Campaigns that deploy across channels are a fundamental expectation from the executive suite. It is expected that marketing efforts are calculated programmes to deploy a message wherever people are and attract (convert) as many as possible to customers.
And rightly so. But the trick now more than ever is to put the message in the right place at the right time.
Pre-Al Gore (aka, before the Internet) it was straightforward. Now it seems like there is a channel to the consumer born every day. No single channel guarantees success so you hedge your bet and play everywhere — from the traditional trio of e-mail, telemarketing, or direct mail, to the extended channels of outdoor (billboards), radio, or television, or the emerging social or geo-push message and beyond.
For all the talk about disaggregation of media content, advertising is worse! Where do you put your money? Does it still make sense to manage the timing of the message across channels?
Where do you start? Budget? Goals? Audience targeted event anchoring? Message delivery vehicle?
Answer: All of the above.
This isn’t the classic cliché of three elements where you have to pick two assuring failure. You really do need to balance all of the questions to form the answer.
For example, if you want to target apartment renters with a subscription, you need to start with forming an understanding of the audience group. Are you going after all apartment complexes? Hopefully not with a single message!
How are you picking the complexes to go after? Is the delivery frequency optimised for the target group? Are the outdoor-weekend types even home on Saturday and Sunday to take a weekend FOD? Maybe they need something on Friday (with Sunday pre-prints!)?
Have you had your data scientist mine through the data and give you a profile of the target geographies to confirm viability? Model out a predicted FOD and rate. Model a conversion rate as well. Should you lead with digital? Can you match an apartment’s renter demographic profile to your product?
Okay, enough about the apartment dweller campaign. Let’s talk about getting the message out.
Regardless of the target group, reaching these people will require a multi-layered programme. Keeping things simple(r), let’s think through a back-to-school campaign that is going to use the classic channels: direct mail, e-mail, and telemarketing.
The message will highlight the return to normal: It’s time to wind down summer holiday travels, the kids will be back in school, and it’s time to re-focus on the world. A calmness. Peaceful reality. Take a breath and get re-focused on the happenings of the world around you.
What better way to plug back into the community with a subscription to The Daily Miracle? It’s easy to enjoy a classic package of news, entertainment, and community happenings and shopping with a subscription to The Local News — via a combination of digital and print delivery.
Data people, it is now your turn. You have seen the message concept, so what can you do?
I’m pretty sure you can find target groups: Let’s say one group is households with kids in high school, the other will be long-term residents who have empty or nearly empty nests.
Both groups are (likely) coming back from summer holiday travels (model the vacation stop demos to see), are getting settled into their homes for the fall seasons, and will turn their attention back to the community and get serious about the election cycle.
Data-wise and in terms of campaign deployment timing, it is time to think about how to sequence the three channels that will be used. It’s a classic scenario: Chicken and egg — or is it egg and chicken? Is it direct mail (DM), telemarketing (TM) followed by e-mail (EM), or EM and TM, with DM last?
The sequencing is sometimes decided simply from a cost management perspective — i.e., go with the free/least cost channels first and only move the remaining “list of prospects” to the pay channels if they haven’t responded on the free channels.
This is not a bad strategy. But, while on the surface this looks like it makes sense, when you dig deeper the cost base sequencing has some problems and inherent response rate distortions within it.
First, typically the e-mail address penetration for an acquisition list is very low, and for a reacquisition it is good, but typically at about 40%-50% of the list. So, at best you are able to reach only 50% of the “list” with e-mail.
Secondly, since e-mail addresses are almost entirely collected from past customers, the likelihood of response is fairly high within the list for people with an e-mail. So, you could see a 5% response rate on EM and a .9% on DM. The realty is that if you had DM’ed first to only those with an e-mail address, you could have seen the 5% with the DM send.
However, if you did the DM first, your cost per order goes way up because the cost per sent “piece” goes to about US$0.50 versus US$0.04 in an e-mail blast. This is quite the compelling story for always leading with e-mail. Low hanging fruit and low cost to deliver the message.
So, now you have the first channel figured out as always EM. But remember, you will get to (at best) 50% of your “list,” with a 30% open rate and a 10% click-through. This leaves a good number of your “list” still un-touched or non-responsive.
So, what is next in the sequence: telemarketing or direct mail? If you are still in a cost-based decision pattern, for the TM channel you need to look at your calling vendor contract.
Do you pay by dial-tone attempted or by order sold? For the DM channel you need to look at production/print cost and postage. Do the math and make the choice.
Remember, though, that, like EM, the TM world is a small section of the list. You are also wrestling with federal and state do-not-call rules and the migration of land to cellular phones, causing huge drops in available numbers, plus the rules of calling cellular are different than dialing land lines.
All in all, for an acquisition list, you just don’t have a lot of dial-able phone numbers available.
On the other hand, direct mail looks expensive mainly because it reaches virtually every address on “the list” while the other channels reach less than 50% of the available list. (Total responses between TM and DM will be close to the same; it is just the size of the denominator in the math that changes. Unfortunately, the shift is the wrong way because the denominator becomes every address on the list, not just those with a phone number or e-mail address.)
So, you’ve lined up your sequence: e-mail first, chase with a phone call, then finally send a direct mail piece. Remember — and I can’t emphasise this enough — that when you do the sequence in this order, the sequence is going to pick up the highest response rates first (in the EM and TM campaigns).
This will put a premium on the data scientists doing some high-quality work on the DM list. It also poses the question of should you make two passes through the list? Should you do an EM, then chase with TM then DM, and then circle back and do a second EM and TM?
After all, the people who didn’t click the first time will get the DM piece, and, maybe with the additional exposure to the campaign offer, will be “softened” and will respond to a second round of e-mail. Scientists, you need to pitch this as a test. If it works, design all of your cross channel campaigns as EM > TM > DM > EM.
With sequence decided, how about timing? How long do you pause between the EM and the TM? How long does the TM part run? And, unfortunately, how early in the process do you have to kick the DM bit off because of the production/mail delivery time?
Data folks, if you are tracking campaign history and response rates, you can measure the time from EM delivery until you’ve hit peak (and cliff) of EM opens/click-through (track them separately). Some folks say the EM life is just a few days. Some say longer.
Roll up your scientist sleeves and pin this down. You may end up with something like this chart (where blue = EM, yellow = TM, pink = DM, and green = started but no source identified).
You’ll see odd patterns. For example, you send the e-mail on Tuesday, but it isn’t opened until after you’ve done the TM and DM portions of the effort, and the order is coded as an EM even though you know the person got all three components of the campaign.
Scientists can then track the pattern and see if they can predict channel preference, which could be a preference to get all three before responding — i.e. those people waiting to see if the offer gets better.
Another thing to consider in the sequencing is in the target groups. In general, each generation and each PRIZM or PERSONICX (to name just two clustering methods) has its own patterns of responsiveness to channels by life stage.
Also dwelling type, age, and length of residence will have differing responsiveness patterns. Depending on how crazy you go in your segmentation of your lists, this can be either an important factor or not. Do remember that just because you can slice a list into 500 segments doesn’t mean it is worth doing. Segment sizes have to be large enough to measure!
That is an overview of channel sequencing. If you want to get over to the leading edge of using channels, how about using geo-sensing within your apps? Or, how about figuring out how an app/site registered user behaves on your site to do channel preference predicting and then custom sequencing.
For example, if someone clicks on a restaurant’s phone number, maybe he prefers phone interaction. That is another path in the science of optimising marketing based on behaviour.
For now, have fun, work the data, and be safe out there.