What Works? An Analysis of Campaign Results and Best Practices.
Part 2: Campaign Parameters
Dr. Tony Rizzo, CMO, Marquis
One of the questions I get asked frequently is, “What works?” So, taking that question to heart, we executed a very extensive analysis of campaign performance throughout campaigns that we managed and produced over 2019. I’m going to share those results with you today.
We stratified the results also by assets. I was curious to know that the smaller you were, did you do worse than someone much larger? You can see the numbers. Not really. So consumers are consumers. Your asset size does not really play a factor in your ability to be successful with this approach.
We looked at the strategy as well. We stratified across three primary strategies: acquisition, cross sell and retention. And here’s how to think of these – red to green. Red, the farther away you are from the product or institution. Green, the closer you are to the product or institution.
So what does that mean? I’m going to do a checking promotion, a prospecting acquisition campaign. Or I’m going to do a checking retention campaign where I already have the checking account. But you’re going to sell me something like access, convenience services, perhaps a safe deposit box and add on a bolt so I’m closer to the product.
When I’m closer to the product, I already understand that you’ve already captured my attention. My psychological fence, if you will, has been opened versus the acquisition side of that where I may or may not know you. I don’t have this product. I have to think harder. So it’s kind of common sensical that you would think that acquisition all the way to retention, you’re going to have lower to higher response rates, lower to higher marginal ROI.
Next, we looked at the difference between a campaign and a matrix. What does that mean? Campaign – one-time event. Matrix – ongoing frequency – onboarding, reboarding, for example. Or perhaps it was an auto loan campaign that had multiple flights and multiple touch points. In every case, one-time versus multi-times, multi-times outperforms in spades versus the one-time wonder. And you can see that’s true amongst response rates, accounts open and average balances.
Now let’s take a look at channels by direct response and indirect response. Remember direct response, you promoted an auto loan, you bought an auto loan. And you can see we’ve stratified this across direct mail only, direct mail plus email, and then email only.
And here’s the thing that’s interesting. You look at these from a response rate, they’re all fine. There’s nothing wrong with them. They’re pretty good response rates. But look at the balances. If you look at the balances, they’re 126% better when you use those two channels. You’re going to see that throughout the rest of this presentation. The combination of channels substantially, I think, in every case but one, lifts balances.
Why do I care? I’ve never seen an annual report that says we had 7% response rate. I have seen annual reports that speak to marketing’s ability to bring in dollars. So I’m going to focus on dollars – balances per account that come in the door. The numbers hold up when I also look at direct plus indirect response. Same thing, while not as dramatic, same kind of premise will hold.
Let’s take a look at data use and package design.
When we talk about the use of data, we categorize data six different ways, in terms of how we weaponize that data to communicate with the consumer. Could be FICO based data. Could be based on the timing, that’s marketing automation, the timing of the offer. Could be based on variable use of content. In other words, I can go in and send you a letter that says your home, based on what sold within your neighborhood, is worth X. Is it time to buy a new home? Is it time to refi? That’s using content in a variable sense to drive a better connection with that consumer.
Could be branch based. Could be payment based. We do an extensive amount of work with payments, to be relevant, to be personal, and more importantly, to be specific to answer questions ahead of time that the consumer, the member, might have.
I would also use the life stage as a variable driver. We’ll show some examples of that as well. One of those you’re seeing here.
So our typical direct mail letter package looks very similar to what you’re seeing here. A lot of variable photography, variable copy. I’m going to have different offers going in based on what the data is telling me. I’m using a non-window envelope. And in a lot of cases we use a lot of live stamps. It reduces a lot of the clutter on the envelopes.
Again, I’m trying to be personal with you. This envelope that you’re looking at is more personal than one that, say, has a mailing indicia. So that’s our standard package. That’s when I talk about letters. This is really what I’m talking about.
As we go through this now, if I’m doing postcards, same kind of approach. They’re highly data based. They are all oversized. We do oversized cards so they stand out in a stack of mail. Your postcard rises to the top.
If I’m doing email, we do all manner shapes and sizes of email. They are all highly databased. But they’re also very what I call code simple, one column designs that are designed to easily render on 90% plus of the platforms that are out there. Keep your email simple. Don’t do a lot of columning. Don’t do a lot of special effects within the email. It ends up not working as well as something very simple and straight to the point.