สำเนาบทสนทนา 🔗
WEBVTT
00:00:00.000 --> 00:00:04.951
Welcome to Shallow DeepDive, where we take
a deep-ish dive into the shallow world of
00:00:05.031 --> 00:00:07.048
branding, advertising and creative.
00:00:07.128 --> 00:00:07.946
I'm Joe Qualls.
00:00:08.026 --> 00:00:12.857
Today, we're covering the quiet cost of
building a brand that can't offend anyone.
00:00:12.937 --> 00:00:17.829
Somewhere in a marketing team right now, a
campaign is being run through a new kind
00:00:17.909 --> 00:00:18.620
of check.
00:00:18.700 --> 00:00:20.657
Not "is it on brief" or "is it any
00:00:20.737 --> 00:00:24.969
" A different question altogether: how
much backlash could this attract?
00:00:25.049 --> 00:00:30.000
There are now models for it — ways of
scoring a piece of work on how likely it
00:00:30.080 --> 00:00:34.852
draw a coordinated pile-on, an organised
wave of complaints, a spell of political
00:00:34.932 --> 00:00:35.644
heat.
00:00:35.724 --> 00:00:38.219
"Backlash exposure," people are calling
it.
00:00:38.299 --> 00:00:41.813
Anything that scores high gets softened,
or quietly shelved.
00:00:41.893 --> 00:00:44.688
It's a completely understandable thing to
build.
00:00:44.768 --> 00:00:49.420
It's also, we'd argue, reshaping brands in
a direction most of them would never
00:00:49.500 --> 00:00:50.498
choose on purpose.
00:00:50.578 --> 00:00:54.211
Let's be fair to it first, because the
fear behind it is real.
00:00:54.291 --> 00:00:57.865
The last few years handed everyone the
same cautionary tales.
00:00:57.945 --> 00:01:02.776
A brand makes one decision, some corner of
the internet decides to make an example
00:01:02.856 --> 00:01:07.388
of it, and weeks of a company's life
vanish into a crisis nobody planned for.
00:01:07.468 --> 00:01:11.401
You can still name the big one from three
years ago without trying.
00:01:11.481 --> 00:01:16.013
Pride this year made the pattern plain —
brands that had spent a decade being
00:01:16.093 --> 00:01:20.026
visible went quiet, trimmed the
merchandise, softened the language.
00:01:20.106 --> 00:01:24.638
Not because they'd changed their minds,
but because they'd run the numbers on
00:01:24.718 --> 00:01:26.974
exposure and didn't like what they saw.
00:01:27.054 --> 00:01:31.646
When the downside is that vivid, of course
you build a tool to spot it coming.
00:01:31.726 --> 00:01:33.982
Nobody wants to be the next case study.
00:01:34.062 --> 00:01:37.156
But here's the thing about a
backlash-exposure model.
00:01:37.236 --> 00:01:42.068
It is very good at pricing one kind of
cost — the loud, visible,
00:01:42.148 --> 00:01:42.859
kind.
00:01:42.939 --> 00:01:43.651
The hashtag.
00:01:43.731 --> 00:01:44.669
The angry thread.
00:01:44.749 --> 00:01:46.226
The phone that won't stop.
00:01:46.306 --> 00:01:51.257
What it cannot see, because nothing on the
dashboard is shaped like it, is the other
00:01:51.337 --> 00:01:52.049
cost.
00:01:52.129 --> 00:01:52.841
The slow one.
00:01:52.921 --> 00:01:55.775
The cost of saying nothing that anyone
remembers.
00:01:55.855 --> 00:01:59.429
There's no alert that fires when a brand
becomes forgettable.
00:01:59.509 --> 00:02:01.945
No spike, no thread, no emergency meeting.
00:02:02.025 --> 00:02:06.736
The brand just gradually stops mattering
to people, quietly, over years — and no
00:02:06.816 --> 00:02:11.288
model flags it, because the absence of a
reaction looks exactly like safety.
00:02:11.368 --> 00:02:16.200
And this is the part worth sitting with,
because it isn't obvious until you say it
00:02:16.280 --> 00:02:16.991
out loud.
00:02:17.071 --> 00:02:22.023
The thing the model scores as risk, and
the thing that makes a brand mean
00:02:22.103 --> 00:02:23.400
are the same substance.
00:02:23.480 --> 00:02:24.778
They're not neighbours.
00:02:24.858 --> 00:02:27.114
They're one thing, seen from two sides.
00:02:27.194 --> 00:02:32.085
A brand becomes memorable by choosing — a
point of view, an audience it's genuinely
00:02:32.165 --> 00:02:36.098
for, a way of seeing its own category that
not everyone will share.
00:02:36.178 --> 00:02:39.272
And choosing is precisely what registers
as exposure.
00:02:39.352 --> 00:02:44.244
The moment you stand somewhere, you become
visible from that somewhere, which means
00:02:44.324 --> 00:02:46.879
someone, somewhere, is now able to
disagree.
00:02:46.959 --> 00:02:51.791
So when you tune the exposure down toward
zero, you're not only removing the risk.
00:02:51.871 --> 00:02:53.528
You're removing the choosing.
00:02:53.608 --> 00:02:55.384
And the choosing was the brand.
00:02:55.464 --> 00:03:00.056
There's a useful distinction hiding in
here, between a position and a posture.
00:03:00.136 --> 00:03:05.028
A position is something you actually
believe — about your work, your customer,
00:03:05.108 --> 00:03:07.783
corner of the market — and it costs
something.
00:03:07.863 --> 00:03:12.035
It's for some people and, by honest
implication, a bit less for others.
00:03:12.115 --> 00:03:16.947
A posture is built to look like a position
while costing nothing: warm, agreeable,
00:03:17.027 --> 00:03:19.762
engineered to reach everyone and offend no
one.
00:03:19.842 --> 00:03:24.613
The trouble is that people can feel the
difference, even when they can't name it.
00:03:24.693 --> 00:03:27.668
A position lands, because someone clearly
meant it.
00:03:27.748 --> 00:03:32.220
A posture slides off, because underneath
the nice words there's no one home.
00:03:32.300 --> 00:03:37.191
And a brand made entirely of postures —
every edge sanded down, every claim
00:03:37.271 --> 00:03:40.605
every line made safe — reaches everyone
and moves no one.
00:03:40.685 --> 00:03:41.624
It's in the room.
00:03:41.704 --> 00:03:43.421
Nobody remembers it was there.
00:03:43.501 --> 00:03:48.452
What makes this genuinely tricky is that
the safe choice always looks correct in
00:03:48.532 --> 00:03:49.243
moment.
00:03:49.323 --> 00:03:54.275
Every single time, softening the line is
the
Load full transcript
lower-risk decision on the day you
00:03:54.355 --> 00:03:55.066
it.
00:03:55.146 --> 00:03:56.264
Nothing bad happens.
00:03:56.344 --> 00:03:57.582
The meeting ends well.
00:03:57.662 --> 00:04:02.014
And that's the trap — the cost never
arrives as an event, so it never gets
00:04:02.094 --> 00:04:04.530
attributed to the decision that caused it.
00:04:04.610 --> 00:04:09.381
A hundred small sandings, each sensible on
its own, add up to a brand that no one
00:04:09.461 --> 00:04:10.519
can quite describe.
00:04:10.599 --> 00:04:15.371
But you can't point to the Tuesday it went
wrong, because it didn't go wrong on a
00:04:15.451 --> 00:04:16.162
Tuesday.
00:04:16.242 --> 00:04:18.438
It went a little wrong on all of them.
00:04:18.518 --> 00:04:23.470
Now, we're not making the opposite
mistake, which is just as common and a
00:04:23.550 --> 00:04:24.261
embarrassing.
00:04:24.341 --> 00:04:28.274
Having a position is not the same as
picking a fight for attention.
00:04:28.354 --> 00:04:33.305
There's a whole genre of brand that goes
looking for controversy on purpose,
00:04:33.385 --> 00:04:36.600
a cause it hasn't earned, and calls the
resulting noise
00:04:36.680 --> 00:04:38.456
" That's not a position either.
00:04:38.536 --> 00:04:42.170
It's a posture in a leather jacket, and it
ages just as badly.
00:04:42.250 --> 00:04:44.446
The point was never to be provocative.
00:04:44.526 --> 00:04:49.297
It's to be genuine — to know what you
actually think about your customer and
00:04:49.377 --> 00:04:54.089
craft, and to say it plainly, edges and
all, whether or not it happens to be the
00:04:54.169 --> 00:04:55.946
safe thing to say this quarter.
00:04:56.026 --> 00:04:59.480
So what do you do with a backlash model,
if you've got one?
00:04:59.560 --> 00:05:00.438
Use it honestly.
00:05:00.518 --> 00:05:05.409
Let it tell you the thing it's good at
telling you — where the loud, near-term
00:05:05.489 --> 00:05:10.321
actually sits — and then make it earn its
keep by putting the other number next to
00:05:10.401 --> 00:05:11.112
it.
00:05:11.192 --> 00:05:16.143
Not just "how exposed are we if we say
this," but "what does it cost us, over
00:05:16.223 --> 00:05:17.162
to say nothing at
00:05:17.242 --> 00:05:21.714
" The second number is harder to
calculate, and it's usually the larger
00:05:21.794 --> 00:05:26.445
If a piece of work scores high on exposure
only because it genuinely stands for
00:05:26.525 --> 00:05:29.740
something, that is not, by itself, a
reason to kill it.
00:05:29.820 --> 00:05:34.352
Sometimes that reading is just confirming
you've finally got a point of view.
00:05:34.432 --> 00:05:37.227
The job was never to drive the exposure to
zero.
00:05:37.307 --> 00:05:41.120
Zero exposure and zero meaning have a
habit of arriving together.
00:05:41.200 --> 00:05:46.151
Because the safest brand in the room is a
real achievement, when you think about it.
00:05:46.231 --> 00:05:47.409
Nobody's angry at it.
00:05:47.489 --> 00:05:48.846
Nobody's threatening it.
00:05:48.926 --> 00:05:50.883
Nobody's writing threads about it.
00:05:50.963 --> 00:05:54.177
It has successfully avoided every downside
on the list.
00:05:54.257 --> 00:05:58.969
It has also, without anyone quite deciding
to, avoided being the reason a single
00:05:59.049 --> 00:06:01.065
person chose it over anything else.
00:06:01.145 --> 00:06:04.419
Safety, past a certain point, isn't the
absence of risk.
00:06:04.499 --> 00:06:06.635
It's the absence of a reason to care.
00:06:06.715 --> 00:06:10.469
And that's the one exposure no model on
the market is measuring.
00:06:10.549 --> 00:06:13.763
Thank you for indulging us in today's
Shallow DeepDive.
00:06:13.843 --> 00:06:16.159
We hope you'll join us for the next one.
00:06:16.239 --> 00:06:17.476
Until then, take care.