This summer, staff at The National Aquarium’s Animal Care and Rescue Center bid a fond farewell to ‘Albert Einstein’, a grey seal they’d nursed back to health and successfully released into the wild, just four months after he’d been found stranded, sick and injured on the Delaware coast.
Elsewhere at the Baltimore-based non-profit, however, a completely different Einstein has been on the mind of director of business intelligence Michael Grandel. In this case, it’s the artificial intelligence (AI) capabilities of Salesforce.com’s Einstein offering that have been a big focus for his work, as he explained to attendees at the company’s Dreamforce conference this week in San Francisco.
The National Aquarium has been a Salesforce customer since 2013. In that time, it has used the company’s Nonprofit Success Pack (NPSP), along with Sales Cloud, Marketing Cloud, App Cloud and Community Cloud to build out the experience it offers to visitors, donors and volunteers.
Now, it’s the start of its journey with Einstein, Grandel explained, and basically looking to use the Einstein Prediction Builder tool to better understand what interventions its sales and marketing teams might make to maintain (and hopefully, boost) memberships and donations.
Two use cases - memberships and major gifts
With this in mind, he’s identified two key use cases where Einstein could help. The first focuses on using Einstein Prediction Builder to better understand how likely current members are to renew their memberships. These memberships are taken out by visitors in order to get free unlimited admission to the Aquarium for the year, and discounted prices on parking, tours and special events - as well as to support the organization’s mission to ‘inspire conservation of the world’s aquatic treasures.’.
As Grandel explains, data relating to visits, gifts, event participation and where members live can also be used to predict and score how likely they are to renew:
Why we’re trying to do this is that, once we get these groups segmented by likelihood, we want to be able to market to them differently. We’ll be able to get those details into Marketing Cloud and follow different marketing journeys, with different cadences of communications, and be able to leverage different dynamic content for different groups. We’re still kind of working through what we want to do exactly, but it’s about identifying audiences and then being able to try a bunch of different things and see which ones perform best. And the point of this is to help increase conversion rates and reduce the number of lapsed members.
The second use case looks at predicting the likelihood of a major gift from supporters, where ‘major’ is designed as over $25,000:
Because we have thousands of donors and members and ticket buyers, we have a lot of data in our system, so we’d like to be able to take a look and have Einstein Prediction Builder surface for us where new major donors might potentially be found among them.
Here, the idea is to use Einstein Prediction Builder to compare previous major donors - based on their gift history, membership history, any other involvement they’ve had with the National Aquarium and data from third-party fundraising intelligence tool iWave - with the general pool of current supporters, to spot who among them may be on course to eventually making a major gift. From there, it will be a matter of using smart marketing to cultivate closer, more personalized relationships with them.
AI for non-profits
According to Andrea Schiller, Senior Product Marketing Manager at Salesforce.org, the company’s research suggests that some 40% of non-profit marketers are looking to invest in AI in the next few years. In the case of non-profits, key use cases include identifying likely donors and supporters at risk of dropping out, as well as programme beneficiaries who are most vulnerable or considered the highest priorities for receiving help.
For Grandel, one of the key selling points for Einstein Prediction Builder is its ‘no-code’ approach. He’s not a programmer himself, he says, but while he has had access to a developer to assist in his explorations with the tool, he hasn’t needed that help. Building a prediction is simply a matter of determining the prediction you want to make and the records, fields and objects associated with it, in an easy, point-and-click process. Once predictions are made and marketing strategies formulated accordingly, he adds, the National Aquarium intends to use Einstein Analytics with Einstein Discovery to compare predictions against actual performance.
So far, the predictions aren’t informing marketing campaigns - it’s a little early for that, Grandel explains. The National Aquarium uses a consultancy firm in figuring out its marketing journeys and Grandel has only met with them in the past week or so to explain Einstein Prediction Builder to them and discuss how it might be incorporated into those journeys or help to build new ones. But a kick-off can be expected in 2020, he says:
And what we still need to figure out is the best responses to what predictions tell us. If you seem unlikely to renew, for example, should we send you more or fewer emails? What should these conversations look like? And if you visited in the last two weeks, should we jump on that right away with an email - or would it be best to give it some time? These are things for the Membership Team to figure out, but I’ll be engaging with them on figuring out how they can best respond to our findings.