Friday, March 1, 2024

Big Tech decides to "fix" email. God help us all.

As part of my role for a recent client, I have been adding to my email bona fides, which also means going to webinars involving personnel from Gmail and Yahoo. For those of you who do not make a living out of this channel, let me catch you up. I promise it will be more interesting than you might imagine. Also, really infuriating!

For the past few business cycles, Yahoogle has joined the previous party started by Apple to "improve" the email channel... but all of those improvements have more than a hint of sulfur to them, especially if you attempt to do, gasp, business in the channel. And since no one wants to stand up for professional emailers and even if they did, Yahoogle owns their channel and can and will do whatever they please... well, we're just all going to dance to their tune. 

Here's what the Big Tech Big Daddys have planned for us.

1) 1-click unsubscribe in the header. Hey, do you know how you've been able to find unsubscribe links for the vast majority of emails you've ever received in your entire life in the footer? And how, yeah, it might involve a little bit of scrolling, but it means that when you do break off contact with someone, you kind of had to mean to?

Well, that's not good enough anymore. Instead, we're going to have big fat UNSUBSCRIBE links (1 click, too! Heaven forbid you have to confirm anything) in the header area, so you can dramatically increase your chance of doing that by accident, especially on mobile devices. Don't you feel better now that Yahoogle is making your lives easier?

2) They are also going to look at "engagement data" (what data? Why, if we told you that, The Spammers Would Win!) to help determine what should go to the inbox. But since we've already systemically destroyed open as a metric with someone's idea of privacy (thanks, Apple! It's not as if we've been using that metric as a key performance indicator for the success of our emails for the last 30 years or anything), then we should go to click data. Not read rates, not multi-open, not multi-click, not long tail opens or long tail clicks... or maybe yes, who knows. Just clicks. 

Except those can be gamed by bots and bad actors, so probably not clicks either. Next year.

3) These moves have, of course, inspired other smaller email service providers to do the same. Because the not so dirty secret of email service providers is that they'd really love to not actually deliver any email, since that's an expense and all of the Kool Kidz went to social and SMS marketing years and years and years ago, because Email Is Dead and Old and never you mind that the metrics have never, ever actually followed that particular Naturalistic Fallacy.

So what does this make? A world in which unsubscribe rates and spam complaint is wildly more prone to misclick, creating a cycle in which less and less email gets to the inbox, which is, of course, what everyone says they want. And anyone with a rising rate (hint: people who use good dayparting, compelling subject lines, etc.) will run into more of this, and get dinged by these numbers, giving Yahoogle and its ilk all they need to... cut down on the number of emails that get to the inbox.

All without the end user ever likely knowing, because you like it when Big Tech solves "problems" for you, right? You can trust those people, and you'll never be able to get to Inbox Zero by yourself. (Um, I've been at Inbox Zero for my entire life. It's actually not that hard. Don't tell Yahoogle, they don't believe their users can actually do things for themselves.)

Or, TL/DR... literacy? Commerce? A channel where people have to read and write and think? Won't that get in the way of the other enshittification moves that Big Tech is giving you, especially now that the low interest rates dumb VC money has gone so far away, and they need to make more off you while providing less?

Now, look, I get it. AI means more spam. So does a campaign year. People are busy, they get too much mail, they are looking to claw back their off hours. But giving up your personal agency to Big Tech Daddys isn't the way to do that. They are just going to make it harder for legitimate pros, while actual spammers will just find a way, as they always have. 

Perhaps maybe ask pros, rather than just commit your usual hubris? Nah... what do emailers know? They work in email! Unlike... um...

Actually heard on a recent seminar: "We want to help inspire you to be better emailers." Oh, Thank You, Big Tech Big Daddy! We never thought of being better! (Side note: do you work with email pros that aren't constantly trying to get better? Usually by, I don't know, MATH, rather than hubris?)

Personally, it's all a win for M&AD; when the going gets tough, the tough go pro, and our bag of tricks has never been more full. But just because the world is better for this particular email marketing outfit does not mean it's better, or that Big Tech does not, in point of order, Suck. 

So very, very much.

Reach out if you need us! And enjoy those Big Tech layoffs! Maybe they'll inspire the laid off personell to be better!

Sunday, November 26, 2023

Top 10 things we've learned this year

Sorry it's been a while, but the reality of agency work is that the less you hear from us on this blog, the busier we are. Anyway, let's get to the good stuff, and if most of this seems email-centric, well, that's what we've been doing. You are what you eat.

1) Dayparting continues to adjust. 

One of the easiest ways to goose your positive metrics is to send your emails (a) when people will see them, and (b) when everyone else isn't. So if you've never tested your times, don't think that they vary from summer to winter seasonality, and aren't taking advantage of evening and weekend work for highly motivated readers, you are just leaving money on the table. Anymore from 5 to 20% of it, actually.

2) Your body copy probably never mattered, and it really doesn't now.

After a year of extensive testing for a steady client, I've learned that email service providers are not analyzing body copy and content. They are analyzing subject lines, list hygiene, frequency, complaint rates and so on -- but the stuff that your marketing team really sweats over and edits as if it matters? Not so much. Inboxing, open rates, spam filtering -- that's all on the subject line. Plan accordingly. 

3) Your total list isn't helping you.

That's because for most clients, attrition and obsolescence is going to make the unresponsive parts of your list an active detriment. Send to them, and your rates will go down, your sender reputation will tank, and you won't be able to reach the people who actually want to hear from you. Reactivating lapsed users is small beer, by the way -- worth doing, but it's not a game changer. Send only to active members and write the back end off.

4) Go organic, if your brand allows it.

Emojis used to seem like hack work or gimmickry, but since the world has gone away from them, ESPs see them as evidence of human activity again. Odd punctuation, a not quite polished quote copy, something that seems just a little bit off in a way that's more human and less AI? It's all going to work, because it's all going to seem like it's from a human, not a machine. The more tech we have, the more we want humanity.

5) File sizes will not be enforced.

Once upon a time in email, you had to optimize your images to the point of pain, limit animation cycles, worry about deployment times and worry a lot about ESPs gating heavy messages, to the point of always sending plain text test segments to make sure you weren't harming the campaign. Now? Yes, an animated image might take a second or two to pop in, and you shouldn't make your messages huge for the fun of it, but otherwise, don't sweat this. It does not matter.

6) Single KPI sucks, has always sucked, and will continue to suck. 

No marketing department will stand up in public and say that they only care about say, acquisition without conversion... but they will also inevitably prioritize one measurement over all others, because humanity wants to operate on works/does not work binaries, when reality is a continuum. Measure the top of the funnel, the middle, and the close -- or risk your business to myopic moves that will optimize one factor, often at the cost of your entire business.

7) A secret sauce metric: read/glance/skim.

A top client uses Pardot for email, which leaves a lot to be desired, but does have one winning feature: a pixel that loads and tells the marketer if the reader has viewed the email in question for less than 2 seconds (glance), two to eight (skim), and over eight (read). Like all metrics in email, take it with a shaker of salt, but when you compare emails against each other and one has a dramatically better read rate? That's a clue on the efficacy of your body content, formatting, and so on. Don't ignore clues.

8) Check the tail.

Many email analysts will look at metrics after a small period of time and never look again -- but that's another clue that is being ignored. Some emails, dayparts, offers, content and so on is going to inspire interaction for much longer after delivery. That could lead to content changes, different tactics, and so on. It will also impress your client.

9) Email continues to underuse some tech, and overuse others.

Want AI to write subject lines for you? Sure, that's easy (and dear God, just hire a good human already; tech that ends honest work should not be celebrated). How about dynamic content that matches the user's interests, or fits into larger marketing efforts with CTV, AR, QR codes, banners, direct mail and so on? Not so much. (Alas the poor QR code: it's just never really going to be a thing. Especially when a bad actor inevitably uses it to promote malware and it becomes a media firestorm. Inevitable, that.)

10) Painting outside the lines still works.

The single biggest win for one of my clients this year was with an unusual format (a 2 minute broadcast spot), featuring very different creative (a comedy skit) in an unexpected setting (a nation-wide college football broadcast). A steady diet of this would not work, and there was a lot of unique aspects to this that made it happen; the game was a blowout and social media decided the ad was a lot more fun to talk about than the game. Having this be a replicatable success seems unlikely, but it did a lot of good for the client, reinforced the brand, and energized the internal team. It also wasn't an all-in gamble. Consider the wisdom of all of this.

There's more, of course, but we need to save a few things for your next engagement with us. We look forward to working with you.

Sunday, June 25, 2023

On losing and loser clients

Ocean's Gate, Like Heaven's Gate
Many times in my decades of consulting, I've been directed to do a dumb thing by a client that's paying me to, well, steer them away from doing dumb things. (And yes, this post was inspired by a small group of obscenely wealthy people dying in the ocean, causing untold expense to find out what happened to them, all of which is money that could have been spent on so many, many better ideas.)

It's the single worst part of consulting, and the hardest judgment call you'll ever make. If your client hears your counsel and chooses to keep in their bad course, you are honor-bound to execute it to the best of your ability. All while knowing that, like a failed military coup in politics, this refusal to bow to expertise, math, etc., is just setting the stage for a deterioration of the relationship later. 

Most agencies lose 1 out of 4 clients every year, and most marketers don't stay at one gig for life, especially in aggressive categories. You acquire a past as being either too ineffective or too inflexible, the client acquires a past as being obstinate or resentful that you are telling them that they are wrong, and the seeds for a future separation are shown.

I'm not really sure how to solve for this, honestly. I've tried at most of my gigs to make it about the math, under the theory that math can solve a disagreement without emotion, and gives everyone a great way to walk it back from a bad decision. Test don't guess, and the only real way to fail is to not test, or at least, not test effectively.

But the trouble with this approach, and the reason why M&AD hasn't been able to elude the gravity that every other agency is subject to, is this: "the math" isn't binary. Which is an odd thing to say about math, but hear me out. It's never about a single key performance indicator. If it were, you could do your job forever and the only real issue is ducking AI or boredom, because your job is simple AF. 

But for most clients, if I drive leads to your business cheaply, you can say I've done my job -- or you can point to lifetime value and ROI and say I'm done nothing useful. I can say it's your site or your offer or your competition, and now we're back to the world of emotions, not math.

For clients like this, you win so long as you do the thing that makes them happy. You try to make yourself bulletproof by getting them to change the way they look at the world... but reality is that most people do not change the way they look at the world. And if they do change, it's not always for the better.

And that's when the job starts to lose its fun, its vitality, its candor and its color... and you start to think about the client that's going to replace this one. Which is also why agencies are never not looking for new business. Ping us accordingly.