Taylor recognizes that it will take time to win the confidence of CIOs who are wary of teh company's intentions, but he's adamant that Salesforce has no plans to dilute MuleSoft's effectiveness as an all-purpose integration platform:
We need to make sure that both the product and the sales team for MuleSoft can still sell it to connect anything, [including] our competitors' products.
We're very serious about it, because I do think the power of these integration technologies is their neutrality, so we're really trying to make sure that we hold ourselves accountable to maintaining that.
For that reason, MuleSoft will remain an independent business unit within Salesforce under the leadership of its existing CEO Greg Schott, with its own sales, product, marketing and support teams. The intention is to continue to drive the strong growth MuleSoft achieved prior to acquisition — an important part of the justification for the $6.5 billion Salesforce paid to acquire the business.
Integration has become a strategic issue for customers, driving a need to rationalize increasingly heterogeneous IT infrastructures, Taylor explains:
Integration has gone from a CIO discussion to a CEO discussion. In particular, I think the trends driving it are the urgency around digital transformation ... and a mood among IT departments to move to the public cloud ...
So when we're coming in with a big vision for customer experience, we realized that we were leaving that [integration] to our partners or our customers to figure out on the other side. We really thought, you know what, if unlocking all this data from these back office systems and legacy systems is actually one of the key problems that our customers are solving, we should help them do that.
Changing how Salesforce connects its clouds
But there's huge value too in what MuleSoft brings to Salesforce's internal integration landscape. Therefore an important second strand of what happens now is that Salesforce has begun using MuleSoft's integration technology and expertise to improve how data and processes connect across its various clouds. While the Sales, Service and Community Clouds all run on a single infrastructure, the Marketing, Commerce and Analytics clouds each run separately.
Taylor cited two examples launched at the recent Salesforce Connections event — Service for Commerce, which lets service agents working in Service Cloud resolve customer queries by accessing real-time data from Commerce Cloud, and Commerce Journeys, which lets customer actions in Commerce Cloud automatically trigger follow-up responses in Marketing Cloud. There is now a dedicated team in his organization with a specific remit to build cross-cloud products, he says:
We've broken out a team within my organization called the Integration Cloud, that's really a team focused on cross-cloud solutions ...
The interesting thing about that team is, it's really being built at the platform level. So the plumbing that we did, so to speak, to wire those together is something that we can build on for future integrations.
Enabling cross-cloud capabilities — and selling cross-cloud solutions — has become a big focus for Salesforce, he says, because that's how customers are thinking:
Some of our most sophisticated customers, when they're coming to us now, more often than not, they're not coming just for one specific product. [They're] really looking at broadly how they can transform their customer experience, which almost always involves multiple customer touchpoints — or from our perspective, multiple clouds.
What MuleSoft brings to Salesforce is an API-centric, service-oriented model for managing integration. Taylor praised MuleSoft's concept of building an enterprise API infrastructure in the form of an 'application network', which encourages reuse within a consistent governance framework:
The cool thing about it, just like a computer network you set up in your company, [is that] once you set it up, it's available not just for that project, but every future project as well. So the integration work you did for Project A is available for Project B six months later. I really like that concept because it means that, rather than doing point-to-point integrations, when we help our customers integrate these legacy systems — make them available to mobile or make them available to cloud applications like Salesforce — the next project will be cheaper, easier and faster.
I really love that concept of improving the clock speed of future projects. So that's the one that really helps me unlock this data and then help our customers get better at future customer projects, which I think is really powerful.
Headless apps and citizen developers
Looking ahead, Taylor believes this API-based infrastructure will provide a platform that can enable 'headless' applications, whose functions can be made available wherever they're most useful, rather than having to load up the original application to get at them. As he explains:
[MuleSoft] often says, 'We make your whole company look like it's microservices, even if it's a mainframe.' They really start to transform old infrastructure to something that looks and feels modern and gives you that ability.
You're seeing this demand across a lot of our clouds. One of the big trends in commerce right now is headless commerce, or API-driven commerce — it's something we talk a lot about, in that specific segment ...
What I really hope we can do is, when we're partnering with our customers on these cross-cloud solutions is, rather than seeing a set of products, they see a set of services and capabilities that they can use to stitch together the customer experience. I really love the vision that MuleSoft has for this. I think it will influence our product direction, with that services-oriented mindset, which I think our customers are independently pushing us towards anyway.
He also believes the MuleSoft approach will lead to making those functions and data sources available in a low-code/no-code fashion. Then administrators and non-technical 'citizen' developers will be able to build them into new processes, or into conversational interfaces such as messaging tools and chatbots.
I think that so many business processes are done by our administrators and citizen developers, at customers. One of the things I think we get uniquely at Salesforce is the importance of that persona. That, [if] you make something only accessible to programmers, you really cut out a huge percentage of the market.
It is challenging. I don't think we're there yet. I'm very proud of where we are with our bot framework, but it is definitely early stages in that, but it is a huge area of investment for us.
Cross-cloud, we're really co-investing across our Marketing and Service Clouds, recognizing that those multichannel bot interactions actually span multiple clouds. We're wanting to make sure that we actually share resources and share our development there.
These integrations with these back-office systems — and making them accessible to administrators and citizen developers — is absolutely a key part of our product strategy.
Much of the discussion of the MuleSoft acquisition has focused on how it increases Salesforce's ability to help customers integrate other systems into Salesforce. That's also been the main focus of Salesforce's messaging, which has emphasized how MuleSoft helps it deliver a 360-degree view of the customer experience by connecting into other systems that touch the customer.
But there's clearly an important story too around how MuleSoft will provide a new framework to enable integration within Salesforce's own platform, between its different clouds. As Taylor makes clear, customers are increasingly asking for data flows and end-to-end processes that cut across the traditional product demarcations. This is a challenge that all the major enterprise application vendors are having to face up to as business becomes more digitally connected and demands more frictionless processes.
There is much still to be explained about the detail of how all this will be achieved and what it means for existing integration patterns within the Salesforce platform, such as Lightning Connect and whether it means any refactoring of the current set of APIs. But from Taylor's comments this week it sounds as though the MuleSoft notion of an application network is going to play a central role in that new integration landscape.