transform customer engagement and customer experience?
What applications can you ingest IoT data into, and how do you handle non-Salesforce data?
Steele: We're building out connectors to every cloud. We started with Marketing and Service Clouds. We'll build connectors to the App Cloud and the Sales Cloud in 2016.
Non-Salesforce data is going to be a fairly common use case. Some want to bring that data in-house. Bringing this non-Salesforce data into the IoT Cloud will take development and building connectors over time. But also, there is an open ecosystem. We allow not only ourselves, but third-party developers and customers into IoT Cloud, as well as out of IoT Cloud.
Do you mean you will create APIs or other kinds of connectors?
Steele: For now, we're building those connectors specifically through APIs. There may be other standards in the future to push data into the IoT Cloud. It's not only about the method, but also the format of the data.
Will it require more steps if it's non-Salesforce data?
Steele: Over time, I don't believe so. We'll have to prebuild connectors to Salesforce that will make it extremely turnkey. But today, if you're going to a third-party system that we haven't interacted with before, you may have to do custom development to get it set up in a way that's useful to you. Some customers may be able to do it themselves; some may require third-party consultants.
Dylan Steelesenior director of product marketing, App Cloud and IoT Cloud at Salesforce
This year, there has been a lot of discussion about the promise of data integration between clouds. How does this work with the Salesforce IoT Cloud?
Steele: The IoT Cloud will be the glue between all clouds, including Sales, Service and Marketing.
In my mind, a complete view of the customer is composed of two types of data: contextual data, which is more slow-moving. It reveals buying preferences, the last issue a customer had. But that is only half of the customer profile. The other half: How are customers engaging with devices in real time, what are they doing on the website, in your app?
We're blending two classes of data together: this real-time streaming data that's flowing off this world of Internet of customers, and blending that with context about what we know about them in the background. When those two kinds of data come together, companies can really engage them based on what they're doing in real time and give them a more relevant, rich experience.
What does combining these kinds of data look like?
Steele: Think about a connected car. Before, a car dealer might have known when you [were due to] come in for 10,000-mile checkup service. That's all it knew about you as a customer. Now, add all the real-time data: How am I driving, OnStar information. When you blend what you knew in the past with data from the last service check on a car's transmission with real-time data, the dealer can see data from out on the road and how a transmission is working, for example. Think about how that improves the customer experience, and how companies are getting a deeper, richer view of who these customers are and how they are interacting with products.
Another Salesforce executive at Dreamforce brought up connected cars as an example of how IoT devices haven't lived up to their promise yet. His wife has one, but when there was an issue, she had to call the dealer, rather than the dealer reaching out proactively.
Steele: You hit on one of the differentiators and one of the exciting things about IoT at Salesforce: We know about your customers, we are the system of engagement, and now we're bringing in that device data and bringing the two together.
We own not only the customer record, but we own a lot of systems of engagement with customers, rather than building IoT in a silo. If you're a car manufacturer and building IoT on one side of the fence, and then building a CRM in another part of business, are you really solving the problem? My suspicion is no. Can you really solve the customer engagement problem and really get to the heart of the issue you just described when you connect the systems together, and when you share the customer record behind the device?
How does data security work with the Salesforce IoT Cloud?
Steele: We are not collecting data from devices, nor are we aggregating data at the gateway. Once that data gets passed to the gateway, we set up a secure connection through APIs. But it's important to make the distinction on where Salesforce fits: We're not at the gateway, we're not at the device level; we're just dealing with the data once it passes through to the application.
What's on customers' wish list?
Steele: One item is continuing to build out connectors to our clouds beyond the Service and Marketing Clouds. We're also continuing to refine user experience. What's the UI experience inside the IoT Cloud? We work to make it business user-friendly, so you don't have to be a developer to set up the rules and logic behind the different pieces in IoT. As we get closer to [general availability, which is expected to be in the second half of 2016], we'll have not just slick, pretty but also a super-functional and makes it easy for them to set up rules and logic.
Microsoft pilots Salesforce IoT Cloud
Salesforce Health Cloud faces some skepticism in market
IoT Cloud adoption may be slow
Learn how to start an IoT CRM pilot