I've written about good customer service in the past, highlighting my excellent experience with L.L. Bean. I've postulated that quality customer service is actually free, because the incremental sales more than make up for "excessive" talk time and refunds/credits granted by your customer service representatives.

Here's a trivially-observed example of what lousy customer service will cost you.  In this example, from Verizon, it's a minimum of $100.

The saga unfolds after the break.

We made the switch to Verizon FiOS in December, after waiting for years.  It's a terrific product, with great HD picture quality, and speedy and stable high-speed Internet access.  The installer was excellent and courteous and made sure everything was working perfectly before he left.

The ability of Verizon to handle an easy premium offer (the infamous HD Free TV offer)?  Not so much.

I've been working email and phone conversations with Verizon since January trying to get my $200 Best Buy gift card, which I selected in lieu of the 19" HDTV.  Since then, I've amassed a string of emails with the Verizon executive escalation team which have gone unanswered since January.

Finally, in frustration, I resorted to The Consumerist's list of Verizon executive contacts and sent an email to Ivan Seidenberg on Friday at 5:30.

Talk about quick reaction!

At Monday at 9:30, I received a call from a very nice young man who was very apologetic and who was clearly empowered to do whatever it took to make me happy.  He indicated that the $100 gift card I received in January was the first part of a win-back offer and that I'd get the other $100 in month 13.

The problem was that I never responded to a win-back offer.  I went for the FiOS triple play/free TV offer.  However, I didn't want to get something I didn't deserve, so I let the young man know that if they just sent me a $100 gift card, I'd consider us even and not to worry about anything else.

Of course, he had no way to give me just what was owed to me ($100 of gift cards) and had to resort to shipping out a $200 card.  I practically had to beg him not to FedEx it to me.

So Verizon will get stuck paying a $100 penalty for bad customer service, and that's not counting the time and effort expended via the forwarding of the complaint email from Mr. Seidenberg down the food chain and the inevitable executive escalation report that will be fed back up the chain at some point.

Now, want to take a guess what I expect in January or February of 2009?  Yep, I suspect the CRM systems of Verizon have been so fouled up by the free TV offer debacle that they'll spit out another $100 card for me.

The worst part of the whole experience actually took place yesterday afternoon.

I received a very uncomfortable call from one of my prior executive escalation contact's supervisor.  The supervisor apologized to me for her employee's lack of follow-up to me, and let me know it would be "handled appropriately" and that the person would be "reprimanded."  Yecch.  It felt like one of those unconfortable visits when your neighbor's kid hits a ball through your window and they come over with the kid's ear in a vise grip and make them apologize.  I couldn't get off the phone quickly enough and couldn't believe that Verizon was treating its employees like this--both the supervisor, who clearly wasn't comfortable making the call, and the employee.

Can you imagine how this employee--who's a grown adult, with a mortgage, family, etc.--must feel? I wonder how far he/she will deviate from the Verizon manual going forward and actually even try to use their brain?

I also wonder what brand of bureaucratic punishment will be exacted.  Banishment from the department bowling team?  Six months of replenishing the floor candy bowl?

All it would have taken was for Verizon to have a system in place that let their employees solve customer problems the first time, even to the extent of spending a little bit of money to address the root cause.  This one resolution must have cost them in total the price of 2 or 3 TVs; multiplied by thousands, it's got to be one of the all-time promotional disasters.

What are you doing that's L.L. Bean-like?  What are you doing that's Verizon-like?  More importantly, what are you doing that's more of the former and less of the latter?

photo credit: Daquella manera

Bookmark to:

OK, as a CMO you're going to be fired in ~23 months anyway.

But a recent study, carried out by Marketing Management Analytics, Financial Executive and Ed See, makes me think that more senior marketing executives should double-check the shine on their resume and perhaps consider a few more networking lunches in the near future.

A very brief summary of that report found in Ad Age frightens me and brings to mind a few courses of action that you can take today, if you find yourself in a similar situation.

A few quick points from the study follow. The CFOs interviewed said that:

• 9/10 of CFOs don't use marketing ROI metrics in development of marketing budgets.
• 7/10 don't use their company marketing forecasts for earnings guidance or public disclosures.
• 6/10 believe their marketing departments have inadequate understanding of financial controls.
• less than 4/10 said their company marketing forecasts were audit-ready.

From the CMOs and senior marketing leaders surveyed:

• Only 33% of CMOs reported "full cooperation and an open dialog" with finance in setting metrics and methodologies for marketing ROI calculations in 2008.
• Only 10% indicated that they could forecast the effect of a 10% cut in marketing spend.
• Only 14% said senior management had confidence in their firms' marketing forecasts.

To summarize:

1. CMOs ability to estimate their impact on the business isn't even of ballpark (+/- 10%) quality.
2. CMOs don't have a great grasp of providing accurate and reliable forecasts of investments that will pass reasonable audit scrutiny.
3. CMO's don't play well with others.
4. The CFO's have so little confidence in the marketers that they crunch the numbers with relatively little input from the marketing team.

So what do you do, if you're a senior marketing leader and find yourself in a situation like the above?

First, accept you have a problem.  Get with your CFO and let them know one of your top priorities is to fix the problem.  Make sure you both understand what the CFO's specific problems are with your marketing planning, forecasting, ROI calculations and assumption development processes.

Second, triage. Figure out what you're doing wrong, but as importantly, figure out what you're doing right (and I know you're doing a number of things correctly) and learn from those processes that are working well.  Work out a punch list with your CFO, and understand the "A" priorities.

Third, partner with your CFO to improve your process. I insist in the concept of the financial guys driving the calculations and modeling side of the equation, the marketing team leading the assumptions and scenario planning, and then a "team of two" on running the iterations of your plan or forecast.  (The marketing team should pressure test the calculations and the financial team should pressure test the marketing assumptions.)

Fourth, own the issue. As CMO or senior marketing leader, if your financial leadership doesn't trust your ability help get to clear marketing metrics and ROI calculations, it's your problem, not theirs.  Make it really clear that you own the forecasts, plans and budgets and that you'll be the one to be held accountable for the hits and misses--not the financial guys who built the models, nor can you blame the last CMO for problems.  It's on you.

So do you have problems with your planning and forecasting? Are you a financial pro who needs some help in improving the planning and forecasting competency of your marketing team?  Let's talk, here or on the phone.

photo credit: Gary Denness

Bookmark to:

A recent article in the New York Times got me thinking about the psychology of marketing again, and how some basic principles are used or under-used in social media marketing.

"Warning: Habits May Be Good for You" explores how an anthropologist turned to marketing experts from CPG companies like Procter & Gamble to help increase the incidence of hand-washing with soap after using the toilet in the nation of Ghana to improve the health of children.  Obviously, this was an important effort and I was encouraged to see marketing practitioners as instrumental in helping achieve success in this endeavor.

As I was reading the article, it struck me that many of the techniques used are found in Robert B. Cialdini's classic Influence: The Psychology of Persuasion. I'd lent out my copy a few years ago and, thankfully, had to buy a new version which includes an epilogue written by Dr. Cialdini in 2007.  My re-read then triggered a few thoughts on social media.

If the social media crowd can stop navel-gazing long enough to do some quick research and scientific work, boy will money be made. [More after the jump.]

Dr. Cialdini's research caused him to identify six key weapons of influence:

1. Reciprocation
2. Consistency
3. Social Proof
4. Liking
5. Authority
6. Scarcity

Many of his examples focused on face-to-face selling or just plain interactions with others.  When I first read the book (originally published in 1984) for the first time in 1995 or so, I was struck by the fact that my company at the time was using all six of those principles, but in various offline direct response channels.

We didn't intend to use Cialdini's principles; rather, we evolved into using them through years of scientific univariate tests.

Which of the six principles is most important to a social media startup? From what I can see, it's the concept of Social Proof.  Jason Calacanis becomes a Twitter advocate and, because several of his followers also pick up on Twitter, it soon provides proof to a certain circle that Twitter is a good thing.  Boom, the product takes off.

What's the problem with the concept of Social Proof in growing social media? Try this:  Ask anybody over 40 who's not a regular reader of Alley Insider, Valleywag or Paid Content what Twitter is.  My bet is you get a blank stare two out of three times.

Simply relying on Social Proof can get you pigeonholed within a very select segment of people.  Now, for companies that provide great services to that limited target audience, it can be very good business indeed. (See how Rafat Ali made out with Paid Content.)

The key is how to use the concept of Social Proof to expand beyond that initial, lucrative target audience and become more ubiquitous.

Summary and key takeaways

1. What's "old" is new again. Go back and do your research.  Classics of marketing and psychology are relevant today, just as the concepts they cover were 10,000 years ago.  Plus, with Amazon, it's now cheap to really build your library of information!
2. Consider how humans are hard-wired. When building products, services and delivery methods for those products, think about basic human infrastructure and how you can find the path of least resistance to your ideas.
3. Remember that great psychology can't overcome a bad product. You can certainly fool some of the people some of the time using some of Dr. Cialdini's "click, whirr" discoveries.  But that's still only going to take you so far.  Start with great products.

I'll look at more of Dr. Cialdini's concepts in the future.

In the meantime, how have you applied psychology in your marketing?  Or, when have you seen good (or clumsy) attempts been applied to you?

photo credit: HeffTech

Bookmark to:

You plan for offensive operations, while you prepare to play defense.  You'll find this concept in both warfare and sports, and it's applicable in business as well.

I much prefer playing offense, because that's where you score and generate revenues. A strong business offensive plan also limits the amount of places you'll need to prepare to play defense, freeing up more resources for--you guessed it--playing more offense.

What's an example of planning as opposed to preparations in a marketing context?

A great example can be found in the United States Postal Service and the annual postage increases.  If you're using direct mail as a marketing channel, you can be sure of two things:

1. Postage will go up, driven by higher USPS costs and the rapid decline in the volume of first class mail.
2. Automation requirements will become more rigid, which limit your ability to be creative in your mail formats.

One of the key cost components of your direct mail marketing efforts is always going to go up (which it has done for years) and is going to go up more rapidly than in the past because of the USPS's need to cover labor costs and, now, energy price increases.

The USPS is trying to counteract the labor cost increases on their side by providing a bigger carrot for mailers who automate their mail more effectively and using a bigger stick (higher prices) on those who don't.

In the past, we used to counteract higher postage rates by getting more creative in the the formats we'd mail. Moving to heavier packages, with larger non-standard sizes paradoxically would counter the postage rate increases.  Not anymore.

So, what's the difference in "planning" for the USPS's annual price action vs. "preparing" for the same?

In "plan" mode, you understand what you're likely to face in the future and are testing for it proactively today. For example, at a past employer,  we knew the USPS was changing the automation requirements to require a certain amount of flexibility in the mailings or face significantly higher postage.  We were using a very rigid piece that was going to lose its automation discount and the price increase would have tanked our ROI.  So, we immediately began testing a number of different formats that met the flexibility requirements for next year, so we'd know the impact on our marketing ROI and could further develop plans to overcome any degradation.

In "prepare" mode, you've got to be ready for anything and then respond in a reactive way. You could prepare by figuring out what the price increase might be.  Or what the automation requirements might be.  Or any other number of things.  Being ready for too many options would suck away marketing resources into planning and analysis and away from the key task of acquiring more customers.  Plus, once you reacted, you still wouldn't know the impact on your ROI.  Rolling out untested responses to competitive or environmental changes is the number one way to destroy your marketing ROI.

Plan = Proactive = Offense

Prepare = Reactive = Defense

Summary and key takeway
When you have the choice, play proactively and play offense.

I'm not saying you shouldn't prepare for the unexpected, whether it be competitive efforts, consumer preference or even the weather.  Instead, you should be aggressively planning to launch new channels, acquire more and better customers and developing new and better products.  Then, on the planning side, be prepared for the 20% of things that could cause major disruptions to your business.

What do you think about the proper Plan to Prepare ratio?  Have you ever had an example where you suddenly realized you were preparing too much and, hence playing defense when you should have been playing offense?

photo credit: selena marie

Bookmark to:

Today's article is a reminder that customer service, when done correctly, isn't an expense.  It can be your #1 marketing tool and put a wide moat between your business and your competitors.

L.L. Bean is the standard against which your phone support should be measured.

From their website:

Guaranteed.  You Have Our Word^TM.

Our products are guaranteed to give 100% satisfaction in every way.  Return anything purchased from us at any time if it proves otherwise.  We do not want you to have anything from L.L. Bean that is not completely satisfactory.

Unlike a lot of companies, L.L. Bean really lives by their guarantee.

I was preparing for our annual Outer Banks vacation this year and checking out our Bean Sunbuster shelter.  We've had it for four years--it's a fantastic product for use when spending long days in the North Carolina sun.

However, the last time we used it, one of the large fiberglass poles broke.  When I pulled it out today, I remembered that and decided to call for a replacement in enough time to receive it before we left for Nag's Head.

I called and immediately got a very courteous phone representative who found the part number in under one minute.  Secondly, he asked for my name.  The instant I gave him my last name, he confirmed my address--he'd apparently pulled my record up via the ANI when I called--and thanked me for giving him the correct pronunciation of my last name.

Before I could even ask the price, he let me know it would arrive at my home via FedEx on Saturday and was there anything else he could do?

The price? Gratis.

I asked about the price, and was more than willing to pay for a new pole.  No dice. L.L. Bean wanted me to be happy and refused to take my money.

In my entire interaction with the company, I didn't hear a representative worried about talk time, first call resolution rates or any of the nonsense you normally encounter in call center management.  He clearly owned the problem, cared about the problem and was empowered to make me happy.

Will I buy from L.L. Bean again?  Absolutely. (In fact this experience makes me think about why I'm not doing so much more frequently)  Will I tell others?  I just did.

Summary and takeaways

1. Treat customer service as a marketing investment.  Not an expense.  Give those front-line employees all the tools they need to keep your customers happy and let them have at it.  Treat product replacements as marketing, not a cost to be squeezed out ruthlessly.
2. Ignore the traditional call center metrics. Stop beating on your employees for talk time "infractions" and other petty nonsense.  Your customer service team is your key to success.  Treat them and pay them well and let them win for you.
3. Fire a few MBA's. Get rid of the people who come to you with "metrics" about why you can't do #1 and #2 here.  Take the money you save and use it to pay more to your customer service team.

Bookmark to:

After lashing out at Verizon and Vonage a little while ago, I thought I'd move on to other topics.

But Verizon can't help but make simple mistakes in their direct mail efforts.  On Friday, I received yet another first-class mailing, this time promoting Verizon's FiOS Internet service.

Since I've now been a triple-play customer for two months and paid two bills for the service, this mailing is slightly redundant.  Unlike the last promotion, this one has doesn't even include a letter signed by an identifiable person.  I guess they knew this was so poorly done that nobody wanted to sign it?

Of interest is the fact that the return address for the FiOS Internet service is Annapolis MD, while the TV promotion came out of Irving, TX.  This confirms my suspicion that the FiOS product is multiple business units hammered together.

Summary and takeaways

1. Share your internal customer lists across multiple business units and suppress against those customer files when you are undertaking a promotion.
2. Always suppress against your transactional/payment database.  Your internal marketing databases might not be updated, due to either sloth or neglect, but the department that counts the money always knows who's paying and who's not.

Bookmark to:

I've recently received several direct response efforts that are unfortunately poor examples of our science and craft.  But they provide some important lessons that we should all take into account when preparing our next campaigns, whether they be acquisition, retention or winback efforts.

The two efforts I'm going to write about appear to suffer from at least one flaw that I try to address with clients and employees.  I call it "fingertips."

By that, I mean that the nuts and bolts and detail of the marketing effort or analysis need to flow from the written page or the computer monitor through your eyes, be processed by your brain and then exit via your fingertips to the email, instant message, Excel spreadsheet, etc.  It's not about forwarding the print production schedule you get from your printer or assigning the analysis of the media plan to your least-experienced employee because it's tedious.  Having "fingertips" means you know the detail because you've not only seen it, but processed it and then had it exit via your keyboard.

I didn't see the "fingertips" of the marketing managers in these efforts.  Al Pacino, in his famous rant at Kevin Spacey in the movie  Glengarry Glen Ross would have put it more bluntly.

"Where did you learn your trade?"
The problems I've identified by these marketers--Verizon and Vonage--are all very basic and things that, with a bit of training, would have been easy to avoid and would have saved those organizations tens or hundreds of thousands of dollars.

Verizon and the FiOS data debacle

My first example, using Verizon, is that of a two-part mistake.  The first part is not understanding your data and the shortcomings of it.  The second is a merge/purge mixup.

I've been waiting for Verizon's FiOS triple play bundle to arrive in my neighborhood for over a year.  I wanted to consolidate my data, TV and phone service with one company and have heard nothing but rave reviews about FiOS from everyone I know who've made the switch.  So you can imagine my glee, when on November 20, 2007 I was door-hangared announcing the availability in my neighborhood.  I immediately called to place an order and arranged an installation date of December 7, 2007.

Everything went well, including the transfer of our home number from Vonage (more to come on this later) VOIP service to Verizon's product  At the end of the install, I had a 15/2 Internet connection, 2 standard definition TVs, 1 HD TV with DVR and my home phone number ported over from Vonage to Verizon.  All working flawlessly and installed by a thorough and very pleasant technician.

In early January, I paid a bill for my triple play service.

For some reason, on January 25th, I received a 9" x 12" matched mailing from a Thomas M. Crowder at Verizon offering me, a FiOS TV subscriber, the option of adding FiOS Internet and getting a month free of service.  Nice offer, except I've been happily using FiOS Internet service for over a month.  The mailing was a complete and utter waste of money and it should never have hit my home.

Verizon's databases are a mess.  (If you're in doubt, I suggest you spend a few minutes traversing their website.  It's clear to me that the FiOS business is at least three business units hammered together, and it's clear they're not talking to each other.)  But if the marketers responsible for this campaign at Verizon had good "fingertips", they would realize this and understand the risks to their campaigns.  Namely, that the wrong efforts would be going to the wrong people at the wrong time.

How could Verizon have avoided mailing me--and many, many others--an expensive promotion that will always yield 0% response to my segment?

1. When in doubt add more suppression files.  Had the people running this FiOS Internet campaign simply understood that it was likely that some FiOS TV customers were not showing up on their records as having the double-play product, they could have added FiOS TV only customers as a suppression.  I would have showed up as a suppress on any of these.  Of course, maybe the customer databases are really in bad shape.  Then, doing the following is an extra catch;
2. Suppress paying customers. If your marketing database is separate from your transactional records, grab a suppression from the transactional system and suppress people that have paid for the product.  While you're at it, look at the merge output and, if you notice discrepancies, get your IT folks and DBAs in pronto.

Alas, Verizon is not alone in their direct response campaign challenges.  Vonage has winback problems of their own.
Vonage and the print production processThis one is less of a really basic marketing sin (targeting the wrong people for the wrong product at the wrong time) as one of marketing process.

During the process of moving to Verizon, I ended up having to cancel my Vonage VOIP service.  The first thing that happened was on December 7th, Verizon ported my number from Vonage to them.  That means that since December 7th, Vonage knew I was no longer receiving phone calls using their service.

I actually called to cancel Vonage on December 12th and, after enduring a Vincent Ferrarri-like experience with the retention sales consultant, completed the transaction.

I received the predictable win-back efforts via email on December 20th and via direct mail on January 10th.

Yet on January 26th, I received a piece offering me (as a customer) free service if I recommended Vonage to a friend.  One of three things happened here:

1. This is a test they intentionally mailed me.  In which case, it's a case of somebody simply not paying attention to the segmentation.  (After all, why would a non-customer talk a friend into buying a product that I haven't used for almost 50 days, with the reward being a free month of a product that I can't use?).
2. Vonage's print production process is too long. If I assume a drop date of around January 16th, this tells me that Vonage had around 34-39 calendar days to produce the direct mail I received (a 5" x 9" window envelope, with a personalized tri-fold piece with tip-ons) after they knew I cancelled. That's much too long for that piece, even in volume.
3. The suppression file was incorrect.  It's possible, as with Verizon, that they didn't suppress the right names.  A simple add to all of Vonage's merge/purge processes of the LNP (local number portability) change requests as a suppression file might have cut me out of this wasteful mailing.

Summary and takeaways
Make sure you've got people that understand the details of how all your marketing efforts run.  Make sure they understand their craft and are not at the mercy of their agencies or vendors for information and understanding of how their campaigns work.

For example, if your marketing manager can't tell you how a merge/purge or name selection process works, including being able to justify every word in the merge instructions or can't explain the details of the production capabilities of your lettershop vendor, I guarantee you're promoting to the wrong people and/or taking too long to do so.

Other quick thoughts:

1. Over-suppress names in your merge process.  You can always add back those names if it turns out you were too tight.
2. Assume your product databases are out of sync and act accordingly when selecting names for promotion. Further, your transactional and marketing databases, if kept separate, will tell you two different stories.  Look for discrepancies between the two, which will not only give you opportunities to save money, but also find new leads to promote and even new product ideas.

Bookmark to:

Today's example of online marketing waste, which probably went awry someplace in the 4% of process (see Focus on the Four in our featured article section), comes from Samsung in the form of a promotion for their new Blackjack II cellular phone, which is available to subscribers of AT&T's wireless network.

A couple of reasons why it's a good thing I received this promotion:

1. I'm an AT&T customer, meaning no worries about porting my number over or breaking contracts.
2. I used the Blackjack in the past, and Samsung knows that because I registered the phone and my email address with them.
3. I liked my Blackjack phone, although there's no way for Samsung to know this subjective piece of information.

Now, a couple of reasons why it's not a good thing I received this promotion.

1. I'm an AT&T customer, but am currently tied to a contract with my iPhone.
2. I used the Blackjack, but my usage of AT&T's services has gone up substantially since I switched to the iPhone.
3. I love my iPhone and AT&T probably can divine that by looking at my usage pattern with iPhone vs Blackjack.
4. AT&T and Samsung could have avoided reaching out to me by just thinking for a minute or two.

Something's clearly not in sync between Samsung and AT&T Wireless. The Samsung promotion leads to a very nice landing page and the "buy now" link took me right to a page where I could see the offer to upgrade to the Blackjack II. A simple login to my AT&T account, however, reveals the problems that I'm not eligible for an upgrade, etc. etc.

So what would I suggest Samsung do differently in the future?

• Segment. I'm sure Samsung did do a whole series of tests with different headlines, offers, creative and the usual to different target audiences. However, it looks as if they forgot to consider who absolutely wouldn't buy the product. My usage and purchase history of handsets for use with AT&T makes it pretty clear that while I'm possibly in the market for the next generation iPhone, I'm not going to go to the Blackjack, at least not now.
• Share data with your partner. Clearly, AT&T and affiliates sell a lot of Samsung product. I'm sure they've got a close relationship and one that should be able to--once you get your lawyers comfortable that you're on the up-and-up--allow them to merge the email address at a third party. And, at the very least, drop the email addresses of individuals that are not likely to respond, leaving only a small portion of names in as a test to confirm the hypothesis.

Your takeaway: As one of your up front questions when developing campaigns, ask yourself the simple question "who absolutely won't buy my product" and make sure you avoid marketing to them. And then to a clean test to confirm your hypothesis. Then either bank the improved ROI or (my preference) invest it into more big, needle-moving tests.

Bookmark to:

Humans and chimpanzees have a match on about 96% of their DNA.  That's not a lot of difference between you or I in our automobiles, sipping a Starbucks latte while chatting our cell phones and our pan troglodytes relatives in the rain forests of central Africa.

And that 4% is about the difference between dramatic marketing success and dramatic marketing failure.

How can you avoid being a marketing chimpanzee?  Just Focus on the Four.

In my experience, failures in direct response marketing efforts tend not to come from missing the big things.  Although big misses do happen from time to time, they are usually easily identified and rectified through improving the people or process.

It's the problems in the 4% of your efforts (sometimes wrongly thought of as "on the margins" or "blocking and tackling") that conspire to sap your marketing efforts of their ability to dramatically improve results, grow your business, and keep ahead of your competition.

While you can--and should--think about the 4% in all areas of your direct response marketing efforts, there's two areas that I'd advise looking at: Testing and Process.  Having a high degree of focus on the 4% in those two areas means you have to look at the two areas in very opposite ways.

In Testing, you have to focus on creating wide variances of results (-50% to +100%).  The biggest mistake in testing is fear of failure, resulting in lots of flat results that offer no help in setting direction for future tests.

In Process, you have to focus on being able to create replicable and stable results from both your rollout and testing efforts.  The biggest mistake in process is in turning the "routine" or "automatic" parts of the campaign over to less-experienced staff (with little or no training!) or outsourced to vendors (who in turn, turn over that work to their less-experienced staff).  The result is that you have a high degree of variance in your campaign, test and financial results.

Let's look at each in more detail.

Testing

96% or so of your tests will yield nothing or be basically flat. (I'm including in this category the +/- 5-10% results that will regress to the mean in rollout.)  Flat tests, while a natural part of what we do, drive me a little crazy.  They don't immediately show a path to improved profit and they don't immediately show paths to avoid.

You will never succeed in driving your business unless you are willing to have big winners and big losers, and you're willing to analyze both the winners and the losers.  The reasons for big winners are obvious.  You find a new offer or channel that can drive 50% more volume or reduce CPA by 30% or whatever and you roll it out.  Further, that test shows a fruitful path that leads to more tests that can build on your success.

The big losers are even more useful.

The losing tests can help you modify your hypotheses about what might work in the future and how those hypotheses apply to your customers.  However, you have to make sure that your culture makes it safe to bring those big losers to the fore and properly debrief against your findings.  The test that depresses response by 70% teaches a lot, but you have to be willing to look at the message it gives you and not shoot the messenger.

An example from a past life was in testing "straightforward" copy in direct mail solicitations for a continuity product I used to sell.

We used a lot of typical direct response techniques--strong offer, multi-level game/gimmick, double/triple guarantees, longish benefit-oriented copy in the letter, product sample and so on.   For years, we heard from both our customers and creative agencies that they wanted "no B.S." or "just the facts" in the promotion.  Time and again, we tested simplified copy, offer presentment and so on.  And time and again, we yielded depressions in both front-end response and back-end take, yielding massive depressions in ROI.

Because we were getting such a big swing in results, though, I was convinced there was something missing that could move the needle to the other side and generate a big positive result.

The breakthrough hit our creative agency one night at about 1 a.m. after a long, frustrating day of testing concepts in dyad interviews.  What happened?  It turned out we weren't being simple and clear enough in our creative and copy.  We had always been tempted to insert just a bit more benefits or to clarify the offer just a bit more.

Getting rid of the temptation to add "just a bit more" was the key to uncorking a new creative concept that swung the needle 180 degrees and lifted both response and back-end take. The result was a long-term control that worked on a number of products, in a number of channels and that produced a series of additional strong tests.

Testing takeaways:

1. Look to drive big winners and big losers.
2. Prioritize tests that are going to be likely to have big impact and test them cleanly.
3. Make sure you spend time with the big losers.  There can be gold in them thar hills.

Process

Unlike testing, where you need to open your mind and be aggressive to drive big results in the 4% of non-flat tests, process is an entirely different thing altogether.  Here you need to focus on the detail and not drive big variation.

It's the final 4% that you don't cover off on that will kill your campaign, your tests, your product and possibly your career.

Some examples of process problems that can bite you:

Who's developing the merge/purge instructions and analyzing the results of the merge?  If the answer is you don't know, or the lowest-paid or least-tenured person on your team or, worse, the merge/purge vendor or nobody, you can stop reading this article and get on that immediately.  The details associated with your merge and name allocation process can seem somewhat mundane and routine, but the output of your selection and allocation process is 40% of the campaign, as the old DR saw goes.

Panel splits. Are you sure your test and controls comprise nths of the exact same population?  And is that test population the same as the rollout population?  I've seen cases where people were doing extraordinary things in their merge, with very sophisticated testing to ensure proper allocation of names across multiple products, offers and with various timing.  However, the offer, creative and step tests provided inconsistent (sometimes disastrous) rollout results because something as seemingly basic as the panel splits were done incorrectly.

I've always followed the 40/40/20 rule when allocating time and resources to direct response campaigns and planning.  40% of my time was working all aspects of the offer, 40% was on the lists (i.e. the target audience and recipients of the promotion) and 20% for everything else.

Far too often we leave the "glamorous" work, such as the creative (which is only a piece of the 20%) to the most seasoned members of our team and leave the "grunt" work, such as the list ordering and list testing (which is solidly behind 40% of your results) to unseasoned and under-trained staff. You can even flip the responsibilities around and probably immediately improve your results!

Process takeaways:

1. Follow up on the details.  At the very least, know who's responsible for that 4% in the corners.
2. Make sure the responsible person knows they've got the ball.
3. Look at your process periodically.  What worked in a channel last year may no longer apply or there may be an easier and less error-prone way of doing that process now.

Summary

Seek to drive more variance in the 4% of your testing that won't be flat.  Don't worry about looking bad with the losers; as long as your hypotheses are solid, you'll learn from them.

Drive more consistency in the final 4% of your campaign process that has the potential to hurt both your overall campaign performance and your testing.  Make sure you understand it and that your team knows there's nothing "routine" about it.

Focus on the Four, and you won't be left eating bananas.  Unless, of course you like them!

Bookmark to: