My talk from the 2010 Cannes Lions.
For you startup geeks out there I guess you can say I’m giving advertising the “lean startup” treatment. For those of you with real jobs, I’ll be talking about how we can reduce risk, eliminate waste and increase the impact in campaigns where media fragmentization and hyper competition are significant factors.
3. I used to be one of you... my last gig was as a
Planner at Daddy and for a short time at CP+B,
after they acquired Daddy last year.
4. But I decided to give up what was essentially my
dream job to go work on another dream.
5. I’m now the CEO and founder of Burt, a software
company helping marketers to do amazing work.
6. Gustav Martner
Gustav von Sydow
I founded to Burt with Gustav Martner (currently
Exec. Creative Director at CP+B Europe) to adress
challenges we faced together on a daily basis.
7. Mid nineties. Fifteen years later.
Big bang launch Iterative improvements
Static content Rich applications
Archived Real-time
One size Personalized
One way Social
The founding idea was that we had seen the web
change drastically in the last 15 years...
8. Mid nineties. Fifteen years later.
Big bang launch Same, same.
Static content
Archived
One size
One way
But online advertising - at least on a large scale -
remained pretty much the same.
9. Sure there were promising technologies, but they
were complex to use so the tools and processes
that were used on large scale were antiquated.
10. Copybox Meme Machine Rich
The writing tool for Dead simple Campaign metrics
digital marketers. ad personalization. that matters.
www.copyboxapp.com www.mememachine.com www.richmetrics.com
So we set out to create a set of tools that would
make it easy for marketers to update how they
made and measured their marketing activities.
15. BANNERS &
VIDEO MOBILE
RICH MEDIA
Or more specifically, ”online display advertising”.
16. TODAY IS
NOT A
CASE STUDY.
Case studies are often outliers. What makes them
gripping often also makes them irrelevant.
17. ”YES. BUT DOES IT SCALE?”
Evaluating new opportunities in advertising, one
should always ask - ”does it scale?”
18. SCALE ACROSS AGENCIES.
You should be able to apply new technology/
methods across all types of agencies.
19. TECHNOLOGY IS RAW
MATERIAL, NOT AN IDEA.
And beware of when technology in itself is the
idea... great advertising technologies that scale
include ”color”, ”movement” etc.
25. We love talking about the future. Everyone’s a
”FRAMTIDEN”
futurist/expert these days.
26. It’s very much a don’t-miss-the-train kind of
mentality. All the time. Everywhere.
27. ”FIRST MOVER
ADVANTAGE”
WHO CREATED THE FIRST BANNER?
The concept of first mover advantage has been
largely discredited in general business, why are we
still believing it in advertising.
28. AT&T, 1994
Besides the amazing viral effect ;) of being part of
this deck, what have AT&T gotten in return for
pioneering the banner format?
29. TODAY IS
ABOUT...
WELL. TODAY.
Advertising is about getting massive reach today.
Not about predicting the future.
30. In this talk, we’re going to focus on improving
what we have and tweak how already do things in
already big channels. Solve existing problems.
31. NOT LIKE THE
OTHERS.
To sum up: this is not the average advertising
festival speech. When I arrived I got to looking into
the conference magazine...
38. HOW TO MARKET
A MARKETER:
1. WIN A BUNCH OF AWARDS.
AND/OR
2. ANNOUNCE THE END OF THE WORLD.
39. Cannes Lions combines both, which might explain
why it’s the premier event to land new business ;)
40. What I’m saying is that you can take it easy. Don’t
worry too much about the future.
41. OPTIMIZE
FOR NOW.
Great advertising connects with ”the now”.
Advertising is tactics - don’t be fooled even if the
word ”strategy” gets thrown around a lot ;)
42. THAT’S WHAT WE
GET PAID TO DO.
We get paid to get results now. Or at least in the
next quarter. Very few good advertising activities
don’t achieve tangible result within 90 days.
43. YOU CAN
DO IT.
What I’m talking about applies to everyone.
Creatives, planners, account, devs, clients etc.
And you can start using the ideas on monday.
46. THE CREATIVE PROCESS:
IDEA PRODUCE
Though the name implies something super
dynamic and exciting, the creative process is a
very linear process.
47. ME = BIG FAN BTW.
”FINAL DELIVERY”
Even the best practices (I recommend checking
them out btw) emphasize the ”final delivery”.
48. Traditional Product Development
Unit of Progress: Advance to Next Stage
Waterfall
Requirements
Specification
Design
Problem: known
Solution: known
Implementation
Verification
Maintenance
Though every agency use different words to
descibe their unique, amazing process, they’re all
based on the same classic manufacturing model.
50. ”LEAN PRODUCTION”
Doing more with less.
But manufacturing has actually moved on to a
model for creating more value using less work,
and so should we.
51. ”Lean production” originated from Toyota,
emphasizing to eliminate waste by removing all
activities that does not create value.
58. 1. $150B IS WASTED EVERY YEAR.
2. 19% OF ALL ADVERTISING FAILS OUTRIGHT.
3. 67% COULD GET SIGNIFICANT IMPROVEMENT
WITHOUT ADDITIONAL SPEND.
And there’s a lot of waste, risk and potential for
improvement in online advertising.
Source: What Sticks
59. ”LEAN ADVERTISING”?
The ideas have been successful in everything from
cars to software startups (respect to Eric Ries and
Steve Blank!!!). How do they apply to advertising?
60. IDEA PRODUCE MEASURE
Well, the first step is to add a measurement
component to the creative process.
61. IDEA PRODUCE MEASURE
CAMPAIGN 1
LEARN
IDEA PRODUCE ME
CAMPAIGN 2
And then use this to learn for use in your next
campaign.
63. Which is not a new idea of course. In advertising,
it was pioneered by David Ogilvy in the 50s.
64. ”I never skate to where the puck is,
I always skate to where to puck is going to be.”
65. The problem is that the Ogilvy way of doing things
assumes a rather static media landscape, but it’s
now changing faster and faster.
BLABLABLABLA... COMPETITION, CHANGE ETC.
66. 30 sec. 0 sec? 2 sec? 13 sec?
From start to nish, full screen. And no idea when it starts...
The level of predictability makes it exponentially
more complex to predict what will work.
67. Most ads aren’t visible long
enough to make an impact
+10 ads share the
same page
30% of all ads are placed
where users can’t see
There’s just no such thing as average anymore.
68. WHO’S RESPONSIBLE?
MEDIA? CREATIVE? DIGITAL?
PRODUCTION? RESEARCH?
PUBLISHER? AD NETWORK?
THE AD NETWORK’S AD
NETWORK’S AD NETWORK?
AND WHICH PERSON?
69. BLABLABLABLA...
ROI.... MEGA DEEP
INSIGHTS OF INTRINSIC
HABITS... SPECULATION...
GUESSWORK...
BLABLABLA....
”EXPERT”
The way we usually solve challenges in advertising
is to appoint experts. But experts doesn’t scale.
72. ”AGILE ADVERTISING”
The process of being quick
learn, adapt and improve.
So ”Lean advertising” doesn’t quite describe the
high velocity dimension of what I’m getting at. So
”agile advertising” seems like a better phrase.
74. IDEA PRODUCE MEASURE
IDEA 2 LEARN
IDEA 3
IDEA 4
IDEA 5
But instead of waiting to measure and apply the
IDEA 6
learnings on the next campaigns, we change and
IDEA 7 campaigns.
optimize the existing
IDEA 8
77. WHAT?
IDEA PRODUCE MEASURE
IDEA 2 LEARN
IDEA 3
IDEA 4
IDEA 5
The first question is naturally:
IDEA 6
What do we7
IDEA measure?
IDEA 8
78. Metrics is the number one pain for
display advertisers.
Lack of proper metrics Ad quality is too low Other
Source: Forrester
79. IS ”ROI” THE PROBLEM?
There is usually a big focus on ”ROI” when metrics
comes up. But is that really what hands-on
marketers should focus on?
80. INVESTMENT PROFIT
$ $$$
{ ”KPIs”
Our job is to understand the underlying metrics
driving profit, rather than auditing the ROI.
81. INVESTMENT RETURN
FOCUS ON KPIs, $$$
$
NOT ROI.
{ ”KPIs”
82. EXAMPLES OF GOOD KPIs:
AUDIENCE REACH AND RELEVANCE - PEOPLE
”SOFT METRICS” - BRANDING
DIRECT USER BEHAVIOR - CLICK STREAM
SALES - NOT THE SAME AS PROFIT
83. THE THREE A’S OF
METRICS:
1. ACCESSIBLE - THINK ”YOU TUBE VIEWS”
2. ACTIONABLE - OTHERWISE SKIP IT
3. AUDITABLE - METRICS ARE PEOPLE TOO
84. BENCHMARK.
The first step is to benchmark what you’re up to,
to get a decent idea of where you stand.
85. Source: DoubleClick DART for Advertisers, U.S. advertisers, a cross section of major ad sizes only, January – December 2008
Figure 4 Click-through Rate by Industry Vertical
Auto 0.12%
B2B 0.08%
CPG 0.09%
Financial Services 0.06%
Media/Entertainment 0.11%
Retail 0.10%
Tech 0.10%
Telecom 0.08%
Travel 0.09%
Wellness 0.09%
0.05% 0.10% 0.15% 0.20% 0.25% 0.30%
Source: DoubleClick DART for Advertisers, U.S. advertisers, a cross section of major industry verticals only, January – December 2008
But benchmarks are a bit tricky. They vary greatly
by advertiser category.
3
86. Overall Click-through Rate (CTR)
Figure 3 Click-through Rate by Creative Sizes
Half Page Ad (300x600) 0.22%
Wide Skyscraper (160x600) 0.10%
Vertical Rectangle (240x400) 0.33%
Large Rectangle (336x280) 0.12%
Medium Rectangle (300x250) 0.12%
Skyscraper (120x600) 0.06%
Leaderboard (728x90) 0.08%
Square Pop-Up (250x250) 0.18%
3:1 Rectangle (300x100) 0.07%
Vertical Banner (120x240) 0.05%
Full Banner (468x60) 0.05%
Rectangle (180x150) 0.10%
0.00% 0.05% 0.10% 0.15% 0.20% 0.25% 0.30% 0.35% 0.40%
Source: DoubleClick DART for Advertisers, U.S. advertisers, a cross section of major ad sizes only, January – December 2008
Or creative sizes for that matter.
87. INDUSTRY, FORMAT,
CONTENT, TIMING,
COUNTRY, DEVICE ETC.
And a bunch of other stuff.
88. INDUSTRY, FORMAT,
AVERAGES ARE (ALMOST)
CONTENT, TIMING,
MEANINGLESS.
COUNTRY, DEVICE ETC.
Besides using them as a sanity check, averages
are pretty hopeless.
90. SELF DIAGNOSE.
Instead of just comparing yourself with others, you
should become good at helping yourself out.
91. A
VS.
B
Small tweaks can have an enormous impact.
Source: What Sticks
92. Intends to buy Colgate
25 %
20 %
15 %
10 % +10x
5 %
0 %
Creative A Creative B
Creative execution is by far the biggest lever for
creating effective online advertising.
93. BANNER BEST PRACTICES:
1. MEDIA/CREATIVE INTEGRATION BIGGEST DRIVER.
2. SHOW THE LOGO. ALL THE TIME.
3. SHOW THE PRODUCT. IF APPLICABLE.
4. STATIC OUTPERFORMS RICH. THINK PRINT, NOT TV.
5. DIRECT. SIMPLE. OBVIOUS.
94. ”SPLIT TESTING.”
Once you have the fundamentals right, it’s time to
start learning from a live campaign. And the best
way is to compare, not with others, but yourself.
95. ZYNGA AND THE USE OF
DISCOUNT PRE-TESTING.
Social gaming and other viral products have
pioneered rapid testing and optimization.
96. Intends to buy Colgate
25 %
20 %
15 %
10 % +10x
5 %
0 %
Creative A Creative B
Split testing allows you to spot differences early.
97. Creative A, performance by site Creative B, performance by site
25 % 25 %
20 % 20 %
15 % 15 %
10 % 10 %
But the problem is that - as we touched on earlier
5 % 5 %
+10x
- different executions work different under different
0 % 0 %
Site A
circumstances. Site A
Site B Site B
98. A + Site A, performance by Geography Creative A + Site B, performance by Geography
75 %
60 %
45 %
30 %
+10x
And goes deeper...
15 %
0 %
Site A Site B Site A Site B
99. You end up with a tree structure with all these
different variations and outcomes.
100. 1%
0,75%
0,5%
+10x
0,25%
0%
“Worst” Average “Best” “Bestest”
By using the best combination for each exposure
you can create enormous ”lifts”.
101. Which has been well known, and talked about
forever. But for different reasons, ”one-to-one”
marketing is not deployed on a very large scale.
103. HOW?
IDEA PRODUCE MEASURE
IDEA 2 LEARN
IDEA 3
IDEA 4
IDEA 5
And then there’s that little detail of applying the
IDEA 6
ideas and lessons learned in practice...
IDEA 7
IDEA 8
105. SIMPLICITY = FASTER TIME TO MARKET
EASIER TO TEST
BONUS FEATURE: LESS TO UNDERSTAND FOR CONSUMERS
106. Start off with one simple execution of an idea. In a
banner for instance. One format is enough.
107. WHERE’S THE LOGO? PACK SHOT EARLIER?
ANIMATION TOO MUCH
NECESSARY? MESSAGING?
Self diagnose and make that if you’re breaking any
rules and best practices, you should be aware of
what rules you’re breaking, and why.
108. A
B
Create variations for split testing. At this stage I
recommend doing widely different variations. In
some cases completely different ideas.
109. A B
C D
Buy (or convince the publisher to give you) some
media and run a quick and dirty split test, to see
how consumers respond to your variations.
110. Click-thrus
1,5 %
1,2 %
0,9 %
0,6 %
0,3 %
0 %
Creative A Creative B Creative C Creative D
Choose a proper metric (which is a separate
workshop, I’m afraid) and eliminate the waste.
Focus your efforts in the stuff that seems to work.
111. START SPENDING
ON MORE MEDIA,
FORMATS ETC.
Once you’ve condensed your basic idea, you can
start expanding on the with more variations,
formats and progressivly increase media spend.
112. OPTIMIZE.
Monitor and switch on automatic optimization
where it makes sense to personalize the
messages. With the right tool you can optimize
against any kind of effect.
113. BENCHMARK.
Once you’re over and done with, do a quick
followup and benchmark your efforts. Call in the
audit people if that’s your thing. Wait until the next
campaign to make deep analysis.
114. IDEA PRODUCE
IDEA n
MEASURE
LEARN
x4
For those of you paying attention, ”agile
advertising” puts four feedback loops into play.
115. THE FOUR FEEDBACK LOOPS:
LOOP 1:
SELF-DIAGNOSE
LOOP 4:
LOOP 2:
CROSS-
SPLIT TEST
CAMPAIGN
LOOP 3:
OPTIMIZATION
116. WRAPPING
THINGS UP:
1. THE TRADITIONAL CREATIVE PROCESS IS RISKY.
2. ”AGILE ADVERTISING” LOWERS RISK AND
INCREASES OUTPUT BY TESTING IDEAS EARLY.
3. THE KEY TO SUCCESS IS IMPLEMENTING THE
FOUR FEEDBACK LOOPS.
117. SPEED.
The faster you learn the sooner people will pay
attention to you, and not your competitors.
118. Usability testing
PM software Funnel analysis
Scrum Cohort analysis
JIT scalability Real-time alerting
MVP Predicitive monitoring
IDEA PRODUCE MEASURE
Flow optimization
Working conditions IDEA n LEARN
Time management
5 whys
Creative leadership
Customer archetypes
Smoke tests
Ways to increase your velocity thru the loop.
Thanks: Eric Ries and Steve Blank
119. LEARN.
It’s really a simple idea. Spend time testing and
learning, not speculating.
120. YOU CAN
DO IT.
And as I said, anyone can do this. It works better
the more people are involved, but ”agile
advertising” allows even just one person to get a
big influence on the end result.