Marketo: One Platform for Engagement

"One platform for every engagement, through any channel."

That was the promise made by Steve Lucas, well into his second year as Marketo CEO, and putting the emphasis — as you might expect from his track record with SAP and Salesforce — on product.

There was no shortage of product announcements at Marketing Nation 2018, Marketo's annual summit, although they were perhaps overshadowed by acquisition and partnership news: The acquisition, in particular, of marketing performance platform Bizible, announced yesterday; the acquisition of Marketo integration specialists Tibco; the announcement of a strategic partnership with SAP, delivering native integration of Marketo into SAP CRM and Hybris Cloud for Customer; and perhaps dominating the way, the news that Marketo will run exclusively on the Google Cloud platform, with the ability to leverage the Tensor Flow AI software library.

That's quite a lot to digest, and I haven't even mentioned the new UX experience, Marketo Sky, or the long-awaited re-emergence of year-old acquisition ToutApp as Marketo Sales Engage.

But first the overview from Steve Lucas, who — with all his zany digressions — is one of the more entertaining keynote speakers among tech CEOs.

"Something entirely new"

One of Lucas's most striking remarks referred to the transition from CRM to "something entirely new." Namely, a system of record for engagement. That's part and parcel of Marketo's commitment to what Lucas described at last year's summit as "the Engagement Economy." (Echoes of Adobe's "experience system of record" simply show, I believe, where the space is doggedly headed). 

The problem with CRM, which I now hear about almost every week, is that it doesn't record what people actually do, but what they say they do. Or, to be harsh, what the CRM user — however belatedly and incompletely — says they do. It's hard to see, Salesforce notwithstanding, how CRM data can remain central to a platform which seeks to engage with customers immediately, and at any point on their journey.

"Our brands are not controlled by us any more," said Lucas. "The brand is now being defined by the buyer." Marketo's response? "We are curating the buyer experience; curating a journey." And doing so right across what Lucas calls "the engagement grid" — social, mobile, and digital; a space now home to some half of the world's population.

This doesn't seem an idle boast, given the numbers Lucas had on hand. In one year: 30 billion delivered ("delivered, not sent") emails, 21 billion clicks on links, and a boggling 400 billion actions by users of the Marketo platform.

A platform working at that scale needs reliable, fast, secure, and available cloud service, which is one reason the Engagement Platform is settling on Google Cloud. The other reason is to make almost half a trillion transactions smarter. "AI is one of the biggest areas of investment for Marketo," said Lucas. In addition to Marketo Content AI, which recommends relevant content at an individual level, Marketo yesterday announced Audience AI, to automate the segmentation of audiences at scale. Enabled by Google? I asked Marketo CTO Manoj Goyal for enlightenment.

"Audience AI is our first proof-of-concept"

"We are going to be running all of Marketo's infrastructure on Google Cloud. So we are going to be using Google hardware to run our software. Most of our customers will start migrating to Google Cloud this year." Palo Alto Networks, the cybersecurity giant, is the first Marketo customer to have made the leap. "How do we combine our data with Google's Tensor Flow machine learning library so that we can create models?" The first advantage is that the modeling has already been done; the second is that it's available at high velocity for execution; and thirdly, there's access to the Google data science team.

"Audience AI is our first, joint proof-of-concept. Now that we can do that, let's look at all the engagement points and look at what works and what doesn't." Users will be in control, able to choose between their own strategies and the AI recommendations, or "blend them together," but users won't need to be involved with developing or tweaking the models. AI will work behind the scenes, in a black box way, but will offer recommendations along with analysis, so users can transparently understand the reasons for recommendations.