2. Edistynyt Internet-markkinointi
‖Edistynyt Internet-markkinointi on viimeisimpien
tekniikoiden, työkalujen ja tiedon käyttöä
innovatiivisessa markkinointisuunnittelussa ja
-toteutuksessa, joka pohjautuu sekä perinteiseen
markkinointia (erit. kuluttajakäyttämistä) että Internet-
käyttäytymistä tutkivaan teoriaan.‖ (Salminen 2012)
EdInMa on:
– innovatiivista
– edelläkäyvää (tekniikoiden
soveltaminen ennen massoja)
– tehokasta (parhaimmillaan
‖ilmaista‖, huonoimmillaankin
halvempaa kuin muissa medioissa)
– oikeasti mitattavaa (tulokset
mitataan koko asiakkuuden ajalta)
– markkinointietiikan mukaista
2
3. Miksi startup-markkinointi on ”advanced”?
Startup on täällä
• ei rahaa
• ei markkinointi-
osaamista
• ei kasvukäyrää
• ei toivoa…
Mutta ei myöskään
• byrokratiaa Niukkuus
+ Mahdollisuus
• organisaatioinertiaa innovointiin!
• vanhoja oppeja paine
3 • rahaa haaskata
4. Miksi startup-markkinointi on ”advanced”?
1. Useimmat markkinoinnin teoriat koskevat
amerikkalaisia suuryhtiöitä (korporaatioita);
2. näillä on rajattomasti rahaa ja resursseja, joten ne
eivät välitä markkinointiaktiviteettien optimoinnista.
3. Kuitenkin monen suomalaisen yrityksen, etenkin
aloittelevan, suurin haaste on saavuttaa kasvuvaihe,
ei sen ylläpitäminen (pl. Rovio!).
4. Erityisesti startupit taistelevat niukkuuden kanssa
äärimmäisen kilpailussa ympäristössä, joten ne
joutuvat innovoimaan tavoitellessaan nopeaa
skaalautumista.
4
5. Esimerkkejä startup-markkinoinnin
innovaatioista…
• AARRR (Dave McClure) (acquisition, activation, revenue,
retention, referral)
• markkinatutkimuksen uusi teoria
– Customer Development, Steve Blank
– Lean Startup, Eric Ries
• Business model canvas (Alexander Osterwalder)
• Sisäänrakennettu viraalisuus
– Double-side referral incentives, Dropbox
(kaksisuuntainen suosittelukannustin)
– verkostoefektit (network effects; Google vs. Facebook)
• kaikki Andrew Cheniltä ♥
5
6. Customer development (Blank 2009)
Yhteensopivuudet
• ongelma-
ratkaisu
• tuote-
markkina
• liiketoi-
mintamalli
2. järjestelmällinen tapa
1. rinnakkainen 3. tuotteen ei tarvitse
testata asiakkaisiin
tuotekehityksen olla valmis, vaan
liittyviä oletuksia ja
kanssa, ei ennen minimihyvä
korjata tuotetta saatujen
6 tai jälkeen varhaisille omaksujille
tietojen perusteella
8. Startup-markkinoinnin ongelmia
• monet startupit päätyvät antamaan
tuotteensa ilmaiseksi käyttöön,
koska
– on kova kilpailu
– asiakkaiden ei uskota haluavan
maksaa/asiakkaat eivät oikeasti
halua maksaa
– monetisoinnin ajatellaan olevan
helpompaa kun ollaan saatu
miljoona käyttäjää
• ”ilmaisen kaljan syndrooma”
8
9. ”Build it and they will come!”
(startup fallacy)
NO Why?
THEY
9
WON’T
10. ”We’ll just make it viral” (startup fallacy)
“Let‟s imagine the
conversation at the marketing
STARTUP MARKETING department of the wireless
phone companies. „Let‟s see.
Should we spend $4 Billion on
advertising this year…or
should we just make it viral?‟.”
Virality is something that has to
be engineered from the
beginning…and it‘s harder to
create virality than it is to create
a good product. That‘s why we
often see good products with
poor virality, and poor products
with good virality. The reason
‖We just make it viral!‖ that over $150 Billion is spent
on US advertising each year is
because virality is so hard. If
virality was easy, there would
be no advertising industry.‖
(Kopelman 2010)
10
11. Kumpaa on helpompi markkinoida?
t
o
t
u
u
s
o
n
t
ä
lt
Markkinointi ei ä Hyvä tuote ei ole
ole taikaluoti v taikaluoti
ä
li
(Markkinointi- (Hyvän tuotteen
lt
harha) ä harha)
11
12. Kysymys: Millainen olisi täydellinen
markkinointijärjestelmä
• vaatimukset:
– reagoi automaattisesti sekä kysynnän vaihteluihin,
kilpailijoiden toimiin ja makrotason trendeihin
– toimii erilaisten sääntöjen perusteella automaattisesti
– testaa erilaisten markkinointiviestien toimivuutta ja
muokkaa toimintaansa sen mukaan (machine
learning)
• haasteet:
– hehe :) ….mistä alkaa?
12
13. Maailman ensi-ilta…
Täydellinen markkinointijärjestelmä (perfect
online marketing system) (Salminen 2012)
Starring:
13
14. • profiilit
• benchmark
• …
• kaikki data
Mitä se osaa?
APIs, open graph THE
• budjetin säätäminen • mielipidevaikuttajien
K CORE ja allokointi löytäminen ja
• kampanjoiden kontaktointi
sisällön • asiakkaan
mukauttaminen tunnistaminen,
• sisällön ja kustomointi/
päätöksenteko- kohderyhmän personointi
järjestelmä (DS) sovittaminen • monitoroi omaa ja
• automaattinen • järjestelmällinen kilpailijan brändiä ja
persoonien placement‘ien reagoi
Materiaalin rakentaminen läpikäynti ja valinta sentimenttimuutoksiin
• some-tiedon Required Science?
tuotanto yhdistäminen
content team profiileihin (vrt. • AI (HAL 9000)
Performable) • data mining
(”content • input/output- • machine learning
factory”) funktiot • sentiment analysis
• Taguchi-metodit
crowdsourcing-alustat …ja/tai • choice modelling (unstated
preferences)
(vrt. Transfluent, 99designs, community • small world networks (opinion
14 Mechanical Turk)
(UGC) leaders)
15. Tulevaisuuden visio: Täydellinen
monimuuttujajärjestelmä
• Käyttötapauksia (use cases):
– syötä eri versioita eri elementeistä
• copy-tekstit
• kuvamateriaali
• koko layout (fluid grids)
– kohdista ne tuotteille/sivuille
– järjestelmä luo sivut dynaamisesti ja jakaa liikenteen niiden välillä
– anna järjestelmän laskea automaattisesti parhaiten menestyvä
variaatio (optimal design)
– anna järjestelmän suositella potentiaalisia variaatioita historiallisen
datan pohjalta (ts. ehdottaa hypoteeseja)
• siis: tuota vain materiaali ja lataa järjestelmään, järjestelmä
päättää miten ja missä se kannattaa esittää
• järjestelmä kykenee analysoimaan itseään ja oppii onnistuneista
kampanjoista (esim. värit, sijoittelut, semanttiset valinnat)
15
16. Vrt. Webhooks (Performable 2010)
―Webhooks are a really powerful way to automate your marketing,
but it‘s not always easy to know how to use them or what to use them
for. Here is a list of some really cool webhooks that our customers
have created.
– Push In-app Notifications: Send a push notification to a user through an iOS
native app.
– Send SMS Alerts: Text a sales person when a hot lead returns to the pricing
page.
– Auto-Follow: Auto-follow users who follow your account on Twitter (uses
Performable pre-integration with Twitter)
– Trigger Offers: Trigger a special on-site offer for qualified returning
customers.
– Follow Up: Leave someone a voicemail saying that their support ticket has
been received.
– Facebook Posting: Post to someone's Facebook Wall thanking them for
mentioning you on Twitter.
– Internal Chat Alerts: Send alerts to your company‘s internal chat stream that
lets employees know something interesting/important has happened.‖
16
17. Automated sharing (Carter 2012)
‖One of the things we know about viral marketing is that
it requires two things: a click and a share. We had use
interesting compelling content to get both, but the fact
that there were two actions required made it harder
to achieve the viral effect. What a custom action will
do for you is basically eliminate one obstacle - since you
only opt-in to an action once, every following action you
take is automatically shared with all your friends.‖ (Carter 2012)
Evolution of
sharing in the
Internet amount of
(Salminen 2012) sharing
effort of
sharing
Copy-paste Click Open graph
17
18. ”Open Graph apps allow third-party
developers to create „frictionless‟ apps that,
after a user provides permission once,
automatically share users’
engagement with the app on
Facebook. Furthermore, users‟ friends are
easily able to join in on this shared activity.
So if one Facebook user was listening to a
song on Spotify and a friend saw that story
in their News Feed or Ticker, they could
start listening to that song, too, with just a
click.”
19.
20. • “One-click sign up eliminates virtually all
friction in the sign up process (other than
the anxiety over potential privacy concerns), and
solves the „dang, I forgot my password again‟
problem. Most Facebook users are perpetually
logged in, or they log in frequently enough their
login info is committed to memory.”
• “Depending on your business, you may require
customer information that does not exist
in a Facebook profile, like unique account
number, postal code, industry, account type
(business or consumer) or mobile number (telco
companies).”
21. “With access to profile data, web sites can
personalize based on keywords in both the
Connected user’s profile and his/her social
graph (gift suggestions, birthday reminders,
etc).”
24. “Connected users can view what their friends
have viewed, commented on, or reviewed on
your site. This ‘social proof’ builds trust, as
people value their friends’ opinions over
strangers’.”
25. •
•
•
“Folks who frequent Facebook more than their
email inboxes may prefer to receive product
back-in-stock or shipment notifications
through Facebook, especially when email
inboxes are already overflowing.”
26. • “Remember that most community
features require active participation
from a large number of users to
make them useful.”
• “Without a critical mass of Connected
and active Facebookers, these features
add little value, which may cause the
Connected to disconnect.”
27. ”While reviews, endorsements, and
activities from real friends are more
trustworthy, these are few and far
between; even Levi’s was unable to
meaningfully aggregate Likes from
within our social graphs so as to
aid in purchase planning.”
28. Miten motivoida asiakkaita antamaan
palautetta? (tai tuottamaan sisältöä?)
Tai YLIPÄÄNSÄ tekemään mitään?
Jos viraalimarkkinointi on Graalin malja, niin tämä on
Viisasten kivi…!
28
29. 1. charity
”Linas bought an experience gift and by doing so
financed a microcredit in Bangladesh. Hooray!”
2. gamification
”Almantas has browsed all romantic gifts and has
received the Don Juan badge. All romantic gifts to his
friends -10% TODAY ONLY.”
3. co-shopping
”This gift has been ticked [4] times in the last [5]
hours… [1] more tick needed to receive [15%]
discount, you have [10] minutes left. Tick now!”
(shows countdown, renews automatically, applies to
select products)
4. invite friends to experiences
Like sharing, but with particular friends.
5. sky‟s the limit… really!
30. Avoin graafi ei ole vastaus kaikkiin
maailmankaikkeuden markkinointiongelmiin
―As brand marketers, your task now is to get your
fans and customers to want to make the decision to
share their activity with you automatically from that
point forward. How do you make them so proud of their
association with your brand that they‘ll want to do that?
How will you reward them for doing so? How will you
make sure your product is so good, your fan‘s friends
will also dig it if they try it?‖ (Stiles 2012)
• helpotetaan
• se, että voidaan tehdä jotain, ei
asiakkaan
takaa että se kannattaa tehdä
prosesseja
• näkevätkö asiakkaat hyödyn?
• lisätään kysynnän
• asiakkaan intentiota ja
ja tarjonnan
tavoitteita voi ohjata, mutta
kohtaamisen
niiden muuttaminen on erittäin
todennäköisyyttä
vaikeaa
30
31. Haasteita
• Algoritmit vs. ihmiskognitio (HITs)
ratkaisu: crowdsourcing
• Google testaa 40 sinisen eri sävyä, kuitenkin design
on jotain muuta kuin osiensa summa
• Vaikka tekisi rationaalisempia päätöksiä, miten kone
voi olla luova?
Mitä on luovuus? AI osaa yhdistellä asioita
useammalla tavalla kuin ihminen – jos se kykenee
lisäksi testaamaan niiden tehon, eikö se osoita
luovuutta?
• Parametrointi: hyvien päätösten tekemiseksi tarvitaan
PALJON parametreja, osaamista ja tietoa
rajapintojen täytyy olla avoimia kanavakitkan
välttämiseksi
• Resistanssi (rise against the machines…)
31
32. Esimerkki automaattisen järjestelmän
problematiikasta: fluktuaatio (Libby 2010)
• Bidin kehitys
avainsanalle,
jonka lähtöarvo
on 3,80 $ ja
optimiarvo 4,00 $
(jossa CTR=6 %)
• Automaattinen
sääntö: ‖Jos CTR
on korkeampi tai
matalampi kuin 6
%, nosta tai laske
bidiä 39 %.‖
• Bid ei milloinkaan
asetu optimiin. (Vrt. Edgeworthin hintasykli; ja
automaattiset sijoitusbotit)
32
33. Käyttäytymisen tulkinnan ongelma (Stiles
2012)
‖Some obvious questions naturally come up. Just
because I read a book doesn‘t mean I liked it. Just
because I read an article doesn’t mean I’m into that
subject. Maybe I‘m just doing research…on
lingerie. The point is, it‘s still up to the user to craft what
activity they want to share, it‘s just that the sharing will
take off on its own once that decision has been
made.‖
– epäsuorin (ja joskus myös suorin…) menetelmin saatu
tieto asiakkaasta ei välttämättä kuvaa preferenssejä
luotettavasti, eikä sitä näin ollen voi käyttää
käyttäytymisen ennustamiseen
– esimerkiksi mainonnan kohdentaminen hakuprofiilin
perusteella ei välttämättä lisää relevanssia
33
35. What is good content?
―That‘s a great question. While there are many things in
the minds of Google engineers that I will never
understand, a lot of what they do is observing what
users do — in other words, they look for things like
incoming links, how long readers are spending on pages
and sites, and social sharing in venues like twitter and
Google+.
– That‘s how we measure ―quality‖ as well. We look at
what our readers do — which articles they‘re most
likely to read, send traffic to, share with social media
friends, etc.‖
Joudumme käyttämään
etämittareita (proxy) laadun
selvittämiseksi; liika testaaminen
voi kuitenkin vahingoittaa brändiä.
35
36. (cont’d…)
─ ―This inequality of heavily linked content != best content is exactly
what the non link related ranking factors are supposed to
mediate. Finding the actual best content, not necessarily just the
best linkbait. Whether or not they are succeeding is a whole
another conversation.
– We‘d like to think so, but search engines evaluate quantitatively -
which is why they're so heavily reliant on the link graph to determine
quality/relevance.
− The day when a machine has true semantic understanding - when it can
read a page and decide, without measuring external signals, the ‗quality‘
of that page - that will be the day I go off the grid. Because next we‘ll
have machines creating this ‗quality‘ content, and that is too much like an
Isaac Asimov story to be a good thing.
– I totally agree Mike. Semantic understanding could determine the true subject
matter, but when it can actually determine the content to be factually based or
even sound theory that will be the day.‖
36
37. Tulevaisuuden visio: Automaattinen
markkinointijärjestelmä
• hankkii tietoa asiakkaista ja sivun vierailijoista
• luo profiileja (marketing personas)
• luo automaattisesti lähteviä viestejä
• kampanjat alkavat ja loppuvat automaattisesti
• luo sääntöjä, joiden pohjalta variaatioihin
tehdään muutoksia Dave Bowman: Hello, HAL. Do you read me, HAL?
HAL: Affirmative, Dave. I read you.
Dave Bowman: Open the pod bay doors, HAL.
HAL: I'm sorry, Dave. I'm afraid I can't do that.
Dave Bowman: What's the problem?
HAL: I think you know what the problem is just as
•
well as I do.
Laplacen demoni? Dave Bowman: What are you talking about, HAL?
HAL: This mission is too important for me to allow
you to jeopardize it.
• FRANKENSTEIN?? Dave Bowman: I don't know what you're talking
about, HAL.
HAL: I know that you and Frank were planning to
• SKYNET!!!!?
disconnect me, and I'm afraid that's something I
cannot allow to happen.
Dave Bowman: [feigning ignorance] Where the hell
did you get that idea, HAL?
HAL: Dave, although you took very thorough
precautions in the pod against my hearing you, I
could see your lips move.
Dave Bowman: Alright, HAL. I'll go in through the
”I will not stop until emergency airlock.
HAL: Without your space helmet, Dave? You're
going to find that rather difficult.
you buy, human” Dave Bowman: HAL, I won't argue with you
37 anymore! Open the doors!
HAL: Dave, this conversation can serve no purpose
anymore. Goodbye.
38. Online-mainosverkot (advertising networks)
―The key function of an ad network is aggregation of ad
space supply from publishers and matching it with
advertiser demand. The phrase ‗ad network‘ by itself is
media-neutral in the sense that there can be a ‗Television Ad
Network‘ or a ‗Print Ad Network‘, but is increasingly used to
mean ‗online ad network‘ as the effect of aggregation of
publisher ad space and sale to advertisers is most commonly
seen in the online space. The fundamental difference between
traditional media ad networks and online ad networks is that
online ad networks use a central Ad server to deliver
advertisements to consumers, which enables targeting,
tracking and reporting of impressions in ways not
possible with analog media alternatives.‖ (Wikipedia 2012)
38
39. Mainospalvelinten edistyneisyys
―The typical common functionality of ad servers includes:
– Uploading advertisements and rich media.
– Trafficking ads according to differing business rules.
– Targeting ads to different users, or content.
– Tuning and optimization based on results.
– Reporting impressions, clicks, post-click & post-impression activities, and
interaction metrics.
Advanced functionality may include:
– Frequency capping so users only see messages a limited amount of time.
(Advertisers can also limit ads by setting a frequency cap on money-spending)
– Sequencing ads so users see messages in a specific order (sometimes known
as surround sessions).
– Excluding competition so users do not see competitors‘ ads directly next to
one another. (Usually done by bidding on keywords)
– Displaying ads so an advertiser can own 100% of the inventory on a page
(sometimes known as Roadblocks).
– Targeting ads to users based on their previous behavior (behavioral
marketing or behavioral targeting).
– Targeting specific IP-addresses i.e. targeting specific individuals or
companies‖
39
40. Uudelleenkohdentamisen ”strateginen”
dilemma (vrt. Bloch 2010)
―Imagine this scenario – an online shopper visits your site and
gets as far as the checkout process, then scoots. The shopper
then goes to a competitor‘s site and does the same thing and
moves on to yet another site. All other aspects being equal, if
the first competitor is using remarketing, they have a far
better chance than you of picking up that sale.
– The other thing we‘ll see happening with this I think is that
consumers will become conditioned to abandon carts in
the hope of the follow up email offering a better deal – so
timing could be important. Using the same scenario as
above, but this time you have implemented email
remarketing, if your competitor has his email to the shopper
before you, you may still miss out.‖
Automaattisen
järjestelmän hyväksikäyttö
(gaming)
40
41. Mitä on ’retargeting’?
(uudelleenkohdistaminen)
―Retargeting, or remessaging, is an advertising strategy
that allows you to target users who have visited your
website with ads to entice them to return to your
site. It is generally used to encourage users who didn‘t
convert to come back to complete a purchase or other
conversion step. But, it can also be used for a variety of
reasons including product upsells, branding and social
engagement.‖ (Lund 2011)
―Online shoppers looking for new running shoes visit a popular online
store, FastSneakers.com, to browse the different styles. Some shoppers
leave without buying anything. FastSneakers.com could add these
shoppers to a ‗Site Visitor‘ list. This will enable FastSneakers.com to
reach out to these potential buyers while they browse other websites,
with a compelling call-to-action or offer that will encourage them to
return to FastSneakers.com to complete a purchase.‖
41
42. Miten uudelleenkohdistaminen toimii?
―It basically starts with a client who visited your site, then before
completing a conversion, the client decided to leave. With
retargeting or remarketing in Google AdWords, a tracking code or a
third part cookie is inserted into that client‘s browser, enabling you to
advertise to them the product they have previously been interested in
your site. This only works with a website that has also enabled
retargeting. For example, a customer abandoned his shopping cart
in Website A. When he browses to another page, Website B, which
also has a retargeting mechanism like Website A, a form of
advertising either in pop-up style or banner style of the product that
was in his abandoned shopping cart will prompt him, thereby
persuading him to go back to Website A.‖ (Payne 2011)
42
43. ”Online advertising channel” (Salminen
2010), kuva toimijoista (actors)
• goals relate to • goals relate to securing • goals relate to
marketing strategy quality of advertising maximizing profit
• is the source of (end user experience) • is dependent of
revenue within chain • is dependent of revenue provided by
• wants access to revenue provided by network (however,
customers through advertiser several potential
mediators, end goal • aggregates publishers sources)
is usually sales and sells them to • joins network to gain
advertisers access to advertisers
Matalampi transaktiokustannus!
43
Mutta suhdespesifisyys ja informaatioasymmetria
45. Workflow dilemma and, subsequently,
ethics of scheduling
―The problem lies with the fact that social networking is
an endless task and so if a person starts his day with
twitter, then he/she is bound to lose time which was
required for doing an important job. It has been a
problem faced by many and the solution came in the
form of scheduling tweets. Through scheduling tweets
one can schedule a tweet at a time, when he can do
some other job. So is it ethical to schedule tweets?‖
(Hans 2010)
• keskusteluun osallistuminen on tärkeää, mutta jos
• koko työpäivä vierähtää somessa,
• onko oikein ajastaa/automatisoida osallistumista?
45
47. This was pretty close…
―A Performable user simply selects an action he wants
to carry out—he may want to capture visitors‘ e-mail
addresses, or get them to download an application or a
whitepaper. Performable gives the user a template
optimized to induce that action, along with a URL that
can be integrated into a marketing campaign. ‗And if you
click ‗clone,‘‘ says Cancel, ‗you can duplicate that page
and start testing multiple versions with different
marketing copy. You could create 20 versions of a page
in 10 minutes, if you wanted to. We‘d automatically start
routing traffic to the different pages, then tell you which
ones are performing better.‖
Eri alustojen ominaisuuksia
yhdistelemällä voi luoda
saavuttaa täydellisen järjestelmän
etuja ilman täydellisyyttä!
47