“Pics or it didn’t happen” has become a common challenge to outlandish, seemingly baseless claims. Although the phrase is more likely to appear on an Internet message board than in a boardroom, it still offers a valuable lesson to credit union marketers on the nature of skepticism and the importance of backing up efforts with meaningful, verifiable data.
At Firstmark Credit Union ($841.1M, San Antonio, TX), chief marketing officer Fred Hagerman circumvents roadblocks to his campaigns and departmental initiatives by following one golden rule: If you can’t prove it works, don’t do it.
In support of this philosophy, the credit union has invested more than a half-million dollars in tools and strategies to guarantee better results, particularly in the convoluted space of digital marketing. According to Peer-to-Peer analytics by Callahan & Associates, the investment is fueling results. As of first quarter 2014, Firstmark’s annual membership and loan growth is up 5.9% and 13.2%, respectively, and its average member relationship of $13,929 is roughly double the size of similar-sized credit unions between $500 million and $1 billion in assets.
Fred Hagerman, chief marketing officer, Firstmark Credit Union
Here, Hagerman talks about the credit union’s marketing investments, the impact of these advanced capabilities on every facet of the organization, and why big data is just as important for offline communications as online ones.
Tell me about your credit union’s history and what the organization looks like today.
Fred Hagerman: We are a state-chartered credit union founded in 1932 as San Antonio Teachers Credit Union. In 2003, we were granted a community charter for Bexar County and we also serve education-related employees in 11 surrounding counties along with their relatives. Firstmark currently has over 93,000 members and 14 branches.
How have your approaches to reaching members changed over time?
FH: We have a six-person marketing unit and our departmental budget is split about 50/50 between our digital marketing efforts and traditional channels like billboards, television, and radio. In either scenario, we prefer to spend everything we can in measureable channels first, then move backward to whatever is left.
We’ve come a long way on the digital side. When I joined the credit union in 2009, we had about 6,000 email addresses for 88,000 members. Today, thanks to efforts by our branches, we have over 70,000 emails for our more than 93,000 members.
CU QUICK FACTS
firstmark Credit Union
data as of 3.31.14
HQ: San Antonio, TX
12-MO SHARE GROWTH: 4.66%
12-MO LOAN GROWTH: 13.18%
How has an emphasis on ROI shaped your strategies and investments?
FH: Our investment in web analytics is a good example. We did some experiments with paid search testing early on but ultimately suspended them because of the cost. And although we could measure views and clickthroughs, we had no ability to measure conversation rates, including whether an application was started, finished, or funded.
We had used free web analytics tools for years, but it became clear we needed more powerful tools to understand our current and potential conversion activity, so we began using Omniture, which is now owned by Adobe. That investment increased our analytics expenditures by $50,000 a year. When you spend that kind of money, you need to be able to make things happen. Thankfully, the investment has paid off.
We changed our online application software and in April of this year began using CUNA Mutual’s LOANLINER, a product that works on online, mobile, and tablet. That’s important because our web traffic is almost 40% mobile now and email open rates on mobile devices is over 60%.
In May we moved an incremental $400,000 in funded loans through these channels; by June it was up to $700,000. We expect to see it increase further as home equity activity begins appearing in July.
Firstmark could potentially see $1 million in incremental online sales a month in the near future, so it now makes more sense to push email and paid search and we’ve ramped that activity back up.
What else have you been working on?
FH: We’ve implemented a new marketing automation system, and unlike the pure email service providers we’ve used in the past, this system uses a daily feed from our core to allow us to generate lists, track online and offline campaign performance, test emails and landing pages, deploy batch and triggered campaigns, and throttle up or down the volume of communications going to any specific members or population segments. Over time, we intend to completely load the system with campaigns, but that doesn’t mean we’ll be promoting everything to everyone everyday.
By using segmentation strategies that focus on combinations of behaviors rather than demographics — such as members with no checking account but an auto loan and savings account or members with a credit card but no auto loan — we’ve identified groups we want to reach out to as regularly as possible. A typical email response rate for members is 40 to 75 basis points, but our trigger campaigns have up to an 8% response rate.
We’re also moving to a SQL database system later this year so we can tie data from all our separate platforms in with data from our core. This will let us see where and how members are spending money, tie that into our CRM system, and build better predictive models and next-most-likely offerings. Synapsis updates nightly, so we can track performance on a daily basis.
What have you spent on this strategy so far?
FH: Our total annual spend on software and digital marketing channels is now close to $500,000 annually.
These resources are not cheap, but their value extends beyond the marketing department. For example, the ARCU system comes with over 300 standard reports for use throughout the credit union. This will replace numerous manual reports in different departments. We are also currently working to apply our web analytics software to the credit union’s intranet to better understand internal usage and we intend to investigate adding more operational emails to our marketing automation system.
There have been missteps along the way and we’ve sunsetted software that didn’t deliver as planned. But you can do that fairly easily as long as you don’t tie yourself up into long-term contracts with expensive vendors.
Do you have any tips for tracking ROI in social media?
FH: I keep an open mind because social media can serve several purposes. Social isn’t any good if you don’t have an audience, so we spent the first two years on Facebook building more than 12,000 likes, giving us the opportunity to reach a broad number of people locally.
We built much of or social audience using contests, which are inexpensive but still require you to measure what you’re doing and understand the costs. With our web analytics, we can track whether people went from Facebook to one of our landing pages and applied for a loan or took other positive action. ROI has all been positive but has varied significantly — from 50% to 722% — on different campaigns.
One of our most popular recent campaigns started when LeBron James was carried off the court due to leg cramps during game one of the NBA finals. A #LeBroning meme started almost immediately online, with photos featuring people being hoisted up in a similar pose by their friends.
We considered this a news-jacking opportunity and we had our CEO LeBroning the morning after the game.
Photo courtesy of Firstmark Credit Union
It sounds like goofing off, but that particular campaign had a reach of 14,000 impressions on Facebook and 71,000 on Twitter as well as numerous likes, comments, and shares across both channels. Fun can equate to distribution, and although it’s harder to measure, that activity spills over into generating goodwill and brand awareness.
What about in your offline channels?
FH: We have been able to use our historical advertising data to build some seasonality and attribution models that help us understand the effectiveness of traditional advertising options like outdoor, television, radio, and newspaper. It’s about determining whether these channels are effective as a whole as well as when they are most effective and if they are effective at the levels at which you can afford to spend.
For example, our models forecast future loan application volume by week, allowing us to identify application peaks and valleys and allocate marketing dollars accordingly. These models also suggest how much each advertising medium (online and offline) contributes to application volume. We use this information to time our advertising and allocate expense across all media.
We feel good about the money that we can measure and pretty good about the money that we can’t.
— As told to Aaron Pugh