This Request Is Already Overdue

Part I of my story: How iterative workflow improvements helped me keep my sanity

Wednesday, March 30, 2016

My third day at my new job as director of communications for a large non-profit organization, I got my first artwork request. At the top of the Excel form were the words, "Direct mail requests are due a minimum of 6 weeks in advance of the event date." This request was pretty much right at the exact 6-week mark, it turned out.

I'd already had a little orientation, although, as a single incumbent in a relatively new position, I was still finding my way. So I called my rep the print house and said, "Hey, I just received a new request for a direct mail piece from [business unit]. The event date is [redacted]. Can you tell me when you need the artwork from me?"

The phone was silent for a few seconds. I'd learn, later, that silence from my rep was usually a bad sign. But at the time I thought he was just looking at the calendar.

"Technically, yesterday," he responded, after consideration. "But I can probably give you until Monday."

I coughed up a little bit of coffee. "So I have two days?" I said, wiping weak brew off the desktop. (People sometimes say they did this, but I really did, in fact, do a spit-take.)

My rep laughed. "It's past noon on Thursday, so really, you only have a day and a half."

That conversation was my white-knuckle introduction to an utterly broken process. What follows, in three parts, is an inside look at the impact of building a more robust process, where to look for weaknesses and stress points, and how (literally) the design of calendars seems to discourage human beings from evaluating date ranges.

Chapter I: Already Overdue

The underlying principles for marketing are universally applicable. Every industry has a twist: Food suppliers rely heavily on coupons (especially in mail and newspaper inserts) to draw in new customers and keep current customers buying their product rather than the competition; technology companies use digital ad campaigns narrow-targeting their audiences at the point of contact, whether it be The Deck, AdWords, or paid sponsorships; and I could tell you what oil and gas companies do, but honestly, I have come to believe that it's a waste of money, based on my experience living in Phoenix during the Kinder Morgan Oil Pipeline Rupture of 2003.

My organization, like, well, all of our competitors, relies very heavily on direct mail to bring prospective clients on site, for events on a variety of topics. Direct mail marketing is, to paraphrase Winston Churchill, "the worst form of marketing except for all the others." Industry-wide, the response rate for direct mail marketing averages within a range from 0.25% to 0.75%. It increases to about 1.5% if paired with telemarketing, which, well, I'm not going to say anything else about that. Not only that, but in this particular industry, direct mail is almost always our third-largest source of new leads, behind referrals and Internet, and our second-most productive source behind referrals. (Internet leads convert to a sale at a lower rate, because there is less friction to becoming a lead, and because direct mail marketing targets qualified households.)

I should add that Part I of this story takes place in 2012 and 2013. So the ongoing shift into Internet advertising wasn't as far along as it is now.

This is all a long way of saying, our marketing efforts relied very heavily on direct mail at the time.

So, let's skip back to this phone conversation with my vendor. The clock is ticking, and I have a day and a half. Here's a rough idea of the gauntlet that a direct mail piece has to run:

  1. Design. Review. Iterate. Finish.
  2. Send proof to clients: in this case, the business units.
  3. Get revisions from clients. Review. Iterate. Finish.
  4. Send revised proof to clients.
  5. Eventually, clients are done revising. Clients sign off.
  6. Deliver file to print house.

In this context, what does "overdue" mean? Well, the print house and my organization have a guideline for when we want pieces in mailboxes: 14 days in advance of the RSVP deadline. So, assuming a minimum of heroics, the print house gives us a date when they have to have the artwork in order to rip the PDF, get us to approve the file (either digitally or in print; print takes 3 additional days), put it on press, dry, cut, address, sort into boxes and deliver to the USPS, plus the USPS's required 10 business days to deliver and the 14 days' notice before the RSVP deadline.

So this request was actually made after what I later came to think of as the last possible day to deliver artwork for a "standard" timeline.

And it was delivered right at the 6-week mark stamped on the top of the form.

Now, I panicked a bit. I'm not even going to pretend. It's my third day of work at a new place of employment, and as a single incumbent, there's not even anyone to tell me how to do my job. I'm already thinking, geez, have I made a terrible mistake?

But (after, very quickly, working my way through Steps 1 through 6 outlined above) I came to a realization that I'd learned in my previous life, as editorial systems editor at the Orlando Sentinel:

We had a process problem.

Chapter II: A Broken Process Is Worse Than None At All

Those first couple of months are a blur. I could reconstruct it by looking at file time stamps and the like, but honestly, I cranked out as much as I possibly could, as quickly as the requests were coming in. I reviewed the quarterly marketing plans, which were suspiciously full of gaps and inconsistencies, and sent out e-mails to my clients, asking them to please e-mail me requests early if possible, because I couldn't tell what was missing based on the marketing plans.

Eventually I managed to get ahead, after a lot of late nights at work and a lot of pleading with my clients. Then I was able to take an anatomy of what was wrong. This is an excerpt of my own notes:

I wrote at the bottom, "It does not seem like anyone has evaluated this process in a long time. Or ever."

One of the lessons I learned, after working for seven years on the design and copy desk in a newsroom, is that having a bad process is worse than not having an identified process at all. I assume that that's true in most business environments, although not necessarily all. In a highly regulated environment, having a bad process may not be as efficient as it could be, but presumably it works, whereas not having a documented and tested process could lead to a citation, or if, say, you run a nuclear power plant, death. (Yay!) But within the confines of a process whose outcomes are not regulated, it's often worse to have a process where the wheels come off routinely than to simply wing it.

Let me cite a specific example. When I worked at the Sentinel, we had this extremely complicated, Web-based in-house tool for reserving resources for special sections. Everyone who used it -- ad sales, marketing, production, systems and editorial -- complained about it constantly. It provided e-mail deadline notifications and could produce worksheets for the various departments, but didn't seem to flow into any other systems. So at best, it replicated something that each individual department could do with a spreadsheet. Worse, it gave us a false sense of security that the other departments were on the ball, and I can think of a few occasions when we weren't informed of an important change to the production schedule and were caught with our proverbial pants down at the last minute.

Then, a deus ex machina intervened: The central IT department for Tribune Company replaced an outdated version of Internet Explorer, breaking the ActiveX controls that this tool relied on. We had no choice but to stop using it until it could be fixed or replaced. (For all I know, it's been resolved by now -- this was late in my tenure.) Shortly after, I had a special section that I was responsible for, and I was very concerned that we were going to fail. Instead, in the absence of the special section planning tool, we knew we had to e-mail and meet more often to make sure we were on track. It became clear to me that we were actually doing better than before, because we were no longer being shielded by a bad process and poorly functioning tools.

So back to the process at hand: Artwork requests. I identified a few major weaknesses in the process, both in the front-end planning and the back-end request process. It looks a little like this:

The last point, about producing a quality product, is a key one. When you have 12 working hours to execute something, including revisions, you have no time to consider a second possibility or to revamp something that doesn't work. All you have time to do is make sure that it's complete, isn't wrong, and gets the job done at its core.

That's not quality marketing. Marketing is a creative product, not a manufacturing process, and so there's a certain amount of inefficiency inherent in the process. (An example of another process where there's inherent inefficiency: A friend of mine who's an engineer at Lockheed-Martin says that sometimes his job is to sit at his desk and think about how to solve a problem, and that how long that takes isn't necessarily predictable.)

But the inefficiency should be on the creative side: Thinking about what the right solution is, brainstorming artwork ideas or display type. Our inefficiency was on the planning side, and that was squeezing out any time for creative.

So my next step was to think through a model of the ideal process was, at least for where we were. Another excerpt from my notes:

Next stop when you want to give a process a makeover: Buy-in. For a process where you impact an entire department, you might have to have a brainstorming session, but I was the whole department, so we skipped that step.

I went to my boss, the chief operating officer, and had a discussion with him about the process. He was very supportive and said that I could do whatever I needed to, in order to make our work more efficient. He gave me a little background that I hadn't had: The marketing planning process itself was new, because it was clear that the processes in place before that were costing us money and, in the cases where we had multiple business units in one market, were stepping pretty freely on each others' toes.

He also gave me my first big assignment: At an all-leadership meeting the following month, he wanted me to present about leadership's responsibility over the marketing process and their need to buy into planning. Having modeled what I thought the ideal process was, this was the venue to roll out our process improvements.

My grandmother always says, "You make plans and God laughs." That couldn't have been more true.

Chapter III: If Change Were Easy, Everyone Would Be Different

I wrapped up my presentation on the need for my clients' leadership in the marketing process and their need to take ownership. To set the scene, we're at a hotel for an offsite meeting, I'm in a suit and tie, and I am standing front of the entire operational leadership of the organization: the heads of all the individual business units, the various vice presidents, COO and CEO. Public speaking makes me sweat, quite literally, so I probably looked like I was wilting under a heat lamp.

And I said, "Do you all have any questions about this?"

One of the executive directors says, "Wes, you're saying you need to have these requests to you earlier. My team says that they can't get their speakers lined up any earlier. Can't you just work faster?"

I took a deep breath and decided that my best response was to outline the challenge at hand. We don't record these meetings, so I'm sure there was a lot more "um" and "uh" and "oh" than I remember it. But here's how I remember my response: "The root of the problem is how our mail timelines work. Right now, your team requests RSVPs seven days in advance. Working backward, we request that the post office deliver pieces 14 days in advance of the RSVP deadline, so that's already 21 days, or three weeks. We have to have it to the USPS ten days in advance of that, so now we're at, so we're over four weeks. It takes the print house at least a week to print, cut, address and sort. So, door to door, that's about five and a half weeks. But right now, your team sends me the request at six weeks from the event date, or five weeks from the RSVP deadline. So by the time I have the request in my hand, the artwork is already several days overdue."

"So can we send you the request with 'Speaker TBD'?" he asks.

I took a deep breath. "That doesn't really help me. The topic determines what the piece is going to look like, you know?"

Another executive director raises her hand. "How much earlier do you want requests?"

"Two weeks earlier," I said. "Eight weeks. Earlier still would be better, but I need at least two weeks to schedule my work. What I do right now is literally just drop everything when the requests come in. That's really inefficient." I thought a minute. "Imagine if, every time you got an e-mail from me, you had to answer it right then and there."

Laughter. Broke the ice!

I continued, "So what I guess I'm asking for is that you ask your teams to really commit two weeks earlier. The other thing I need is for the sales teams to either live with the dates that are in the [marketing plans] or to send me a revision."

Next, I described what I was going to do on the process side of things. We were going to create a new form, and stop accepting submissions using the old form. We were going to cross-compare the marketing plans and the requests. And most of all, we were going to create a new approval form that my clients would have to sign and return to me to put on file, stating that they approved of the artwork as completed.

I saw some heads nodding with that one, so I took a minute to expand. "I'm really big on accountability, and I don't want you feeling like I railroaded you into anything, or worse, having something that you don't approve of in the field."

"Any questions about this?" I asked.

Nothing. I grabbed a handkerchief and more or less literally mopped about a gallon of sweat off my head, and went back to my seat.

Mission accomplished! Now, it was time to implement.

Chapter IV: Roll Out The Barrels

I'd already identified that I wanted to make a few concrete improvements:

  1. Make deadlines earlier.
  2. Simplify reporting of new artwork requests.
  3. Capture better information.
  4. Compare requests to marketing plans.

Now it was time to figure out how to do that. Part II of this series contains a lot of the details, wherein I put on my 'editorial systems editor' hat to find the tools and create the process (and continue to iterate it). Here, I think a discussion about philosophy is more relevant than mechanics.

I'm really big on quality product. That's a natural outcome of my background as a journalist and as a designer and copy editor, where my job was at its core quality control and getting the facts right. I take great pride in the fact that in my seven years at the Sentinel, I only had one published correction.

The objective of these improvements, which were, as I alluded at the beginning of this essay, the first step in an iterative process, was to improve the

I'm not an introspective person by nature, so I don't spend a lot of time reading business theory books or studying organizational theory. Perhaps I should! But I don't. However, even I'm familiar with Six Sigma. You might recognize the underlying goal as being somewhat Six Sigma-inspired, even if, since this is not a manufacturing process, not all outcomes are measurable or, when they are, are not (given the very small sample sizes in any given market area) statistically significant.

By requiring the teams to live by their marketing plans, making deadlines earlier, simplifying reporting and capturing better information on the front end, I was trying to create a stable process.

As the in-house agency, I figured that I could learn a lot from how outside agencies did their work. It turns out that I was right and wrong at the same time: Agencies don't necessarily have well-documented processes to produce quality outcomes, because they have large teams of designers and art directors and copywriters and QC departments to look for flaws. What they do, instead, is build a lot more time into the process. I've since worked with several agencies and found that they generally need two to three times as long as my team requires.

So, with that revelation in hand, I dug into what exactly I needed. What were the questions I needed to ask? How could I make sure the teams lived by their deadlines?

Once I felt like I had an answer, I took the next step: I built a submittable PDF form, with a really, really, really simple CodeIgniter service that recorded the data into a database (to prevent lost requests) and sent me an e-mail with the PDF.

Done, right?

Oh, no. That was just the first iteration. But for now, we were done.

I sent a fateful e-mail to my clients, notifying them of the new process, as well as new deadline requirements. It was late March 2013, and we were getting ready to work on artwork for June. And it worked! For the first time, I was getting useful information about what we were going to do, and we were getting the requests weeks earlier, giving me enough time to actually consider what something would say or how it would look before I did it. It also gave me enough time to take a three-day vacation, which I proceeded to do in April.

Then, failure struck. Let's pick up with Part II.

Later: Malformed XML File Cannot Be Parsed