At RBC Wealth Management, the dashboard item its advisors use the most is the one that shows all their clients' birthdays for the current month. That tells you a lot about their relationship with data — they want to see, at-a-glance, the data that helps them get on with their job of building client engagement. With that in mind, all of the dashboards and analytics they use have been built to help them get things done. As Greg Beltzer, Head of Technology at RBC Wealth Management, says:
Analytics has to be actionable. If they can't click on something, and take action, I don't really want it on the screen. I'm not interested in a bunch of pretty charts and graphs. I want them to be able to put them into a Marketing Cloud campaign. I want them to be able to do a follow-up task to their staff, like, ‘Make sure to get this distribution done’ when they're looking. And it needs to be within, I always say one click, obviously sometimes it's a few clicks, but that's the goal.
This is not just a matter of satisfying a demanding user base. It also makes economic sense. Beltzer adds:
Think about financial advisor salaries. That's our biggest expense, the salaries. These are highly compensated individuals. If I can give them back a couple of hours of the time, that's a meaningful ROI, on a weekly basis.
'I can't find anything'
But this user base is hard to satisfy, impatient, resistant to change, and, when it comes down to it, their personal relationship with the client is more meaningful than any data. To show value, the data must enhance what they do without imposing a burden. Beltzer says:
One of the common complaints we get is, 'Okay, great, Greg. You built these great boards, you have all this data — I can't find anything. I don't know where it is. I know it's there, maybe on the financial account page, or maybe it's on the client page ...' And obviously, real estate is precious.
To that end, Beltzer's team has created a set of dashboards that bring together information from the underlying Salesforce Financial Services Cloud about client accounts, along with external services such as FactSet and Refinitiv to deliver real-time stock price information, news and other data. For example, there's a dashboard that summarizes information for each investor household. He explains:
That presents the most common data attributes an advisor would get if a client were to call in unexpectedly. Like, 'What's my balance? How much do I need to take in retirement? How are my goals doing?' Those type of comments that usually then [you] have to tap back and forth, we put it right here. They're all real time calls [to the data], so they're getting up-to-date information.
It's taken several years of iterative development to reach this stage. The bank first rolled out CRM Analytics (then known as Einstein Analytics) alongside Financial Services Cloud in the fall of 2018, having already implemented MuleSoft as its integration hub, and completed the implementation by the end of 2019, just ahead of the pandemic. This replaced a home-grown CRM system with limited integration to the various data sources at the bank. Beltzer classes the initial roll-out as a "heavy MVP," in that it already included a lot of features, but user feedback and evolution of the underlying platform has since led to further changes. There's a regular update cadence of 25 releases a year. He says:
If I were to show a screenshot from December of 2019, I guess it'd be recognizable, but you can definitely see the evolution of the platform — which is one of the reasons why we have Salesforce as our platform.
Never miss a birthday
Much of the work has been devoted to building advisor confidence in the system — demonstrating the accuracy and timeliness of the data, at the same time as providing insights that help them be more efficient. This has meant focusing on what matters to them, which is not the transactional element of the job, but the relationship element. Beltzer explains:
If you ask them the number one thing that they are most afraid of, it's not actually losing a client, it's disappointing the client. It's missing a birthday. It's missing a withdrawal. It's those day-to-day items that they really want to make sure that they don't want let down that client ...
So we've really focused on the service aspect. This is not sales, this is the stuff that you most likely aren't thinking about on a daily basis, that hopefully you can spark a more meaningful conversation with the advisor.
Combining external datasets with the Salesforce account information has been important for putting data in context to help advisors answer client queries faster. He says:
Your client calls in and [says], 'What's going on with Apple?' Instead of you having to swivel-chair, it's literally, click on the account, and not only can you talk about Apple, you can talk about [for example] CRM ticker symbol, you can see exactly what's going on contextually within the application.
Over time, AI-powered capabilities have been added, for example to identify suggested actions, such as adding missing beneficiaries to an IRA (Individual Retirement Account), or to analyze wallet share. A client attrition model has been useful in helping build confidence in using AI, since it's been designed so that advisors can drill down into the factors driving variations in the score, and so understand why the system has made the change.
Doctor vs surgeon
A crucial element in getting the design right was the creation of a Center of Excellence (COE) for the analytics team, where developers sit next to business analysts (BAs), rather than the old model of sending requirements to IT with no further interaction. This not only creates immediacy, it also builds a team culture. Beltzer explains:
The developer can literally lean over and say, 'Is this what you wanted?' And the BA can go, 'No.' They can collaborate then, versus having a meeting, or versus the developer spending time writing code for something that's not meeting the business needs.
You also build a culture there as far as, you tend to want to hang out with people that you work with all day, if you're doing cool things. The pace we were going at was very different for the organization, there was an energy around that pace, and bonds were formed that serve us today and allow us to iterate and keep pushing. And then you just grow the team.
Today, Salesforce admins play a key role too, especially when it comes to user experience. He explains:
What a data scientist and a developer don't have a good feel on, is that user experience. Where is the best fit for that data? They can come up with the best algorithm in the world. But if I can't actually surface that in context, who cares? That's where I use those admins, to be the bridge on where it shows up in the application — where it shows up in that sidebar, what that front page experience looks like.
The ongoing evolution of the platform has allowed for more declarative, point-and-click development, which admins can take on. That leads to an interesting dynamic in weekly meetings with the COE, as he explains:
It's kind of like when you go to your doctor versus a surgeon. A surgeon wants to cut. The doctor says, 'Well, maybe make some lifestyle changes, you know, take these pills.' That's a developer versus an admin.
To see that dichotomy between the two leads to some really interesting questions — because there's never just one way to do something in Salesforce, not with the toolset that they provide to you. So that way, you get to the best solution. It could be a mixture, it could just be all code, it could be all declarative. But you want to take that user experience into into that mix.
A monthly advisory board gives advisors an opportunity to see proposed changes and new features before they're rolled out in the field. Beltzer says it's important for developers to be able to have these conversations:
I expect my developers to be business-facing. You need to be able to be in a room, you need to be able to understand the business enough to have the conversation. You don't have to lead it. We'll have a BA or an admin there with you, but you need to be able to understand what that end user is experiencing.
From batch to real-time
The revenue potential of freeing up advisor time and helping them target their actions is probably the most significant ROI impact of the analytics rollout, but Beltzer says the largest cost savings have come from being able to reuse MuleSoft APIs:
In that onboarding, we built a ton of APIs. When you onboard an account within financial services, I've got to have a power of attorney API, I've got to have a trusted contact API, an address change API ... A lot of these services are really into the reuse phase. I built them for the onboarding application, and now I'm going go ahead and build it into a flow within Salesforce. Same service, just exposing it to a different endpoint.
But moving data in bulk isn't the best solution going forward, so Beltzer is now evaluating Snowflake and Salesforce Data Cloud as an alternative mechanism for querying large datasets. As a regulated industry, financial services has been slower than others to read and analyze data on the fly in the cloud rather than moving it into separate silos. But customers are no longer happy to wait for overnight batch processes to run, so Beltzer believes it's only a matter of time and putting the right guiderails in place. He comments:
I still think financial services is in a batch culture, just because of the settlement period around securities. But our clients aren't. Our clients are used to real time in everything they do. I could probably get away with the advisors, because they expect batch. But if someone gives you a beneficiary, I want to be able to go to my online portal, and see that it's updated. I want to be able to see that it's done. I don't want to have to worry about it and check on it the next day.
Beltzer was speaking at last week's TrailblazerDX conference and met diginomica for an interview after his session.