Tactic 1: Pick A or B system, not C

Wells Fargo IT teams needed to work on every major system in the bank to integrate Wells and Wachovia operations. So was this the golden opportunity to improve those systems, to seek out the industry's best of breed?

No.

"We tried to keep it A or B. We didn't want a C," Mekjian says.

Wells Fargo was bigger in mortgages, so the integration team picked that system. Wachovia was bigger in brokerage, so they kept that one. The teams considered each system's capabilities, capacity, and the staff to support them.

Only two reasons were used to consider a new system: if neither existing system would scale to two or three times the current transaction volume, or if both systems were due for an upgrade. In one case, IT brought in a new voice-response system because the lines of business wanted new capabilities. But in all, fewer than a dozen systems were moved to a new platform. For the most part, any features gained in adding a system weren't deemed worth the risk of a customer disruption.

On occasion, the bank's IT organization would port over functionality from the losing system. For example, Wachovia had a "Way2Save" program in which the bank does automatic transfers to savings, such as moving $1 from checking to savings when a person bought something with a debit card. Some customers loved it, so Wells Fargo coded it into its card programs.

This same question comes up when a company's doing a major application consolidation project: Pick the best app we have, or pick the best that we could get? IT leaders need to be crystal clear on whether they're open to bringing in new software and if so what criteria they'll consider. Otherwise, they'll debate this same question hundreds of times.

Tactic 2: Decide how the business will run, then what tech will run it

Wells Fargo and Wachovia were in 80 different businesses. For each one, business and tech teams created what they called a "target operating model," from which they would decide the tech needed to execute the model.

It forced business units to make the big choices that would drive IT decisions early. IT needed those calls made quickly, because the tech integration needed to run months ahead of a business transition. Since each project built on the next, IT had to minimize last-minute strategy changes that added new tech work. "It wasn't on day one you had to pick A or B," says Chen--but it was within about the first 90 days.

In all, Wells Fargo eliminated about 1,000 applications from the 4,000 the combined companies had. For data centers, IT had its own target operating model that involved virtualizing and consolidating infrastructure, letting it go from seven Tier 4 data centers to three and from 13 regional data centers to 10. The combined company also went from 17 LAN environments to two. Virtualization has brought down IT's server provisioning time from months to 10 days.

Tactic 3: Be wary of emerging tech

In terms of emerging tech, the Wells Fargo-Wachovia integration is pretty boring. The integration team believed new features or systems that hadn't been tested in full production weren't deemed worth the risk amid so many moving pieces that needed to stay on schedule. "All technology has a burn-in period," tech integration chief Davis says.

But the integration group did make a few tough calls to embrace new technology. Among the biggest was how to handle checks given to tellers at the retail banks: Should it stick with the tried-and-true, "burned-in" manual processing of paper checks, or make the leap to capturing images of those checks, digitizing that age-old process on a huge scale? Digitizing is clearly the future. But was now the time to push it, when it had only been used in about 200 Wells Fargo branches?

The retail banking team (which Wells Fargo calls community banking) went for digital image capture. "It took some courage on the part of the community bank folks," says Jerry Enos, Wells Fargo's head of operations and a long-time Wachovia employee who had headed its operations and technology group. It meant installing new hardware at the banks, retraining tellers, modifying the process for check pickups, and changing IT systems to accept images.

Tactic 4: Build teams equally from Wells Fargo and Wachovia

Wells Fargo execs made a conscious decision to draw the integration pros as close to 50-50 from Wells Fargo and Wachovia as it could, says Scott Dillon, head of the bank's technology infrastructure services group. It set up those teams quickly--within 90 days after the acquisition. That decision brought advantages in terms of knowledge sharing and integrating the cultures. They now had the collective knowledge of two fraud-prevention teams, for example, so they could pool their tactics.

It also helped with retention of critical people, since they were now immersed in a vital project with a three-year timeline. That's important, since the bank made clear it would cut overlapping positions, including technology and other staffers who didn't directly interact with customers, in order to cut costs.

The long-term, feel-good benefits that came from building 50-50 teams came at a cost that's often overlooked, paid in lost short-term productivity. Blending people from the two companies broke lots of long-standing communications patterns and unwritten rules for how things get done, "which was really hard, because there were a lot of people who had a long tenure working the informal network," Enos says. "We've had to rebuild some of those networks"--which took six to 12 months, he says.

Companies doing any kind of reorganization need to recognize and build in this kind of cost.

Tactic 5: Use the muscles integration built

So it's done--Wells Fargo pulled off probably the largest bank integration of all time. The team successfully completed the work early this year, just a few months after the three-year anniversary of the acquisition. But chances are there will never be another merger approaching this size again, so is that a skill that's worth anything?

Yes, Davis argues. Banks are good at executing vertically--the mortgage team working within its process to create a new product and the technology needed to support it. What they're not very good at is executing horizontally--coordinating teams across product groups--and the integration honed that skill.

"That's a muscle we can flex," Davis says. "We can plow back into our DNA how we ran the integration."

Enos says the company built skills around creating target operating models--envisioning how they want something to run and backing technology out from that. One place IT and business units might use that is to create technology and process "utilities" that many of the bank's 80 business units can share--so each doesn't replicate the tech and staffing needed to support common tasks such as opening an account, processing payments, or taking deposits. Wells Fargo had done a bit of this utility approach with tasks like printing, but Enos thinks there might be more opportunity. Teams also will keep pushing digitization of processes as it did with check image capture.

A very public test of how well that muscle is developed is happening right now in the form of Wells Fargo's Project Compass.

Project Compass' goals include cutting the bank's quarterly noninterest expenses to $11 billion, from about $12.5 billion during the integration. Part of that will come because the bank's no longer spending $400 million to $500 million a quarter to run the integration, but a big part will come from reducing staff and back-office functions, including technology. For example, Wells Fargo has cut tech expenses 3% compared with early 2011 and cut by 11% non-customer-facing staff and contractors in "high-cost geographies," according to its latest 1Q earnings call.

"The key here is if we're more efficient on things that customers don't want to pay for, we can add salespeople, add more value to things that customers want to have, so we can compete for revenues and grow our franchise and do all the things that add value," CEO John Stumpf told equity analysts.

Davis has now morphed the integration organization into one that drives the Project Compass mission for the technology and operations organization.

Future projects likely will use some of the same muscles the Wells Fargo teams used during the massive integration, but the bank will need a different mindset. Innovation will have to take higher priority, whether that innovation is aimed at cutting costs or adding IT-powered features that directly touch customers. Asking "A or B?" was an efficient approach for integration. Now the questions get more complicated.

Chris Murphy is editor of InformationWeek. Reach him at [email protected] or on Twitter: @murph_cj.

InformationWeek: June 11, 2012 Issue

Download a free PDF of InformationWeek magazine
(registration required)