สำเนาบทสนทนา 🔗
WEBVTT
00:00:00.000 --> 00:00:04.837
Welcome to Shallow DeepDive, where we take
a deep-ish dive into the shallow world of
00:00:04.917 --> 00:00:06.886
branding, advertising and creative.
00:00:06.966 --> 00:00:07.764
I'm Joe Qualls.
00:00:07.844 --> 00:00:12.563
Today, we're covering the word Steve Jobs
wouldn't let Apple say, and why the most
00:00:12.643 --> 00:00:15.900
loyal customers are made on the day
something goes wrong.
00:00:15.980 --> 00:00:20.817
This week Ron Johnson, the man who built
the Apple Store, published a short piece
00:00:20.897 --> 00:00:23.802
Fast Company ahead of his new book, Shop
Different.
00:00:23.882 --> 00:00:27.431
Most of the attention will go to the
anecdote in the headline.
00:00:27.511 --> 00:00:32.231
We want to spend a minute on that, and
most of our time on a number buried
00:00:32.311 --> 00:00:36.855
down, because the number is the part that
should change how a business behaves.
00:00:36.935 --> 00:00:37.967
The anecdote first.
00:00:38.047 --> 00:00:42.825
Around 2001, Jobs asked Johnson to put the
stores' service philosophy in four words
00:00:42.905 --> 00:00:43.621
or fewer.
00:00:43.701 --> 00:00:46.138
Johnson came back with two: love
customers.
00:00:46.218 --> 00:00:50.645
He meant it in the old sense, love as
working for the good of another person.
00:00:50.725 --> 00:00:51.640
Jobs pushed back.
00:00:51.720 --> 00:00:53.981
Cupertino wasn't a place that said love.
00:00:54.061 --> 00:00:55.737
Could he use a different word?
00:00:55.817 --> 00:01:00.596
Johnson landed on "enrich lives," and
built a set of steps the staff could
00:01:00.676 --> 00:01:05.454
take: a warm welcome, questions to
understand what the person needed,
00:01:05.534 --> 00:01:08.907
the concern they hadn't said out loud, and
a fond farewell.
00:01:08.987 --> 00:01:11.541
There's a small brand lesson in that
already.
00:01:11.621 --> 00:01:13.356
Jobs didn't object to the idea.
00:01:13.436 --> 00:01:18.039
He objected to the word, because it had to
be said by thousands of people, every
00:01:18.119 --> 00:01:20.321
day, without anyone rolling their eyes.
00:01:20.401 --> 00:01:24.068
A service philosophy is a piece of writing
with a very hard job.
00:01:24.148 --> 00:01:28.926
It has to be short enough to remember,
plain enough to act on, and true enough
00:01:29.006 --> 00:01:31.443
people will say it without feeling
foolish.
00:01:31.523 --> 00:01:34.896
"Love customers" was right, and unsayable
in that building.
00:01:34.976 --> 00:01:38.643
"Enrich lives" was sayable, and still
pointed at the same thing.
00:01:38.723 --> 00:01:41.745
Most company values fail at least one of
those tests.
00:01:41.825 --> 00:01:46.486
They're too long to remember, too vague to
act on, or too grand for anyone to say
00:01:46.566 --> 00:01:47.715
with a straight face.
00:01:47.795 --> 00:01:48.593
Now the number.
00:01:48.673 --> 00:01:53.334
In 2007, Apple Retail was one of the first
retailers to adopt Net Promoter Score,
00:01:53.414 --> 00:01:57.139
the question that asks how likely you are
to recommend something.
00:01:57.219 --> 00:01:59.539
The early results were what you'd expect.
00:01:59.619 --> 00:02:01.471
Good experiences got recommended.
00:02:01.551 --> 00:02:02.407
Bad ones didn't.
00:02:02.487 --> 00:02:05.919
Then Johnson started doing something that
wasn't a strategy.
00:02:05.999 --> 00:02:07.031
It just felt right.
00:02:07.111 --> 00:02:11.890
When a customer reported a bad experience,
he'd write back to apologise and copy in
00:02:11.970 --> 00:02:12.885
the store leader.
00:02:12.965 --> 00:02:17.802
The store leaders followed up and fixed
the problem, and because the boss was on
00:02:17.882 --> 00:02:19.733
email, they did it straight away.
00:02:19.813 --> 00:02:24.416
The customers whose bad experience was
fixed quickly were, by Johnson's account,
00:02:24.496 --> 00:02:29.099
seven times as likely to recommend Apple
as customers who simply had a good one.
00:02:29.179 --> 00:02:30.679
Sit with that for a second.
00:02:30.759 --> 00:02:33.840
A fast fix beat a flawless visit, and not
by a little.
00:02:33.920 --> 00:02:38.698
Johnson's way of putting it is that a
great experience makes customers happy,
00:02:38.778 --> 00:02:40.688
resolved failure makes them loyal.
00:02:40.768 --> 00:02:45.137
Those are different outcomes, and most
businesses only budget for the first.
00:02:45.217 --> 00:02:47.185
We'd hold the exact figure loosely.
00:02:47.265 --> 00:02:51.868
It's one company's internal read, recalled
twenty years later in the run-up to a
00:02:51.948 --> 00:02:56.726
book launch, and the research on what's
called the service recovery paradox is
00:02:56.806 --> 00:02:58.834
mixed than a single number suggests.
00:02:58.914 --> 00:03:03.458
Recovery doesn't always beat getting it
right, and nobody should start breaking
00:03:03.538 --> 00:03:04.511
things on purpose.
00:03:04.591 --> 00:03:09.370
But the direction matches everything we've
seen, and the reason is simple enough to
00:03:09.450 --> 00:03:10.165
trust.
00:03:10.245 --> 00:03:13.736
A smooth experience tells a customer that
your process works.
00:03:13.816 --> 00:03:18.653
A problem fixed fast, by someone who
clearly had the authority to fix it, tells
00:03:18.733 --> 00:03:19.448
that you do.
00:03:19.528 --> 00:03:22.258
That's information they can't get any
other way.
00:03:22.338 --> 00:03:26.765
You only find out what a business is
actually like when something goes wrong.
00:03:26.845 --> 00:03:28.755
This is the part most brands skip.
00:03:28.835 --> 00:03:33.497
They spend real money on the good path:
the store, the website, the unboxing, the
00:03:33.577 --> 00:03:34.550
onboarding emails.
00:03:34.630 --> 00:03:36.891
They design the moment everything works.
00:03:36.971 --> 00:03:41.633
The moment something breaks gets handed to
a policy, a queue, a form, a script of
00:03:41.713 --> 00:03:46.491
approved phrases, and a staff member who
is allowed to apologise and not
Load full transcript
allowed to
00:03:46.571 --> 00:03:47.287
do anything.
00:03:47.367 --> 00:03:51.852
The complaint is the one moment a customer
is paying full attention to how you
00:03:51.932 --> 00:03:52.648
behave.
00:03:52.728 --> 00:03:57.565
It's the most carefully read page of your
brand, and most businesses have left it to
00:03:57.645 --> 00:03:59.496
whoever wrote the returns policy.
00:03:59.576 --> 00:04:02.891
I learned a version of this at my agency
back in New York.
00:04:02.971 --> 00:04:07.457
We built a full takeover campaign for a
residential development in New Jersey:
00:04:07.537 --> 00:04:09.213
digital, out-of-home, the lot.
00:04:09.293 --> 00:04:13.896
The whole thing was space-themed, and one
of the subway and NJ Transit car wraps
00:04:13.976 --> 00:04:16.939
carried a modified version of a line from
Star Trek.
00:04:17.019 --> 00:04:20.334
CBS, which owns Star Trek, sent a cease
and desist letter.
00:04:20.414 --> 00:04:24.900
Our legal advice was that we could fight
it, because the line had been changed
00:04:24.980 --> 00:04:25.696
enough.
00:04:25.776 --> 00:04:30.378
It was also that CBS was known for
dragging these things out, and had far
00:04:30.458 --> 00:04:31.549
pockets than we did.
00:04:31.629 --> 00:04:32.837
So we didn't fight it.
00:04:32.917 --> 00:04:37.754
We reprinted the posters at our own
expense with one of the alternate lines
00:04:37.834 --> 00:04:40.036
campaign, and swapped them out quietly.
00:04:40.116 --> 00:04:43.314
What the client remembered afterwards
wasn't the letter.
00:04:43.394 --> 00:04:47.704
It was how quickly it was handled, and
that it never turned into bad press.
00:04:47.784 --> 00:04:51.099
That's the seven-times effect, at the
scale of one client.
00:04:51.179 --> 00:04:55.782
Johnson names the condition underneath all
this, and it's the uncomfortable one.
00:04:55.862 --> 00:04:59.294
None of it works if your people are afraid
to make mistakes.
00:04:59.374 --> 00:05:01.577
His example is from Apple SoHo in 2003.
00:05:01.657 --> 00:05:05.264
An author came in with a broken Windows PC
and a book deadline.
00:05:05.344 --> 00:05:09.713
A specialist working from a sales script
would have sold her a new computer.
00:05:09.793 --> 00:05:14.513
This one asked whether it would be easier
to finish the book on one of the store's
00:05:14.593 --> 00:05:16.444
machines while hers was repaired.
00:05:16.524 --> 00:05:18.434
She came in every day for a month.
00:05:18.514 --> 00:05:21.888
There was no policy for that, and there
couldn't have been.
00:05:21.968 --> 00:05:26.746
It happened because the specialist
understood what the store was for well
00:05:26.826 --> 00:05:27.542
improvise.
00:05:27.622 --> 00:05:32.225
The same freedom produced Personal Setup,
which started with one store leader in
00:05:32.305 --> 00:05:36.966
Minnesota noticing that customers wanted
their new computers set up for them, and
00:05:37.046 --> 00:05:38.371
ended up in every store.
00:05:38.451 --> 00:05:42.644
Which puts this in a slightly awkward
conversation with our last episode.
00:05:42.724 --> 00:05:47.443
There, we argued Coca-Cola was right to
move its brand rules out of a PDF and into
00:05:47.523 --> 00:05:52.302
the design tools, so the rules get
followed without anyone having to remember
00:05:52.382 --> 00:05:53.472
We still think that.
00:05:53.552 --> 00:05:57.277
The two ideas fit together once you notice
what each one governs.
00:05:57.357 --> 00:06:02.135
The identity should be locked: the colour,
the type, the mark, the things that must
00:06:02.215 --> 00:06:03.599
look the same everywhere.
00:06:03.679 --> 00:06:08.457
The behaviour can't be locked, because the
moments that matter most are the ones no
00:06:08.537 --> 00:06:09.452
rule anticipated.
00:06:09.532 --> 00:06:12.145
Rules for the things that should never
change.
00:06:12.225 --> 00:06:16.710
Judgement, and permission to use it, for
everything that happens in front of a
00:06:16.790 --> 00:06:17.506
customer.
00:06:17.586 --> 00:06:21.018
So, for a business that isn't Apple, three
practical things.
00:06:21.098 --> 00:06:25.818
First, write your service idea in four
words or fewer, and test it by whether
00:06:25.898 --> 00:06:28.276
newest staff member would say it out loud.
00:06:28.356 --> 00:06:31.495
Second, look at what happens after a
complaint arrives.
00:06:31.575 --> 00:06:36.237
Count the hours before a person with the
authority to fix it is involved, and cut
00:06:36.317 --> 00:06:40.158
that number, because speed is most of what
the customer is judging.
00:06:40.238 --> 00:06:45.017
Third, give the people closest to the
customer something they can do without
00:06:45.097 --> 00:06:49.699
permission, whether that's a refund, a
replacement or an hour of their time, and
00:06:49.779 --> 00:06:54.499
make it clear that using it and getting it
slightly wrong is better than not using
00:06:54.579 --> 00:06:55.295
it at all.
00:06:55.375 --> 00:06:58.456
Johnson ends his piece by saying love
creates loyalty.
00:06:58.536 --> 00:07:02.377
We'd put it in plainer terms, which seems
fitting, given the story.
00:07:02.457 --> 00:07:06.475
Customers stay because they saw what you
did on the day it went wrong.
00:07:06.555 --> 00:07:09.694
Thank you for indulging us in today's
Shallow DeepDive.
00:07:09.774 --> 00:07:12.035
We hope you'll join us for the next one.
00:07:12.115 --> 00:07:13.323
Until then, take care.