Making Offers On PFM Platforms

Congrats to @pglyman and @kplynch for their quotes in an American Banker ($) article on customized Web ads. According to Lynch and his eCommerce manager, Steve Kruskamp, the PFM platform:

[W]ould let First Mariner pitch products only to people that might be receptive…the bank might use Geezeo’s software to note customers’ auto insurance payments, even those made to a rival insurance provider charged to another bank’s credit card. The software could then present an ad for First Mariner’s own insurance — and possibly persuade a sticker-shocked motorist to dump his or her current insurance provider.”

That certainly is the promise of PFM. But there are a couple of things that I believe that banks and credit unions will need to do to really capitalize on this promise. The best way to think about these “things” is to put them into two perspectives:

1. The FI perspective. Before pitching an offer for some other bank/credit union (CU) product — especially an insurance product — the FI needs to know: Is this customer a good prospect for this product for us? In other words, seeing that someone pays a car insurance bill is nice, but that’s no big deal — you’d be making a good bet if you bet that 90% of your customers paid car insurance.  The more important questions to address are “does this customer meet our underwriting criteria?” and “would we make this customer a pre-approved offer?

FIs also need to know whether or not they’ve made an offer for that product to that customer recently (or ever). Good marketers establish rules for how many times an offer will be presented to a prospect/customer, and the frequency with which those offers are made.

PFM presents a scenario to blow this out of the water. See a payment, make an offer. Disregard past (or current) marketing activity.

One of the biggest problems I foresee for FIs’ PFM implementations is the desire on the part of the online channel group to prove the “ROI” of PFM by making offers willy-nilly, then beating their chests heralding the “incremental” sales generated.  The potential for an even more disjointed marketing effort than what exists today is looming large.

2. The customer perspective. Kevin (or Steve, I forget who said it) said that the bank would make offers to people “that might be receptive.” That’s definitely the right perspective. But easier said than done, I fear. How do you know when someone is receptive? Simply because they just paid for a product or service? That could be too late, no? If the car insurance payment that the PFM platform captures is that customer’s first payment, then you’ve pretty much missed the boat on this customer, no?

Let me oversimplify things here: There are providers I do business with today. And providers who I might do business with in the future. For a change in providers to occur, there has to be an impetus for change.

If you’re under the delusion that simply putting an ad under someone’s nose is sufficient impetus for change, please stop reading this, and go back to Ad Age, or some other advertising blog.

A bank or CU using PFM as a platform to make offers must provide some impetus for the customer to make a change. Conceivably, it could be as simple as “We could save you 15% on that insurance payment.” Or — and I like this one even better — “Our other PFM users’ car insurance payments are, on average, 15% less than yours. Click here for more info.”

In other words, the bank or CU should use PFM to help make a customer become receptive. And if it’s going to promise a savings, it better damn well better be able to deliver on that promise — or the credibility of the PFM platform will be tarnished.

——

I’m very bullish about the potential for PFM in strengthening the relationship between banks and their customers, and between credit unions and their members. I see PFM as a platform for engagement.  And properly utilizing the data that PFM promises to provide should help improve FI’s marketing effectiveness and efficiency.

But here’s the lesson: Data is like a bullet. Bullets can be very powerful, and having more bullets is certainly better than having less. But used improperly, bullets can be dangerous and harmful. And having more bullets doesn’t mean that you have to shoot more often.

Advertisement

8 thoughts on “Making Offers On PFM Platforms

  1. Ron,

    It’s a lot more than this. The key competency connecting the dots between a PFM platform and the customer is the ability to craft an appropriate offer. The organizational capability for synthesizing behavioural analytics through to offer management, what I called customer dynamics in my book BANK 2.0, is absent from most banks.

    Until banks build something more than campaign marketing capability, the power of PFM as a platform will elude them.

    BK

  2. “Kevin (or Steve, I forget who said it) said that the bank would make offers to people “that might be receptive.” That’s definitely the right perspective. But easier said than done, I fear. How do you know when someone is receptive?”

    Easier said than done for sure. There are ways, although only the best of breed data analysis banks (or vendors) will be able achieve it. At a previous employer, we mined millions transactions to identify behavioral segments. The finer you segment, and the more behavioral the segmentation is, the more likely you will know who would be receptive to a particular offer. We achieved consistently higher than average response rates to direct mail campaigns through thorough analysis, and the same could be done online, even easier.

  3. Ron, I love this post. Great stuff. I just want to point out that FI’s will have to go even further than ““Our other PFM users’ car insurance payments are, on average, 15% less than yours” because the user in question might have a higher value car resulting in higher premiums. So people with the same year/model of car in the same type of neighborhood would have to be analyzed for that case to be really relevant.

  4. I would expect that if one only has a high level, i.e. not too detailed level of analytics, not as specific as I suspect Paul is mentioning, one can make a “Lending Tree” level offer, good but not great, and make it so its contingent, and then close the deal when/if the prospect responds and the underwriting or behavioral modeling plays out.

    Then, segment by segment, as you build Paul’s capabilities you can go right to the offer (given that your model’s success is profitable when the inevitable errors are included).

    Of course, a TRULY sophisticated system would detect if the prospect is Ron Shevlin, or Shevlin-like, and then quickly and quietly run away…. 🙂

  5. Pingback: Tweets that mention New blog post: Making Offers On PFM Platforms -- Topsy.com

  6. Bankers are sitting on a plethora of transactional data. Learning how to synthesize this data is paramount NOW! The type of tool set @Geezeo is offering brings another critical data set to bankers, the financial transactional data we do not have. Now is the time for us to pull all this data together and make it work for our customers to provide a better banking experience. As Ron says, it will be the center for our “platform for engagement”.

    @dmgerbino

  7. Ron, you make some great points. The difference we see in the data from the PFM (versus the internal data mining of bank customers) is it is across many different account types (deposits, credit cards, loans,401(k)’s, etc) so we are seeing a more holistic view of the customer’s financial life. And we can engage with the customer in a place of their choosing and, hopefully, present relevant marketing when they need it. We’ll be learning a lot over the next few months and I anticpate that Pete and our team will be working closely on it’s ongoing development. @kplynch

  8. @Brett: Your comment about the ability to “craft an appropriate offer” is spot on. It makes me shake my wonder to see so many “marketers” looking to Twitter or the newest shiny object (in this case, PFM) as the holy grail of marketing offers. And in the meanwhile, throwing away years of experience and testing on what offers/creative works best.

    @Paul: You’re hitting on something I didn’t make explicit in the post. Banks/CUs should be using the PFM data to integrate into and improve the response models they’ve (likely) already developed.

    @Morriss: Your point is valid, but when I’m trying to suggest is that banks/CUs should use PFM to engage customers, i.e., invite them to interact, so that the appropriate questions can be asked. I’m thinking that the peer comparison data helps create a reason to interact.

    @J-Eye: You joke, but a good model would predict that I’m unlikely to respond online. Even though I’m frequent user of the channel.

Comments are closed.