News Stay informed about the latest enterprise technology news and product updates.

Morgan Stanley: Analytics drives CRM success

CARY, N.C. -- Do you think the "big bang" approach to CRM implementation is ready for a comeback? You'd have been better off putting Gigli in your Oscar pool. Tony LoFrumento, executive director of business intelligence and CRM with financial services giant Morgan Stanley, weighs in on the big bang bust, why some CRM projects crash and burn while others flourish, and how user-friendly analytics software could unleash greater CRM success for his company. interviewed LoFrumento last week at the launch of SAS 9, where he was a featured speaker.

Are there any innovations in CRM you're watching closely? Anything that makes it easier to bring all the information...

together to create a holistic view. We have created a powerful CRM data mart, which brings information about our clients from about 35 different systems into once place. Companies have to be able to view their clients holistically. You also need the right IT professionals to manage the system; the right level of statisticians and analysts and the right application software. Then you have to be able to get the information out to people quickly. It's all about getting information to people who can use it effectively.

Tony LoFrumento has spent 15 years in retail banking. Before joining Morgan Stanley in 2001, he worked at Chemical Bank, Chase Manhattan Bank and CitiBank, holding various positions in CRM, finance, marketing, strategic and business planning, and online operations.


Featured Topic: CRM case studies

Best Web Links: CRM for financial services

Who champions CRM in your organization? Are you the chief champ?
We've been fortunate in my situation, because when new management came in early 2000, their goal was to change the way we did business. They realized they needed business intelligence to run the business. I didn't have to go in and sell the idea, I was brought in [in 2001] to do it. They didn't tell me how, but they knew they needed it. The key is delivering results -- giving senior managers what they need, then they become the champions. They know they can't live without this, and they can't go back. When you evolve into a business based on knowledge and fact, you cannot go back to what in essence are educated guesses. Do you have any plans to extend analytics capabilities to non-statisticians?
We have business intelligence analysts who are trained in both SAS and Business Objects, but it's definitely not gone down much below that. That is something I am interested in looking at because I don't want to create a bottleneck, which could happen if you're one organization and get 30 analytic requests in one morning; my team can't work on all 30 requests. If each group had an analyst who could handle the queries, then that would be the best because then they don't wait. What I heard about SAS increasing the ease of use [with the release of SAS 9] was interesting -- I interpret that as meaning they're going to get [analytics] down the chain a little farther. There aren't a whole lot of statisticians in most companies, much less departments. If a person can write me an e-mail and tell me what they want, they should be able to put that into a front end and get an answer. That would unleash the acceleration of organizational learning and allow my organization, which has all these highly trained people, to focus on the more advanced stuff that only they could tackle. The day-to-day BI requests that are the meat and potatoes of the business -- if we could move that down to [non-statisticians], the organization gets a lot more throughput.

A query begets another query. When you have to keep e-mailing your queries back my team, you lose your continuity and train of thought because we might not get back to you for a couple of days. If you can do the queries and drill down, then you go through the discovery phase and become more of an expert in your area. That's why I am excited at any vendor that can increase the ease-of-use of these analytics tools.

Did CRM come in under or over budget at Morgan Stanley?
We have a very competent IT team that's very independent. We really didn't contract with either of our vendors for help with [the] implementation. We actually came in under budget. It's much more controllable when you're on the analytics side. I have a 14-person team that handles analytics for the retail side of the organization, so initially I didn't have to train hundreds of people -- when you do a sales force automation, you can have 15,000 people you've got to train. The potential for overage is far greater than my team saying, 'Let's load the software and have the statisticians start mining the data.' Once you put it in, that's it -- now it's just the fixed cost time of your people leveraging and utilizing the information. What about the incremental approach to CRM versus the 'big bang' approach?
There is no big bang anymore -- it's impossible. No one will let you get away with it -- the failure is too much of a risk. I think the key is to get a prototype up and show results along the way. I'm currently implementing the client profitability/lifetime-value capability at Morgan Stanley. If I had said, 'I'm going to build the whole thing in a few years, and when it's done I'll show it to you,' senior management would not have been receptive -- versus delivering in pieces, starting with the revenue portion. My first phase was to show every dollar of revenue that comes into our side of the organization at the account/client level. Management is seeing results and progress -- people love to see progress. With the revenue side done, we are currently focused on the expense side. When the expense side is completed, we can subtract one from the other, and then we have calculated current profitability. Then our chief statistician will model based on demographics and other input to derive lifetime value for each of our clients. So they've seen deliveries over various phases of the project -- if you get a prototype up, it's very effective. Showing live data is powerful in getting management to buy in. What about CRM failure, what do you think is to blame?
Some of the more well-publicized failures of CRM were in the contact management space. I don't believe that had anything to do with the software -- it may be the sequencing of how you implement the various components of CRM, and obviously I'm a fan of implementing analytics first. My personal opinion is, if you build a solid contact management system but don't have the analytic CRM house in order, you've built a very expensive Rolodex, and you're not going to make back your $30 million. I think back to the mid-'90s when it was called 'sales force automation' -- these companies threw these things out there without getting the consent and buy-in of the sales force, plus they didn't have the information analytically to drive the interaction with the client. I think if you get the analytics house in order, it just increases the chance for success with the other side.

I also think that a lot of failures were probably caused by not getting the input of the people who were going to use the system. A lot of salespeople have a natural tendency to not want to divulge all the information they know about a client -- they think maybe they won't be with a company forever, so they want to keep the names in their own books. The way you [give them incentive to use the system] is to say, 'If you put information in, our analytics will give you more insight into your client base so you can get more sales.' Then they think there's something in it for them. I can't believe any analytics program really was a failure. If you can pool and analyze your data, any insight you find will add huge value to the company.

Who are your main vendors?
Our two primary vendors are SAS for all the data mining, business intelligence analytics and campaign management, and Business Objects for Web-based reporting. We have more than 450 branches and about 15,000 people in the field. We have a Web-based solution that allows us to get powerful information into the hands of people who can effect change. A lot of the information comes from the modeling and mining we do on the SAS side. We have the best of both worlds. We feel we're fully covered when it comes to discovering knowledge and distributing it to the people who can make a difference. We also utilize sophisticated house-holding algorithms from a company called DataMentors -- they look at data and make sure we're finding all the accounts related to each household to ensure a holistic view. We also use Informatica for ETL process. We have not yet gone with a full-blown contact management solution -- we're focused on the analytics right now. How does the ROI process for CRM projects work at Morgan Stanley?
Operational CRM is much more expensive when you do a multi-thousand seat implementation, and obviously you're going to need detailed ROI analysis to justify that type of expense. When I came to Morgan Stanley, I had a clean slate and said, 'Let's start on the CRM analytics side. It's much less expensive, and we'll derive incredible value by having all this business intelligence about the company we didn't have before.' How would you value transforming a company to one that's run on knowledge and fact versus gut feeling and intuition? I'd say you could value that as a good percentage of the franchise. I view it as our job to make management and senior management decision-making 'boring.' I want to be able to put all the facts on a table when it's time to make a business decision, so it would probably take you a half hour to say, 'Well, it's obvious that Option A is preferable to Option B.' If you can make decisions with authority and confidence because you're backed up with facts, then that's a huge value benefit to the organization.

You have to know exactly whom to roll out the red carpet to. The 80/20 rule is usually a little conservative. You'll often find that 10% of customers deliver a majority of the value. Companies had better understand who they are and do everything they can to retain them.

Dig Deeper on Customer analytics

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.