What is the #1 most successful office tool used by the top-ranked wealth management advisors. Click here to find out.
Onovative | Level Up With Core iQ

Measuring Marketing ROI: It’s Complicated (In a Digital World)

Liberty Bank for Savings found out how hard it can be to define the "customer journey" and analyze marketing attribution, especially when every customer seems to take their own, unique path.

Subscribe TodayA nagging problem in bank marketing is identifying the media that drives customers to open an account or purchase a product. Marketers are hard-pressed to put together cost-efficient media plans without being able to separate the productive from the wasteful. Most banks use surveys to assess effectiveness, but new, analytical tools suggest surveys can be alarmingly inaccurate and superficial.

Years ago, print and radio were the big advertising channels. Big black furnaces that gobbled up money, they created lots of steam and churned out vague results. Thus was born the tired adage, “Half of my advertising dollars are wasted, I just don’t know which half!”

Enter digital marketing with its attendant analytic tools.

Email, paid search and pay-per-click campaigns, each easily capable of filling spreadsheets with decimal-level accuracy. But all communication channels aren’t so easily digitized; “human nature” tends to throw in a big monkey wrench.

For instance: repeat customers, customer referrals, internal sales, signage and professional referrals. These are important referral sources, but the data collected isn’t always reliable. Consumers are notorious for crediting TV spots or billboards with actions even when the bank doesn’t advertise on those channels.

By combining traditional, digital and referral data to rank and quantify sources, the results are a less than solid foundation for building budgets.

Murky waters indeed…

MARQUIS | TriggerPro

Last year, Liberty Bank took a fresh look at the problem. The bank knew consumers weren’t pushed into action by a single channel — but what role did the other channels play? It was assumed that three or four channels might contribute to acquiring a lead or opening an account although only one source maybe officially credited. Much like the star and supporting cast of a movie, budget allocation based solely on credited sources overlooks the critical contributions of other channels.

Because of the analytical opportunities, we focused our examination exclusively on the digital channels which generated about 65% of all submitted applications.

First, we updated our Google Analytics configuration from “Classic” to “Universal,” which allows more robust multi-channel tracking. Then, we used Google Tag Manager which provides flexible, powerful tagging options. This setup permitted us to accurately track user interaction across multiple digital properties.

During the period from March 12 to April 13, 2015, we reviewed 108 mortgage applications using Google Analytics and Velocify for lead management.

The results were surprising.

Of the 108 applications reviewed, organic search traffic was named the most frequent source by 39% of applicants. Direct traffic — website visitors who type in the URL directly or use a bookmark — came in second at 31%, then referrals at 26%, paid search at 3% and display ads at 1%.

Scratching below the surface, however, we found a much different picture.

The 108 applicants — in one way or another — interacted with the bank 944 times. That’s an average of nine times prior to submitting an application! Some started with a telephone call, then talked with a branch manager and signed-up for an email rate advisory. Others checked rates, visited the mortgage page and made a pre-mortgage application. Some saw signage and plugged the bank name into a search engine, which, in turn, showed up as an organic search. Virtually no two applicants followed the same route to completing the process.

Reality Check: It’s hard to define the “customer journey” when each customer takes a different path.

Tabulating the touchpoints allowed us to determine total bank interactions and the relative importance of each media channel, thereby allowing us to create budgets based on performance.

The study also yielded important insights into user behavior leading to a series of website changes in information architecture, calls-to-action and messaging.

We’ve learned three insights from the study:

1.) Take customer or employee surveys with a grain of salt. Although well-intentioned, results are not reliable.

2.) The consumer purchase path is long, meandering and largely unpredictable.

3.) There’s lot of consumer activity beneath the surface which we must understand and support if we are to maximize results.

As Oscar winners invariably say at awards ceremonies, “I owe so much to my supporting cast. I couldn’t do it without you!” Indeed, that’s true in bank marketing, too.

Kevin TynanKevin Tynan is Senior Vice President for Marketing at Liberty Bank for Savings. You can connect with Kevin on Twitter, LinkedIn or send him and email.

All content © 2017 by The Financial Brand and may not be reproduced by any means without permission.

Digital Banking Report | Challenger Bank Battlefield


  1. Great study. Excellent that you took a data-driven approach.

    A few thoughts:
    – repeat the study on a periodic basis (quarterly, or twice yearly perhaps)
    – rerun the analysis as you have new data
    – see if you can overlay personas or segments, to determine WHO is likely to take which path.

    Also, while the sequence of steps may vary, it sounds like you could identify the key touchpoints. However, when a certain touchpoint was used, did it increase the likelihood of an application being completed? (Captures less frequent, by highly effective, marketing tactics)


  2. Great study. Kudos.

    I strongly suspect you found “Virtually no two applicants followed the same route to completing the process.” because your sample size was only 108. I don’t remember much of the permutation / combination theory I studied in high school over three decades ago but, I guess there will be hundreds of different pathways through the 8-10 touchpoints involved in the customer journey.

    I’m hazarding a guess that, if your sample were to be increased to a few 1000s of applicants, you’d find a few distinct customer journeys into which you can tag your 1000s of applicants, where “few” will be a couple of 100s, not more.

  3. The financial industry has recognised the value of additional revenue generation and brand awareness that lies in search engine advertising, as such the financial industry now tops the list of spend on pay per click marketing by industry.

    The value that is generated from pay per click is so significant that it is hard to overlook. The shift in audiences from TV to online has necessitated the advertising dollars to go where the audiences have gone. TV is referred to as ‘lean back media’ where the audience’s attention may be divided or limited, however online is a ‘lean in media’ which creates greater focus. Ultimately more advertisers would prefer the lean in audience in terms of attention and being able to count the number of people reached very precisely.

  4. So true. Glad you are doing some next-level research. Back in the day, how many times did consumers site the Yellow Pages as the reason they called or stopped-in. It was rarely (if ever) the Yellow Pages ad that drove the business. The phone book just happened to be where they found the address or phone number. Such is the case today with search. Was the consumer already looking for us when s/he typed in the search term? What was the trigger for them to search… a need or an ad?

Speak Your Mind


Show Comments