Why So Many Ads? An Introduction To Live Creative Optimisation
- 2. Why do we need
so many ads?
An Introduction to Live
Creative Optimisation
- 3. 3
Content
Sharing valuable insights into how to make live optimisation work for newbies
and pros:
1. What is Live Optimisation?
2. Understanding the landscape
3. Core Principles
4. Benefits
5. How to get started
- 5. Matt Green
Director, Global Media Practice
World Federation of Advertisers
5
Ads used to work like this
This is how ads used to work - from basic “one big idea” with
static delivered ads all the way to today’s access to real-time
data and to the power of AI.
- 6. 6
The Live Optimisation Model
Our live optimisation model allows brands to continually adapt to
feedback from end customers, improving their ads in-flight. This
multiplies creative and delivers better targeting and more effective
ads.
- 8. A little bit
OF HISTORY
8
8
8
2000-2010s:
Uses behavioural data / Cross channel
Programmatic introduced
Mid 2000s:
Dynamic Creative Optimisation appears
Real time data introduced
1990s to Early 2000:
Basic Banners + New Data
8
2010s - Present:
Real time, environmental and contextual data
AI driven
- 10. The Wake-up Call
Nearly ⅓ of
marketers do not
optimise digital
media in-flight
at all
[Source AC Senior Global Marketers Survey, Sept 2023]
10
10
- 11. 11
Despite the fact that optimising your ads in-flight can
boost revenue and deliver much more creative
effectiveness, our findings show that lots of
marketers still don’t do this.
- 12. 12
= The average added media value
our methodology returns on
campaigns vs non-optimised results
[Anonymised client data from over 6 billion live media impressions]`
17%
- 13. 13
13
Meta say use 2-8 ads*
We have proved that 36.49 for
Awareness and 14.81 ads for
Conversions work best.
This is why you need so many
ads - it works
- 14. 14
14
Nearly ⅓ of marketers are
not optimising
digital campaigns
in-flight at all.
- 15. 15%
46%
18% 18%
Do you feel you have
the information you
need to make
decisions around
creative effectiveness
mid-campaign?”
Survey in focus
TIME TO
DECISION
61%
Don’t feel confident / have
below average confidence
in their information to hand
15
5
VERY
CONFIDENT
2 3 4
1
NOT CONFIDENT
AT ALL
3%
“
👍
👍
- 16. How confident are you in
your team's ability to action
changes to a campaign
based on insight in a timely
manner?”
1
VERY CONFIDENT
5
NOT CONFIDENT AT ALL
2 3 4
7.7%
28.2%
17.9%
Survey in focus
TIME TO
ACTION
33.3%
12.8%
53.8%
don’t feel confident in their
team’s or partners’ ability to
action change in-flight
16
“
👍
👍
- 20. 20
Strategy
We strongly advise against having “set and forget” style of systems - whether that’s one ad you leave
running or using live optimisation that doesn’t do anything but that. You need to know what you want to
know about your audience. This means your media spend can also be your research budget. Your ads can
teach you about your audience.
Technology
There are various tools, Automated Creative included. You need a platform that lets you create ads very
quickly, tag them, and report back on them.
Insight
All your live creative optimisation ads should be telling you something about your audience. You’re
turning impressions into intelligence.
You can do all of this manually - taking a long time! - or you can use a combination of in-house and
partners.
- 22. 22
22
1. Production: scale, speed, ease
2. Insight: Turn impressions into intelligence
3. Performance: Expect 17% uplift, at least
With the right technology, you’ll be able to create
any ad for any audience on any platform, get it live
and then get data from it to optimise that ad based
on the actual reactions and insights from that ad.
- 23. 23
23
1. Production: scale, speed, ease
2. Insight: Turn impressions into intelligence
3. Performance: Expect 17% uplift, at least
Learn more about your audience as you evolve your
marketing strategy.
- 25. A case study we’re proud of:
PensionBee’s TVC was
re-recorded to include
messaging and visual insights
uncovered with AC online ads.
Results:
20.42% decrease in CPA
“[The work] led to us declaring our love for Automated Creative”
Jasper Martens / CMO
CASE STUDIES /
/ PensionBee
- 29. 29
The process and embedding it into your workstreams is one of the most
important things. We help integrate live optimisation into new companies day
in, day out. And we’ve seen that if you can start with some really basic
potential A/B testing or live optimisations in-flight and then it becomes much
easier to scale in time.
Firstly, start with a really clear objective for your business, your
communications strategy as a whole and for the specific campaign that you
will use to test live optimisation with. We’ve come across quite a few situations
where the client has overall objectives but not the metrics to match them. For
example, you could be looking at volume of customers into the business
where it may be more appropriate to measure the value or the number of new
customers vs existing.
- 31. 31
Have the right team members in the right position.
This will vary depending on the business you’re running and whether you work
with agency partners or whether you’re testing things in-house.
We recommend having one stakeholder focused on media, one who’s
focused on creative and the bigger creative mission, and then another
stakeholder who’s thinking about live creative optimisation. It’s essential to
have that third live creative optimisation “head” that can think about what
needs to be optimised in the moment. It might be that’s a balance between
media and creative - and it’s usually quite effective to have this third party
“referee” almost to manage everyone. It’s really important to have this neutral
element balancing the needs of creative and media.
- 33. 33
Focus on answering the big questions for your campaign.
All marketers have hunches about the type of advertising you think is working
for your brand - a message, a design style… anything, really. List out all the
questions you would like answers to.
From questions, you can structure into hypotheses and even lay them out
through a timeline.
- 35. 35
One of the best part of running ads at scale is you can test some wildcards,
too!
You don’t have to pick “one safe bet.” When you’re doing lots of multivariate
testing you can throw in messages or ideas that you think may not appeal to a
wide audience.
Very often, it’s these outsiders that drive performance - because they feel
surprising and generate standout content, particularly in a digital environment.
- 37. 37
Once you have your hypotheses and your wildcards, think about data for
decision making.
It’s really important to have access to data to drive your decisions, based on
the key objectives that you have. The ideal scenario is the way we work at
Automated Creative: we build dashboards and APIs into platforms. This gives
us live data coming through constantly.
Most platforms’ in-built ad management is not particularly user-friendly - so
anything that you can do to better visualise the data is immediately going to
help you make better decisions as you go through the process of live
optimisation.
- 39. 39
Finally, just getting started is really important.
It can be as simple as some initially A/B testing: one message vs another
message. Then scale it out from there. You don’t have to wait for the most
perfect possible testing opportunities. In fact, a lot of our clients start small,
run pilots on certain brands, and scale things out when they start seeing the
results and success.
So don’t think you have to immediately turn this on for every single campaign -
start small and scale it out from there.