By Adam T. Sutton, Senior Reporter
The marketing team at Firstmark Credit Union was often left wondering whether its emails had any impact. The team typically had to wait a month before reviewing campaigns results.
"If you are not getting good data, you cannot make good decisions, and we were walking around in the dark basically," says Fred Hagerman, CMO, Firstmark Credit Union. "Not only were we waiting a month [to compile reports], but it was taking about a week for us to produce those reports, and we were using our precious people resources in both the finance and marketing departments."
Firstmark operates 11 bank branches around San Antonio and it planned to significantly grow its email audience over the next year. But with its current database platform, doing so would likely be a waste of time. The marketing team needed fast and accurate reports before it could hope to convince new contacts to open accounts and take out loans.
Firstmark replaced its email database platform with one that updated daily and created flexible reports. The team expected a flood of data from the new system and had to ensure it helped revolutionize the bank's email program instead of drowning it.
Here are the steps the team took:
Step #1. Find a new platform
Like many financial companies, Firstmark used a customer database that pulled updates from a core transactional system. But Firstmark's database could only grab this information once a month. If a customer opened a new account in response to an email campaign, the team would not know for weeks.
"The database we had was a leader in banking platforms. So we hadn't bought something bad, but it hadn't evolved in terms of functionality over the years," Hagerman says.
Hagerman had prior experience with databases and helped create a list of features the team needed in a new platform and possible providers. Required features included:
- Rest on top of Firstmark's core banking system
- Show proven experience in the banking industry
- Gather transactional, demographic and behavioral data
- Create customizable reports
- Update daily
- Generate lists for sales and customer service teams
- Fit the company's budget
The new system also had to maintain the current functionality the team depended on, such as ability to append data, create lists and track changes. Are you ready to switch?
Regardless of how badly the team needed a new platform, it could not immediately launch one. Several factors had to be weighed beforehand:
- Contract with current provider
- Available IT resources
- Projects and priorities in other departments
The team spent the first three months gathering this information, organizing it, and picking the best window in which to launch a new database given all these variables. Get advice from contacts
Hagerman expected the new database to impact departs well beyond the marketing team. Realizing there were other stakeholders involved, he reached out to the company's CIO and other leaders to get their opinions on possible vendors.
The team queried some of the largest and best-known brands in the industry, and also reached out to the provider of the bank's core transactional system. After many demos and conference calls, the team thought the provider of its core transactional system offered the best solution at the best price. In all, the team spent six to seven months researching vendors and making its selection. Make sure everything works
Not wanting to take the vendor's word for it, the team scheduled more demonstrations to ensure the new platform met its needs. The team uncovered features the new platform lacked, particularly in reporting, Hagerman says. Reports that were standard in the older platform would have to be custom built in the new one.
"We recognized we might lose bits and pieces of functionality… You have to weigh that against the functionality that you are picking up, and you also have to identify processes that change."
Step #2. Convince executives
The team needed executives to provide resources and approval to make the switch. Hagerman was quick to point out that the new platform would cost roughly the same as the older one while providing superior functionality.
"From a financial perspective, no one had a problem with the decision," he says.
With costs out of the way, Hagerman explained how the new system would improve processes throughout Firstmark. For example, he pointed to tasks that would be cut from hours to minutes and also noted how quickly executives could create reports for themselves.
Hagerman recruited coworkers in the organization to help make his pitch. He also scheduled meetings with executives throughout the first year after launch to outline progress and results.
"Our leaders are sharp men and they expect me to oversell. So it was good to have enthusiasm from my peers in Branch and Lending [departments] who could sit down and say, 'This is not just Marketing sugar coating this thing. Here's everything we can do with it,' and there were some genuine enthusiasm for the opportunities."
Step #3. Prepare and educate coworkers
The new database would have far-reaching impact, so the team invited people throughout the organization to join the implementation meetings. These people typically had:
- Experience with similar platforms
- A leadership position in a department affected by the switch
"We added them to the implementation team so they could manage their pieces of it and get the requirements met the way that they needed it to work on their end," Hagerman says. "We were a small group of people who knew the opportunities for the various departments."
For example, the team involved assistant VPs from the branch management and contact center departments. Including these people:
Work with vendor
- Familiarized leaders with the platform so they could explain the changes to coworkers
- Showed leaders the low-hanging fruit to grab once the new platform was ready
- Smoothed the implementation process
- Ensured the platform benefitted everyone involved
Most service providers have experts to help you get more from your investment. Hagerman's team worked with the vendor to identify quick wins to grab once the new system was running. Hagerman and the vendor also showed coworkers how some routine processes would change. For example, some jobs previously managed through spreadsheets would be automated.
"Those that didn't get it right away, we walked them through an explanation showing them the demos, and the light bulbs started going off," he says.
Step #4. Anticipate changes, adjust when necessary
The marketing team would be the most fundamentally changed department by the new database. Hagerman foresaw the platform releasing a flood of metrics that could easily overwhelm the team. In response, the team outlined the following processes to ensure it got the most out of the data. Set measurement intervals
The team set a schedule for measuring the results of every email campaign. This helped ensure uniform reporting, which is vital for strong analysis. After a campaign launched, the team took measurements after:
The team started measurements at Day 3 since Firstmark's campaigns typically generated most of their results in that timeframe. Later measurements were helpful for tracking campaigns that promoted products with long buying cycles (such as car loans) and for measuring latent conversions that trickled in.
Other changes the team made:
Hire an analyst - After launch, the team realized that the benefits of hiring someone to concentrate on data far outweighed the costs. Firstmark hired an analyst to help improve measurement and analysis as well as related processes.
"He took to it like a bear to honey," Hagerman says. "Within a couple weeks, literally, we were starting to see benefits in terms of speed, ability to segment, ability to measure and analyze quickly, and our perception and conversations started to change."
Create testing guidelines - Hagerman planned to improve Firstmark's email program through rigorous testing. It documented the steps to design email tests, launch them and ensure their results were statistically valid. Hagerman launched pilot tests to explain the basics of email testing to anyone who lacked experience.
"Basically we made them give me an A/B subject line test. We sent it to 500 people, looked at the opens and the clicks, selected a winner, and then sent it to everyone else."
Balance tests and analysis - Once the team was comfortable testing email campaigns, tests started flying out the door and the team had to slow down, Hagerman says.
"I think within a month we realized we conducted about 45 different tests without looking up and saying 'hey, we should conduct 45 different tests'… Testing something without analyzing it is a waste of everybody's time."
The team shifted to testing more judiciously and set aside time for results analysis and discussion.
Step #5. Improve campaign analysis
The team soon added behavioral data to its analysis of email campaigns and found exciting surprises. For example, the team sent an email campaign
promoting its personal loans in Sept. 2011. The email was sent to all subscribers in San Antonio and included:
- Headline: Fast Cash For The Holidays
- Holiday-related messaging
- Image of a wrapped gift
- Example loan amount and monthly payment
- Orange "apply now" button
Previously, the team had to wait about 30 days to see a report on results. On the new system, metrics were available in 24 hours with response measured across:
- Three age groups: 21-30; 31-45; 46-60
- Loan status: No personal loan; Personal loan with <$1000 balance
The team found the conversion rate to be 852% higher among subscribers who had a current loan under $1,000 than subscribers who did not have a personal loan (conversion defined as a subscriber taking out a new loan).
Here are the results broken out:
- Conversion rate of subscribers without a personal loan: 0.29%
- Conversion rate of subscribers with a loan <$1000: 2.76%
- Highest converting group: 4.69%, current loan holders <$1000 ages 21-30
This valuable customer intelligence would not have been visible through the team's older platform.
Step #6. Find the best segments automatically
When the team sees great results among a particular audience segment, it sets the database to automatically generate that list on a schedule.
For example, the team found that customers with a balance of less than $1,000 on an unsecured term loan were very receptive to offers to take out another loan. Having proven this, the team now automatically generates this list each month and sends the customers an offer. This would not have been possible with the team's prior database.
"We want to send you the right email to buy that product when you are ready for it," Hagerman says.
While only the list-building portion is automated, Hagerman expects to automate the email-delivery potion in the future.
Step #7. Always test, learn and improve
The team takes steps to ensure it learns from email tests, rather than testing one campaign and forgetting about it. Here are two tactics that have helped:
Test brainstorm sessions - A big challenge in optimization is selecting tests that provide big results. The team combats this challenge by meeting once a month to discuss recent results, possible tests for the future, and which tests could yield the biggest gains.
Email testing library - The team prints a copy of every email test and its results. The emails are put into physical binders and organized by product-type. "This part is making sure that we're identifying the winner in whatever we are testing and making sure that it becomes a new baseline," Hagerman says.
"This year's budget has a major reallocation from the 50% of marketing that we didn't know wasn't working to the 50% that we do know is working," Hagerman says. "I know what I spent this money on and what it's doing for us. From a leadership perspective, it just makes it easier to sleep at night."
In the team's email marketing, average conversion rates increased 192% when comparing:
- 0.51% Avg. conversion rate in May 2010
- 1.49% Avg. conversion rate in Dec. 2011
The team's revenue per email delivered increased more than 205% when comparing:
- $56 Avg. RPE - May 2011 through Aug. 2011
- $171 Avg. RPE - Sept. 2011 through Dec. 2011
The new platform also helped the team improve results across the organization. After witnessing three years of declines, overall membership at the bank increased from 2010 to 2011 by more than 1,400 customers, which Hagerman says is partly due to the new platform's alerts system.
The platform even improved processes that were unrelated to email:
"When you come to a branch and ask for an auto loan, the teller is going to ask you to sit down and she'll have somebody come and get with you. On this new system, you will create electronic referral that would send it to the lending specialist who would then pop out of her office and say 'Hey, let's sit down and talk about your auto loan.' That used to be a spreadsheet type of process that got automated through this system."Fred Hagerman is speaking at MarketingSherpa Email Summit 2012 on Feb. 7-10 in Las Vegas.
Useful links related to this article:
1. Firstmark emailMarketingSherpa Email Summit 2012Complimentary Special Report: 4 Email Summit 2011 speakers - full transcripts Mobile Email Marketing: iPhone-targeted landing pages boost conversion rate 40% for Ritz-Carlton Destination Club Email Marketing: How a triggered alert program maintains 40% open rate, 60% click-to-open rate for millions of subscribers Email Marketing: Global telecom combines email and content strategy to segment database