Creativity, Crystal Balls & Eating Ground Glass

@hannah_bo_banna creativity, crystal balls & eating ground glass @hannah_bo_banna

@hannah_bo_banna hello there!

bad at photos, pretty good at other stuff

@hannah_bo_banna my team’s job is to make things that journalists
want to write about & people want to share

@hannah_bo_banna huh?

this was created for PartyCasino, we analysed forty years of
box office data, to see which actors made the most profitable films

links from over 120 sites

this was created for Lenstore, it’s the world’s first gigapixel
time lapse panorama of London’s skyline

@hannah_bo_banna because it’s a gigapixel photo you can zoom in
wherever you want

@hannah_bo_banna & because it’s a time lapse you can also
see how London looks across a full day

this was created for GoCompare, it’s a tribute to the
modernist buildings we’ve lost to the wrecking ball

@hannah_bo_banna over the past 18 months, I’ve been responsible for
launching more than 100 creative pieces which have generated over 5,300 pieces of linked coverage

@hannah_bo_banna incidentally, you know why do we do this, right?

@hannah_bo_banna to gain links & coverage for our clients

@hannah_bo_banna links from highly authoritative sites, increase the authority of
our clients’ sites

@hannah_bo_banna over time this translates into stronger organic rankings

stronger organic rankings = more money (for the vast majority
of websites)

so, what’s this talk all about?

@hannah_bo_banna towards the end of last year someone asked me:

@hannah_bo_banna “how good are you at predicting whether or not
a piece will be successful?”

@hannah_bo_banna then I was like, wait a minute…

@hannah_bo_banna am I really good at this?

@hannah_bo_banna when it comes to memory humans are hellishly unreliable

@hannah_bo_banna ever put something in a safe place, & then
forgotten where that safe place is?

@hannah_bo_banna I feel like this is pretty common

@hannah_bo_banna & most people would accept they have a poor
memory for that sort of thing

@hannah_bo_banna but how good are you at remembering past events?

@hannah_bo_banna most people think that they’re able to accurately recall
things that happened in the past & how they were feeling at the time

@hannah_bo_banna however

@hannah_bo_banna neuroscientists have discovered that each time we remember something,
we reconstruct the event

we reassemble it

@hannah_bo_banna additionally

@hannah_bo_banna psychologists have noted that we suppress memories that are
painful or damaging to our self-esteem

@hannah_bo_banna what does this mean?

@hannah_bo_banna let me give you an example:

let’s imagine you’ve asked me whether or not I thought
that piece would be a success

@hannah_bo_banna right now my brain’s doing something like this…

@hannah_bo_banna I’m going back in time…

@hannah_bo_banna we’d have been working on it through September 2017

@hannah_bo_banna there was a bunch of other stuff going on
at the same time (we were selling the company, I did BrightonSEO, new client pitches)

@hannah_bo_banna I’m thinking about the challenges we had producing the
piece

@hannah_bo_banna then this pops in my brain

@hannah_bo_banna I’m trying to remember how I felt about the
piece right before it launched

@hannah_bo_banna then this pops in my brain

@hannah_bo_banna I think that I felt the data revealed some
great stories…

@hannah_bo_banna who’s the most profitable actor in Hollywood? Tom Cruise?
Brad Pitt? The Rock?

actually, it’s this guy, in the blue vest

@hannah_bo_banna but am I just reconstructing my memory based on
what happened post-launch?

@hannah_bo_banna remember this?

@hannah_bo_banna psychologists have noted that we suppress memories that are
painful or damaging to our self-esteem

@hannah_bo_banna judging the likelihood of success for creative pieces is
essentially what I’m paid to do

@hannah_bo_banna I am supposed to be good at it

@hannah_bo_banna it would be damaging to my self-esteem to recall
instances where I was wrong

@hannah_bo_banna so it’s conceivable that I would suppress memories like
that

@hannah_bo_banna or reconstruct, or reassemble my memory…

@hannah_bo_banna to create a version of history where I always
thought this campaign was a great idea

{insert your own expletive here}

@hannah_bo_banna so I figured I ought to find out how
good or otherwise I really am at predicting this stuff

@hannah_bo_banna so at the beginning of this year I resolved
to make a prediction about how each piece would perform

@hannah_bo_banna how did I do this?

@hannah_bo_banna this stuff isn’t strictly binary

@hannah_bo_banna so I created a scale:

@hannah_bo_banna Scoring Band LinkScore points A 10,000+ B 5,000 to
9,999 C 2,000 to 4,999 D 1,000 – 1,999 E less than 1,000

@hannah_bo_banna huh?

this is how we measure the ‘value’ of a link
https://www.vervesearch.com/linkscore/

@hannah_bo_banna scores range from 0-500 points

@hannah_bo_banna scores are calculated based on a range of factors
including: site authority, language, follow vs nofollow & more

@hannah_bo_banna here are a few examples of scores…

this link on the Guardian scores 456 points

this link on Fast Company scores 284 points

this link on the Independent is nofollow so it scores
42 points

@hannah_bo_banna to help you understand what this looks like in
terms of numbers of links

@hannah_bo_banna Scoring Band LinkScore points Equivalent number of links A
10,000+ more than 100 B 5,000 to 9,999 50-99 C 2,000 to 4,999 20-49 D 1,000 – 1,999 10-19 E less than 1,000 less than 10 this is just an indicator – an average link is not 100 points

@hannah_bo_banna so I began to make predictions about all of
our campaigns ahead of launch

@hannah_bo_banna wait… how do you make predictions like this?

@hannah_bo_banna I consider 3 things:

@hannah_bo_banna resonance

@hannah_bo_banna resonance noun the power to evoke emotion

@hannah_bo_banna initially I might consider this at a topic level

35x articles are written about AI vs RPA & these
articles get 8x more engagement

@hannah_bo_banna but it’s really important to think about this at
a human level too

@hannah_bo_banna how many people are likely to care about this?

@hannah_bo_banna how many people are likely to be touched in
some way by this?

@hannah_bo_banna more on resonance here: bit.ly/feels-2016 & bit.ly/time-machine-2016

@hannah_bo_banna breadth of appeal

@hannah_bo_banna how many publications are likely to cover this?

@hannah_bo_banna can we sell it in to different verticals?

@hannah_bo_banna different countries?

@hannah_bo_banna can we use the piece to tell a variety
of stories?

@hannah_bo_banna past experience

@hannah_bo_banna how have I seen similar pieces perform before?

@hannah_bo_banna so…

@hannah_bo_banna in January I began to make predictions about all
of our campaigns ahead of launch

@hannah_bo_banna fast forward around 8 months

@hannah_bo_banna & I’m pitching this talk to Will…

@hannah_bo_banna how good are your predictions?

@hannah_bo_banna back then I had not yet checked my own
numbers

@hannah_bo_banna & this rather lovely quote would prove to be
more true than I’d have liked …

@hannah_bo_banna “she who lives by the crystal ball soon learns
to eat ground glass…” ~Edgar R. Fiedler, The Three Rs of Economic Forecasting: Irrational, Irrelevant and Irreverent, 1977

@hannah_bo_banna ok, let’s do this…

@hannah_bo_banna this talk is in 3 parts: - how good
am I at predicting the future accurately? - what seems to affect my judgement? - what do I think I’ve learned?

@hannah_bo_banna how good am I at predicting the future accurately?

@hannah_bo_banna really not very good L

10,000+ more than 100 B 5,000 to 9,999 50-99 C 2,000 to 4,999 20-49 D 1,000 – 1,999 10-19 E less than 1,000 less than 10 as a reminder – these are the scoring bands – I predict a band for each campaign

@hannah_bo_banna how did I do?

@hannah_bo_banna I correctly predicted the LinkScore band for just 47%
of campaigns

@hannah_bo_banna which means I am wrong more often than I
am right

@hannah_bo_banna ok, so I’m wrong a lot… but how wrong?

@hannah_bo_banna how often am I a bit wrong?

10,000+ more than 100 B 5,000 to 9,999 50-99 C 2,000 to 4,999 20-49 D 1,000 – 1,999 10-19 E less than 1,000 less than 10 a bit wrong = wrong by one scoring band

@hannah_bo_banna for around 33% of campaigns my prediction was out
by one scoring band

@hannah_bo_banna here are some examples of where I was a
bit wrong

we created this for GoCompare Travel Insurance, to highlight a
variety of languages which are in danger of extinction

I thought this campaign would do better than it actually
did… Prediction: band B (5,000-9,999 points) Actual: band C (4,957 points & 48 links) Prediction: B Actual: C

we created this for IG Index, it compares the salaries
of various world leaders to GDP, population & average earnings

Prediction: C Actual: B I thought this campaign would do
worse than it actually did… Prediction: band C (2,000-4,999 points) Actual: band B (8,844 points & 74 links)

@hannah_bo_banna & how often am I very wrong?

10,000+ more than 100 B 5,000 to 9,999 50-99 C 2,000 to 4,999 20-49 D 1,000 – 1,999 10-19 E less than 1,000 less than 10 very wrong = wrong by two or more scoring bands

@hannah_bo_banna for around 20% of campaigns my prediction was out
by two or more scoring bands

@hannah_bo_banna quick recap?

47% 33% 20% 0% 10% 20% 30% 40% 50% 60%
70% 80% 90% 100% All Campaigns Correct A Bit Wrong Very Wrong

@hannah_bo_banna dear reader, this does not make me feel good

@hannah_bo_banna I’ve been doing this stuff for close to ten
years now

@hannah_bo_banna I really feel like I ought to be more
right than this

47% 33% 20% 0% 10% 20% 30% 40% 50% 60%
70% 80% 90% 100% All Campaigns Correct A Bit Wrong Very Wrong

@hannah_bo_banna eeek, let’s move on

@hannah_bo_banna how good am I at predicting a winner?

10,000+ more than 100 B 5,000 to 9,999 50-99 this is just an indicator – an average link is not 100 points

@hannah_bo_banna actually I did a bit better here J

@hannah_bo_banna I correctly predicted the LinkScore band for 75% of
these campaigns

@hannah_bo_banna which means I am right more often than I
am wrong J

@hannah_bo_banna but for the remaining 25% I wasn’t a bit
wrong, I was very wrong* (*very wrong means I was out by two or more scoring bands)

@hannah_bo_banna so I was very wrong about one in four
of our best performing campaigns

75% 25% 0% 10% 20% 30% 40% 50% 60% 70%
80% 90% 100% Best Performing Correct A Bit Wrong Very Wrong

@hannah_bo_banna quick recap?

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Best Performing All Campaigns Correct A Bit Wrong Very Wrong

@hannah_bo_banna I am simultaneously more right AND more wrong

@hannah_bo_banna I thought you might be interested to see one
of the campaigns I was very wrong about

we created this for GoCompare, using 20 years of IMDb
data, we calculated the most filmed locations on the planet

I predicted less than 1,000 points; the campaign actually generated
over 29,000 points (392 links) Prediction: E Actual: A

@hannah_bo_banna so what was my problem?

@hannah_bo_banna remember I said I consider 3 things when making
predictions? (I’ll look at each in turn)

@hannah_bo_banna resonance

@hannah_bo_banna have you ever wondered what the most filmed locations
on earth are?

@hannah_bo_banna hmmmm

as a topic I guess it’s somewhat resonant…

@hannah_bo_banna In short, I’m not 100% sold

@hannah_bo_banna but I thought we had a bigger problem

@hannah_bo_banna breadth of appeal

I didn’t feel that the results were surprising, and, as
such, I felt that the piece would have limited appeal

@hannah_bo_banna past experience

@hannah_bo_banna but I think it some pretty lazy thinking about
past experiences that really led me awry

I didn’t feel that the results were surprising, and, as
such, I felt that the piece would have limited appeal

I think I was hoping for a result like this…

or maybe this…

I was so concerned about this, that when we reviewed
the data I suggested we drop the campaign

the client wasn’t convinced about this one either…

@hannah_bo_banna so what’s going on here?

@hannah_bo_banna when you create a lot of pieces (& we
really do)

@hannah_bo_banna you start to come up with a sort of
shorthand

@hannah_bo_banna which can be either helpful or harmful depending on
how lazy your thinking is

in my head, this was ‘Director’s Cut’ for filming locations

@hannah_bo_banna & perhaps you can kind of see where I
was going with that

@hannah_bo_banna but actually, the piece is closer to this

this tells you exactly what you would expect…

@hannah_bo_banna but it’s compelling nevertheless, because we’ve proved it to
be true

this piece is not ‘Director’s Cut’ for filming locations it’s
closer to ‘Unicorn League’ for filming locations

the results aren’t surprising… but journalists covered it because we’ve
proved it

@hannah_bo_banna but that wasn’t the only direction I was wrong
in…

@hannah_bo_banna the piece had huge breadth of appeal (which somehow
I just missed)

topical angles…

niche angles…

@hannah_bo_banna & many, many local angles

@hannah_bo_banna I’m very glad my team talked me round J

29,611 points (392 links) we’ve generated coverage from this piece,
every single month, for the past nine months…

@hannah_bo_banna so we’ve seen that lazy thinking, or shorthand like
‘Director’s Cut’ for film locations affects my judgement

@hannah_bo_banna & I wondered what else might affect it

@hannah_bo_banna if the campaign is my idea how accurate are
my predictions?

@hannah_bo_banna when something is your idea, there’s a danger of
falling in love with it

@hannah_bo_banna so I wondered if maybe my predictions for my
own ideas were wildly optimistic

@hannah_bo_banna but actually…

@hannah_bo_banna my predictions for my own campaign ideas were 78%
accurate

@hannah_bo_banna & where I was wrong, it was only by
one scoring band

@hannah_bo_banna LinkScore points Equivalent number of links 10,000+ more than
100 5,000 to 9,999 50-99 2,000 to 4,999 20-49 1,000 – 1,999 10-19 less than 1,000 less than 10 this is just an indicator – an average link is not 100 points

@hannah_bo_banna my predictions about my own ideas are the most
accurate

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% My Ideas Best Performing All Campaigns Correct A Bit Wrong Very Wrong

this really surprised me

@hannah_bo_banna I then wondered if whether or not I ‘loved’
the idea had an impact on my predictions

@hannah_bo_banna I recognise this might be tricky to parse

@hannah_bo_banna I make predictions based on how well I think
each campaign will do

@hannah_bo_banna high prediction ≠ love

@hannah_bo_banna here’s an example:

this piece reveals the most congested roads across the UK

Prediction: A

@hannah_bo_banna how did I come up with this?

@hannah_bo_banna motoring journalists can be tough to pitch…

@hannah_bo_banna unless it’s got an engine, wheels, & goes vroom,
frequently, they’re not interested

@hannah_bo_banna but there are some exceptions:

@hannah_bo_banna this is a resonant topic because pretty much everyone
gets annoyed about being stuck in traffic

@hannah_bo_banna it has reasonable breadth of appeal because we can
go to both UK nationals & regionals

@hannah_bo_banna & I feel confident because I’ve seen other similar
stuff do well before

Prediction: A Love it? Nope Actual: band A (over 16,000
points & 317 links)

@hannah_bo_banna if I ‘love’ the idea, does it impact my
predictions?

@hannah_bo_banna it really does

@hannah_bo_banna if I love the idea my predictions are only
accurate 36% of the time

@hannah_bo_banna also, if I love the idea my predictions are
always high as opposed to low

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Love My Ideas Best Performing All Campaigns Correct A Bit Wrong Very Wrong

@hannah_bo_banna what if I don’t love the idea?

@hannah_bo_banna If I don’t love the idea my predictions are
accurate 71% of the time (& when I’m wrong I’m only a bit wrong)

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Don't Love Love My Ideas Best Performing All Campaigns Correct A Bit Wrong Very Wrong

@hannah_bo_banna what do I think is going on here?

@hannah_bo_banna I’m pretty good at making accurate predictions for my
own ideas & ideas I don’t love

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Don't Love My Ideas Correct A Bit Wrong Very Wrong

@hannah_bo_banna but I’m pretty terrible at making accurate predictions for
ideas I love

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Love Don't Love My Ideas Correct A Bit Wrong Very Wrong

@hannah_bo_banna apparently I don’t fall in love with my own
ideas, but I do fall in love with other peoples

@hannah_bo_banna what else did I learn?

@hannah_bo_banna I may not be predicting the future

@hannah_bo_banna I’m concerned I might actually be affecting it

the campaign generated 9,790 points (200 links) Prediction: B Actual:
B

@hannah_bo_banna but that’s not the whole story

0 1000 2000 3000 4000 5000 6000 7000 8000 9000
10000 April May Demolishing Modernism – Cumulative Points by Month

0 1000 2000 3000 4000 5000 6000 7000 8000 9000
10000 April May June Demolishing Modernism – Cumulative Points by Month

0 1000 2000 3000 4000 5000 6000 7000 8000 9000
10000 April May June July August Demolishing Modernism – Cumulative Points by Month

@hannah_bo_banna I don’t think I did this deliberately (in order
to make my predictions correct)

@hannah_bo_banna but clearly my own belief or bias affected the
outcome here

@hannah_bo_banna what if my prediction was different?

I predicted 5,000 – 9,999 points; but what if I’d
have predicted 1,000 – 1,999 points? Prediction: D? Actual: D?

0 1000 2000 3000 4000 5000 6000 7000 8000 9000
10000 April May Demolishing Modernism – Cumulative Points by Month would I have been ok with stopping here?

0 1000 2000 3000 4000 5000 6000 7000 8000 9000
10000 April May June July August Demolishing Modernism – Cumulative Points by Month we never would have got here…

@hannah_bo_banna I deliberately did not share my predictions with my
team

@hannah_bo_banna they didn’t even know I was making predictions (they
do now)

@hannah_bo_banna but I wonder if I have been unconsciously influencing
them nevertheless

@hannah_bo_banna ok, let’s wrap this thing up…

@hannah_bo_banna I’d like to leave you with a few thoughts

@hannah_bo_banna #1

@hannah_bo_banna start recording your own predictions now

@hannah_bo_banna actually record them

@hannah_bo_banna when I came to run the numbers on my
own predictions I was shocked by some of the numbers I’d written down

I was certain I had scored this & many other
creative campaigns higher

@hannah_bo_banna your memory is unreliable

@hannah_bo_banna each time we remember something, we reconstruct the event

@hannah_bo_banna we reassemble it

@hannah_bo_banna & we suppress memories that are painful or damaging
to our self-esteem

@hannah_bo_banna as such…

@hannah_bo_banna your judgement might be considerably worse than you think
it is

@hannah_bo_banna this is actually a really powerful thing to learn

@hannah_bo_banna over time if you can recognise the situations in
which your judgements are bad

@hannah_bo_banna you might be able to adapt & make better
decisions

@hannah_bo_banna it is really good for me to know that:

@hannah_bo_banna I’m pretty good at spotting a ‘great’ campaign

75% 25% 0% 10% 20% 30% 40% 50% 60% 70%
80% 90% 100% Best Performing Correct A Bit Wrong Very Wrong

@hannah_bo_banna I’m pretty good at making accurate predictions for my
own ideas & ideas I don’t love

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Don't Love My Ideas Correct A Bit Wrong Very Wrong

@hannah_bo_banna & it’s really important that I know that…

@hannah_bo_banna I’m pretty terrible at making accurate predictions for ideas
I love

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%
100% Love Don't Love My Ideas Correct A Bit Wrong Very Wrong

@hannah_bo_banna I don’t fall in love with my own ideas,
but I do fall in love with other peoples

@hannah_bo_banna if you do decide to make predictions like this,
exercise a little caution…

@hannah_bo_banna you might not be predicting the future

@hannah_bo_banna you might actually be affecting it

@hannah_bo_banna your own belief or bias may affect the outcome
(which might be good, or really, really, really bad)

I predicted 5,000 – 9,999 points; but what if I’d
have predicted 1,000 – 1,999 points? Prediction: D? Actual: D?

@hannah_bo_banna ultimately…

@hannah_bo_banna getting things right, is more important than being right

@hannah_bo_banna give people the opportunity to try stuff out

@hannah_bo_banna even if you’re not convinced that stuff will work

@hannah_bo_banna firstly because if things do go wrong people will
actually learn something

@hannah_bo_banna if you’re not allowed to make mistakes, you can’t
learn anything

@hannah_bo_banna but mainly because…

@hannah_bo_banna you know a lot less than you think

@hannah_bo_banna looking back…

@hannah_bo_banna there have been countless times members of my team
have tried things I was *certain* would not work

@hannah_bo_banna but I let them try out those things anyway

@hannah_bo_banna & you know what?

@hannah_bo_banna stopping them from trying those things would have been
a massive error

@hannah_bo_banna because a whole bunch of those things worked out
just great

good luck out there x

@hannah_bo_banna Hannah Smith Head of Creative Verve Search send questions
and/or pictures of cats to: [email protected] tweet things @hannah_bo_banna

@hannah_bo_banna Big love to the team at Verve who helped
me put this together J Images: fortune: http://askbobarcana.blogspot.com/2013/02/the-golden-wheel-fortune-teller.html money: http://fortune.com/2017/01/19/scent-of-money/ good: https://www.youtube.com/watch?v=eHT5KNoWjzY lego: https://www.looper.com/40131/trolls-director-reassemble-lego-movie-sequel/ jesus titty christ: http://www.sickchirpse.com/illusions-that-will-blow-your-mind/ calvin & hobbes: http://inverseintuition.org/phi_2010/areas/cavecliff.html danbo - https://www.flickr.com/photos/nomadic_lass/6889892777/in/set-72157629144987013 unicorn - http://queenofheartsonthesleeve.tumblr.com/post/58736423727/ok-not-much-but-here-it-is-3 Credits

Creativity, Crystal Balls & Eating Ground Glass

Creativity, Crystal Balls & Eating Ground Glass

Other Decks in Business

Featured

Transcript