XPRT Talk: Are banks going to market with the best marketing campaigns?

In this XPRT talk, Gaurav Kapoor (Newristics) leads a conversation with Maks Sobkin (TerraStrat) about the challenges of optimizing messaging in the banking industry. Marketing, especially messaging, has long been an exercise in creativity, but now the scientific and algorithmic optimization of marketing content is necessary to win in the market.

Read this fascinating XPRT Talk to learn how the best banks are optimizing their messaging and winning in the market.

Meet our expert panelists

Maks Sobkin is a leader in using data science to solve tough optimization problems in the banking industry - how to optimize marketing programs, how to optimize bank locations, etc.

Gaurav Kapoor is an expert in applying behavioral science to optimize brand messaging and to scientifically improve the effectiveness of messaging campaigns.


Maks Sobkin

Chief Technology Officer | TerraStrat

  • XPRT Talk

    Over 15 years of experience in Data & Analytics in Financial Services.

  • XPRT Talk

    Leader in data science and machine learning, with focus on customer segmentation models.

  • XPRT Talk

    Ran and deployed branch distribution strategies and network optimization for multiple institutions, including national and regional banks.

  • XPRT Talk

    Published expert within the GeoSpatial community – recognized industry-wide as GIS power-user.

  • XPRT Talk

    Multiple-award winner and recipient of the ESRI Special Achievement in GIS (SAG) Award.

  • XPRT Talk

    Successful entrepreneur, with a proven track-record and keen understanding of small business dynamics.


Gaurav Kapoor

President | Newristics

  • XPRT Talk

    25+ year career in marketing consulting, working with the world's largest brands across industries.

  • XPRT Talk

    Expert in applying behavioral science to optimize brand messaging.

  • XPRT Talk

    Pioneered new message algorithms and research technologies for messaging development and testing.

  • XPRT Talk

    Respected thought leader in applied behavioral science for customer insights and marketing.

  • XPRT Talk

    SME advisor to marketers at leading finserv companies like Capital One, Bank of America, Discover, Fidelity, and more.

XPRT Talk: Are banks optimizing their marketing campaigns with the best messaging science and algorithms available to them?

The fintech revolution is forcing banks to transform their business model and become more automated and human-centric at the same time. Bank marketing is also transforming rapidly and banks have to work harder than ever before to acquire and retain customers through their marketing efforts.

In this fascinating discussion, Maks Sobkin and Gaurav Kapoor address some of the biggest marketing issues facing banks today.

What are the biggest marketing challenges facing banks and credit unions these days and is messaging important to their marketing success?

Marketing is all around us. Personalized messaging has become more relevant than ever to break through and deliver incremental marketing results. Walking the fine line of optimal offers at optimal times without seeming invasive is key. We can also deep dive into additional challenges most banks have to consider:

  • Targeting the right customer/prospect
  • Balancing the right channel mix
  • Customer data (correct, timely, holistic)
  • Deployment and measurement of predictive model performance that feed marketing campaigns
  • Measuring incremental Return-on-Investment (ROI) correctly

Thinking about every aspect of customer marketing, where is messaging most important - acquisition, retention, loyalty, experience, etc. ?

While the entire customer life cycle needs to have a cohesive marketing strategy, most marketing efforts and dollars of the banks I've worked with are spent in acquisition. Next would be cross-sell. This is a product sales model.

  • Acquisition is typically most expensive, so from an investment standpoint you want to get that right first. It’s easier/cheaper to direct campaign to existing customer than capture new ones.
  • Retention marketing is hard, because once the customer decides to leave it’s hard to change that behavior. And a lot is outside of the control of the marketing department.
  • Loyalty, through the lens of LTV and survey analysis is complex and is not as reliable of a metric.
  • Experience always needs to be positive, relevant, and ubiquitous. The incremental Return-on-Marketing-Investment (ROMI) is also hard to measure.

If we define messaging optimization as the right message, right customer, right channel, right time, how well are banks optimizing messaging in your opinion?

Most banks – not great. The top guys, who have done this longest, have had access to amazing talent and large marketing budgets to do it best. It takes time to set, test and prove out. Even at the banks, you see internal competition between managers who have their own product and annual goals running campaigns that compete with other managers’ annual goals. Especially across products that are different lines of business. So, prospects and customers can get hit by many different messages that aren’t necessarily coordinated nor optimized to their needs and their time, relative to the bank’s needs and timing.

Right time is probably still the hardest for banks to get other words getting in front of those life changing moments, “moments of truth” to either sell a product or service or communicate your message effectively. The other factors banks should be able to account for since they have the necessary transactional, financial and demographic data on their customers for targeting.

Banks can segment and build very accurate propensity models. If their customer data is being captured and properly leveraged in a modeling sense the Marketing department at the bank should be very well informed about aligning the right message to the right customer through the optimal channel. The first 4 should be very well informed. *'Please note that the "right message" assumes a Next Best Offer/Action engine has been stood up. Most banks are not there yet and run single campaigns, based on either segmentation or propensity models without considering their role as part of the entire customer life journey.

Also we often see campaigns not having a direct call for action. Or calls to actions aren’t clear or hard to follow. These campaigns often fail. Rule of thumb –if you ask for nothing, or offer nothing, you get nothing.


A lot of messaging is going through digital channels and A/B testing is a popular way to test messages cheaply in digital. But, data shows that a staggering 80% of A/B tests are terminated before they reach a conclusion. So, how are banks finding the best messaging for their campaigns?

Banks just need to keep developing new challengers. Most do not think through implementation and gathering results. I see many teams fail to prove out results because they did not initially stand up the right foundation to track and measure results in a meaningful way to draw conclusions and take action.


You have done a lot of modeling work on marketing campaigns. What have you found the value of Good vs. Great messaging campaigns? Does optimization make a meaningful impact in messaging for banks?

Modeling drives the optimization of eligibility for the campaign in the first place. Then test designs further cut by offer, channels, creative, and frequency. Each is ideally optimized for the customer, and each adds marketing (incremental) lift. Messaging is typically one of the things that's "tuned". That's because the marketers spend a lot of time with creative. Whether or not models and test designs are created is the bigger question.

There are other levers that make a bigger impact on disposition within a campaign like aligning right offer with right person. Anecdotally speaking, when a bank has adapted a “voice” across all its messaging, it seems to resonate more with the market. For instance, ING Direct’s “transparent, simple and quirky” messaging was adopted by Capital One to target a specific demographic before influencing the tone for entire bank.


All messaging campaigns start with ideas and end with execution in the market. Lots of people are involved in the whole process, which means lots of opportunities for things to not go as planned. What are the biggest pain points you have seen in bringing successful campaigns to market?

At the bigger banks, the coordination across all the teams and maintaining timeliness is where there are a lot of opportunities for failure. Another is managers continuing to run what THEY think works vs. using analytics at the bank to truly drive continuous test and learn – not to mention test varied offers/challenger creative and messaging/frequency/channels/etc., that might not be as successful as champions, and to hold out control groups instead of maximizing returns by only marketing to treatment populations. At the smaller banks, I see a lot of one-and-dones. Either the campaign did not perform as anticipated, and efforts are rerouted elsewhere, or results could not be clearly interpreted.

Also banks often leverage third party data and vendors to run such campaigns for a first time. In these cases, I see poor documentation, little incentive internally to coordinate and oversights in data quality and its applications. Also, a surprising number of banks don’t have a rigorous regular budget review where business cases compete against each other for marketing dollars, using analyzed past results and new campaign forecasts.


One school of thought in marketing campaigns is to get it right first time. Another school of thought is to just get a campaign out and tweak it on the fly. Where have you seen more success?

Tweaking on the fly implies the campaign has the response data coming in across all channels very frequently (daily or intraday) and is always being analyzed. Then it assumes each channel can be tweaked midflight. And that there is some overall coordination and decision making to redeploy campaign dollars. That just isn’t the case except for the biggest banks.

What is more practical is to use past analytics, rigorous planning, a solid business case process, and accurate forecasting to deploy the best multichannel campaign in the market and have a rigorous test and learn culture that will make the next campaign better, and the next one better after that. A surprising number of institutions are not close to being there. Another set of institutions tends to sit on the data and over-analyze things until it’s too late. Either competitors have taken action (although sub optimal) to increase their share of wallet or data becomes stale and unusable.

If it was up to me, I’d go with deploying a campaign that was 80% there in a timely manner in a test market, applying learnings and adjusting going forward. In other words, don't let perfect be the enemy of good, but make sure your good is really good first. Hopefully in this case, the campaign also goes through a vetting process carefully through all appropriate channels and business owners to avoid any potential issues.

About Newristics

Newristics is famous for helping brands optimize messaging using a combination of behavioral science and machine learning algorithms. Newristics is the only vendor who has mastered the advanced use of human Decision Heuristics + human-in-the-loop AI for deep consumer insights, message generation, message bundle optimization, segment activation, and category performance benchmarking. In the past 10 years, Newristics has optimized messaging for 100s of world leading brands generating $100s billions in revenue every year.

Learn More

About TerraStrat

TerraStrat strategically redefines the banking industry through data-driven network optimization and predictive modeling. TerraStrat was founded by experts in Financial Services and Data Science. Through their Data Science as a Service (DSAAS) offering, TerraStrat acts as an analytic extension of their banking clients' teams to optimize how they service and market their customers across digital and physical channels. Every bank and credit union client has different strategies and goals, and TerraStrat customizes their business intelligence, profiling and segmentation, machine learning modeling, and geospatial location analytics to best help them achieve their objectives.

Learn More