Engagement = Proactivity?

Digital Assistants turn bank-customer conversations upside-down.

Do you know that feeling when you read a message notification on your phone which is from a beloved one? Thrill. 
What about those that are from your bank? Hassle: problems, things to take care of, pushy offers, hateful reminders. Banks have been on the wrong emotional side of the communicational playing field for too long. It is high time to change that!

Investigating the WHY?

It is a huge challenge to become a brand or service that customers love. You have to be expressly good in the followings:

  • Personalized, relevant content
  • Toned and phrased for the customer
  • Right on time
  • On the most appropriate channel

Frankly, most of the banks are failing in each of these points: they tend to send out generalized, official toned messages (in best case) targeted for a large segment. Send out times mostly depend on the capacity and schedule of IT legacy systems that control email and SMS gateways. In case the bank is able to send push messages, the process is often desynchronised from the other communications, causing duplications and inconvenient deliveries.

Tutorial on message duplication :)

Let me tell you a short story of mine to demonstrate the demolishing effects these processes have on your customer relations: I have credit cards at two banks but mostly use only one of them but kept the other just to have a back-up for emergencies (despite its monthly fee).

My bank is quite aware of this situation since I described it for them at least twice in the recent years. I had that naive idea that they will try to keep hold on me by low frequency, relevant offers on the only channel that I use (digital banking). Instead of that their automated service constantly calling me from different numbers and push personal loan products that I have absolutely no interest in. After a while I get tired of this and cancelled my credit card (which was easier than opting out from the marketing message service!).

How to lose a customer in ‘10 days’?

Why did this happen? I know a trillion reasons, including dull sales strategies and pushy agent routines, unachievable targets, department silos and so on. What bothers me the most is that my bank expected me to listen to them (constantly) why they failed to listen to any of what I communicated (I just keep the card as a backup, not interested in any loan offer, have a primary account where my salary arrives at another bank bound by mortgage conditions).


What I was for this bank? A missed opportunity. They could easily harvest the monthly fee (plus some annual nonsense) if they would just let me forget about the whole thing. Let’s see how they could have done it!

Listen. Learn. Be proactive (but not all the time)

Returning to the thrill you get when someone you like messages you: they know what and when to send. You are thrilled because of your good past experiences, all the interesting things, the fun you enjoyed. Is it too personal to be achieved by brands? I don’t think so.

There are numerous good examples:

  • Starbucks: everyones coffee. DELIGHT
    Oldie but goldie (but it works). Banks always express their doubts about bold communication because their products are not sexy. However, as we know, even a boring coffee can be turned into a piece of delight:
A personalized message to cheer up my mood.
  • Innocent Drinks: entertain your customers. ENTERTAIN
    This beverage company dedicated it’s social channels to entertain their customers: jokes, memes, videos, cute dogs and funny facts. They constantly learn what their audience like and improve: banks can do the same in a personalized way, push various ‘light’ content to their customers and auto fine-tune them for each customer based on the feedbacks and usage data.
A soft drink company entertains users.
  • easyJet: your travel analytics. INFORM
    This is an absolute favourite of mine that can be easily copied by banks: the airline uses personal data to create every travellers personal infographics to make an appetite for their next booking. Banking data creates a perfect opportunity for such individual summaries.
A snippet of easyJet’s personal history email
  • Poncho the weather app. INFORM & ENTERTAIN
    Poncho the weathercat is a great example of brand communication through a well-built persona. It is sassy and funny, but not only for the sole purpose of entertainment: it also makes sure that you always aware of the weather forecast. The company heavily invests in its ‘cat-sona’ with their writer team of copy/script/gag writers. What banks can learn from their strategy is that building a loveable and helpful character shows a great return over time through increased engagement and traffic.
When discussions are not about the main-topic anymore

Some entertain you, others give you great insights or inspiration. Their shared characteristic is the value their communication carries. Values that customers appreciate instead of blocking your callcenter’s number.

How can banks adapt to this?

As the title of the previous example section suggests: listen, learn, be proactive. Banks are ‘inactively’ listening to their customers for a long time, but only a handful of them started to understand what they heard, which is indispensable to the learning phase. Just a few examples how captured data can be turned into valuable insight:

  • Transaction data withholds the patterns and characteristics of each user. Advanced banks are not creating any segment aspects on their own but let the machine learning algorithms come up with the clusters because they know that 200 mini segments are perfect fit for various sales campaign visualizations, communication tone or the examples used for reasoning in a newsletter.
  • Financial data is not the only hero of insights. As customers merge towards digital channels, banks got the opportunity to record a vast amount of usage data. They can learn the hard truth about conventional naming when their customers fail to find a function, or forecast the destination of my next trip by looking at the last currency I selected in their currency converter.
  • Classic channels are somehow still separated from everything else, which is a huge mistake. They are also less monitored and exploited. This results in the really disturbing phenomena when one hand doesn’t know what the other is doing. Just like in my case, it did not matter how many times I expressed my situation, my bank kept bugging me. Callcenter, branch and mail data should be extensively used to identify customer preferences (e.g. backup credit card), sales opportunities (mentioned vacation plans) or frustration.
Information get easily overlooked or lost in traditional channels

These learning processes, either if they are supervised or not, will lead to insights that will help banks to come up with ideas how to nail all of the things I listed at the beginning. They will get a sense of what is working and for who, on which part of they day, etc. To complete this guide I would like to emphasise the best way to do it according to our (Labs team) experience.

Narrators may come up with excuses, brands don’t really

As I mentioned before the best way to learn and experiment is by testing multiple versions of the same or similar message. It means that in some cases users will get disappointed or even frustrated but in many other cases they will be surprised, delighted and entertained. It is pure luck to come up with a good enough solution for the first time so the goal is to minimise the ‘damage’ on your way there.

Segment dependent A/B testing (or rather A/B/C/X) is nothing new, it is one of the digital industry’s testing standards. Although it is exponentially challenging for banks to create, approve and deliver various message versions due to their long and overcomplicated processes and strong brand and communication guidelines. Those guidelines exist for a reason: it is easy to damage the brand when the message comes directly from the brand.

Morgen Freeman is the most beloved narrator of all time.

Our experiments investigate the effect of changing the direct brand-customer conversation by introducing a 3rd player, a middlemen who narrates the content and offers. Assistants seen as private concierges are capable to achieve this mental alteration. We encourage banks to invest in proper Assistant persona creation, feed it with interesting topics and set tones that differ from their general communication. By doing this, they will have a multifunctional, trusted channel to their customers that can be dynamically controlled, changed without a hassle.

Such a persona can (should) be proactive, reach out for the customers and push messages without driving the target crazy right away. If they make any mistakes, step across a line and feel intruding or overwhelming then they can learn from it automatically and correct themselves on the spot. Customers will have a significantly easier time turning an Assistant down than real human beings, they will be less frustrated and feel less guilty.

Its not easy to initiate conversations. Nor to maintain them.

Make no mistake, Assistants won’t turn your bank’s app into an entertainment center that people tap when they wait for the train or relax in their lunch break. But they are the current best shot for financial institutions to say ‘Hi’, interact, gather data and learn about customers on a daily basis. All in all I believe proactive Assistants are the best way to create/restart and maintain customer engagement in digital banking today. I encourage you all to start experimenting and drop us a message if you are stuck and need help!