SuperIQ Analytics

7/30/2019 SuperIQ Analytics

1/77

>SuperIQAnaly/cs


2/77

>Smartdatadrivenmarke/ng

June2010 DataliciousPtyLtd 2

MediaA;ribu/on&ModelingOp/misechannelmix,predictsales

Tes/ng&Op/misa/onRemovebarriers,drivesales

Boos/ngROI

TargetedDirectMarke/ngIncreaserelevance,reducechurn

Usingdatatowidenthefunnel


3/77


CliveHumby:Dataisthenewoil


4/77


Oilanddatacomeataprice


5/77

>GoogleNgram:Privacy



6/77

June2010 DataliciousPtyLtd

Collec/ngdata

forthesakeofitortoaddvalue

tocustomers?

6


7/77

>Metricsframework

101011010010010010101111010010010101010100001011111001010101

010100101011001100010100101001101101001101001010100111001010

010010101001001010010100100101001111101010100101001001001010



8/77

Awareness Interest Desire Ac/on Sa/sfac/on

>AIDAandAIDASformulas


Socialmedia

Newmedia

Oldmedia


9/77

Reach

(Awareness)Engagement

(Interest&Desire)

Conversion

(Ac-on)

+Buzz

(Sa-sfac-on)

>SimplifiedAIDASfunnel



10/77

Peoplereached

Peopleengaged

Peopleconverted

Peopledelighted

>Marke/ngisaboutpeople


40% 10% 1%


11/77

Peoplereached

Peopleengaged

Peopleconverted

Peopledelighted

>Addi/onalfunnelbreakdowns


40% 10% 1%

Newprospectsvs.exis-ngcustomers

Brandvs.directresponsecampaign


12/77


Newvs.returningvisitors


13/77


AUNZvs.restofworld


14/77


15/77

Exercise:Metricsframework



16/77

Level Reach Engagement Conversion +Buzz

Level1,

people

Level2,strategic

Level3,

tac/cal

Funnel

breakdowns

>Exercise:Metricsframework



17/77


18/77

>Establishingabaseline


Switchalladver-singoffforaperiod

of-me(unlikely)orestablishasmaller

controlgroupthatisrepresenta-veof

theen-repopula-on(i.e.searchterm,geography,etc)andswitchoffselected

channelsoneata-metominimise

impactonoverallconversions.


19/77

>Importanceofcalendarevents


Trafficspikesorotherdataanomalieswithoutcontextare

veryhardtointerpretandcanrenderdatauseless


20/77

>Mediaa;ribu/on

101011010010010010101111010010010101010100001011111001010101

010100101011001100010100101001101101001101001010100111001010

010010101001001010010100100101001111101010100101001001001010



21/77

Directmail,

email,etc

Facebook

Twi;er,etc

>Campaignflowsarecomplex


POSkiosks,

loyaltycards,etc

CRM

program

Homepages,

portals,etc

YouTube,

blog,etc

Paid

search

Organic

search

Landingpages,

offers,etc

PR,WOM,

events,etc

TV,print,

radio,etc

=Paidmedia

=Viralelements

Callcenter,

retailstores,etc

=Saleschannels

Displayads,

affiliates,etc


22/77

TV/Printaudience

Search

audience

Banner

audience

>Mediachannelsoverlap



23/77

>Indirectdisplayimpact



24/77

>Indirectdisplayimpact



25/77

>Successa;ribu/onmodels


Banner

Ad$100

EmailBlast

Paid

Search

$100

Banner

Ad

$100

Affiliate

Referral

$100

Success$100

Success$100

Banner

Ad

Paid

Search

Organic

Search$100

Success$100

Lastchannelgetsallcredit

Firstchannel

getsallcredit

Allchannelsget

equalcredit

Print

Ad

$33

Social

Media

$33

Paid

Search

$33

Success$100

Allchannelsgetpar/alcredit

PaidSearch


26/77

>Firstandlastclicka;ribu/on


Chartshows

percentageof

channeltouch

pointsthatlead

toaconversion.

Neitherfirst

norlastclickmeasurement

wouldprovide

truepicture

PaidOrganicSearch

EmailsShoppingEngines


27/77

Closer

SEMGeneric

BannerView

TV

Ad

>Fullpathtopurchase


Influencer Influencer $

BannerClick Online

SEOGeneric

AffiliateClick

Offline

SEO

Branded

Direct

Visit

Email

UpdateAbandon

DirectVisit

SocialMedia

SEOBranded

Introducer


28/77

>Searchcalltoac/onforoffline



29/77


30/77

>PURLsboos/ngDMresponserates


Text


31/77

>Offlinesalesdrivenbyonline


Websiteresearch

Phoneorder

Retailorder

Onlineorder

Cookie

Adver/singcampaign

Creditcheck,fulfilment

Onlineorderconfirma/on

Virtualorderconfirma/on

Confirma/onemail


32/77

>Understandingchannelmix



33/77


34/77

>Websiteentrysurvey


Channel %ofConversions

StraighttoSite 27%

SEOBranded 15%

SEMBranded 9%

SEOGeneric 7%

SEMGeneric 14%

DisplayAdver-sing 7%

AffiliateMarke-ng 9%

Referrals 5%

EmailMarke-ng 7%

DedupedCampaignReport

}Channel %ofInfluence

WordofMouth 32%

Blogging&SocialMedia 24%

NewspaperAdver-sing 9%

DisplayAdver-sing 14%

EmailMarke-ng 7%

RetailPromo-ons 14%

GreatestInfluenceronBrandedSearch/STS

Conversionsaributedtosearchterms

thatcontainbrandkeywordsanddirect

websitevisitsaremostlikelynotthe

origina-ngchannelthatgeneratedthe

awarenessandassuchconversion

creditsshouldbereallocated.


35/77

>Adjus/ngforofflineimpact


+15+5 +10

-15-5 -10


36/77

>Smarttarge/ng

101011010010010010101111010010010101010100001011111001010101

010100101011001100010100101001101101001101001010100111001010

010010101001001010010100100101001111101010100101001001001010



37/77

Therightmessage

ViatherightchannelTotherightperson

Attheright-me

Targe/ng



38/77

CaptureinternettrafficCapture50100%offairmarketshareoftraffic

IncreaseconsumerengagementExceed50%ofbestcompe-torsengagementrate

CapturequalifiedleadsandsellConvert1015%toleadsandofthat20%tosales

Buildingconsumerloyalty

Build60%loyaltyrateand40%salesconversion

IncreaseonlinerevenueEarn1020%incrementalrevenueonline

>Increaserevenueby10-20%



39/77

>Newconsumerdecisionjourney


Theconsumerdecisionprocessischangingfromlineartocircular.


40/77

>Newconsumerdecisionjourney


Theconsumerdecisionprocessischangingfromlineartocircular.

Changeincreasestheimportanceof

experienceduring

researchphase.

Onlineresearch


41/77



42/77

>Coordina/onacrosschannels


Off-sitetarge/ng

On-sitetarge/ng

Profiletarge/ng

Genera/ngawareness

Crea/ngengagement

Maximisingrevenue

TV,radio,print,

outdoor,search

marke-ng,display

ads,performance

networks,affiliates,

socialmedia,etc

Retailstores,instore

kiosks,callcenters,

brochures,websites,

mobileapps,online

chat,socialmedia,etc

Outboundcalls,direct

mail,emails,social

media,SMS,mobile

apps,etc


43/77

Offsitetarge-ng

Onsite

targe-ng

Profile

targe-ng

>Integra/ngtarge/ngplaorms



44/77

Campaignresponsedata

>Combiningdatasources


Customerprofiledata

+ Thewholeisgreaterthanthesumofitsparts

Websitebehaviouraldata


45/77

>Transac/onsplusbehaviours


+oneoffcollec-onofdemographicaldata

age,gender,address,etc

customerlifecyclemetricsandkeydatesprofitability,expira/on,etcpredic-vemodelsbasedondatamining

propensitytobuy,churn,etchistoricaldatafromprevioustransac-ons

averageordervalue,points,etc

CRMProfile

UpdatedOccasionally

trackingofpurchasefunnelstage

browsing,checkout,etc

trackingofcontentpreferences

products,brands,features,etctrackingofexternalcampaignresponses

searchterms,referrers,etctrackingofinternalpromo-onresponses

emails,internalsearch,etc

SiteBehaviour

UpdatedCon/nuously


46/77

>Samplecustomerleveldata



47/77

ThestudyexamineddatafromtwooftheUKsbusiestecommercewebsites,ASDAandWilliamHill.

Giventhatmorethanhalfofallpageimpressionsonthesesitesarefromloggedinusers,theyprovidedarobust

sampletocompareIPbasedandcookiebasedanalysisagainst.Theresultswerestaggering,forexampleanIPbasedapproachoveres-matedvisitorsbyupto7.6-meswhilstacookie-basedapproachoveres/matedvisitorsbyupto2.3/mes.

>Uniquevisitoroveres/ma/on


Source:WhitePaper,RedEye,2007


48/77

>Maximiseiden/fica/onpoints

20%

40%

60%

80%

100%

120%

140%

160%

0 4 8 12 16 20 24 28 32 36 40 44 48

Weeks

Probabilityofiden-fica-onthroughCookies

June2010 48DataliciousPtyLtd


49/77

>Customerprofilinginac/on


Usingwebsiteandemailresponses

tolearnalilebitemoreabout

subscribersatevery

touchpointtokeep

refiningprofilesandmessages.


50/77

Geo-demographicdata

>Enhancingdatasources


3rdpartydata

+ Thewholeisgreaterthanthesumofitsparts

Customerprofiledata


51/77

>Geo-demographicsegments



52/77



53/77

Onsite

segments

Offsite

segments

>Combiningadplaorms


CRM


54/77

>TheDataliciousSuperTag


SuperTag

AdServers

PaidSearch

AffiliatePrograms

BehavioralTarge/ng

ABTes/ngHeatMaps

LiveChat

WebAnaly/cs

MediaA;ribu/on

Easilyimplementandupdate

anytagonanywebsiteswithout

ITinvolvement.

Deduplicateconversionsand

collectmediaaribu-ondatatoboostreturnonadspend.

Implementcomplexretarge-ng

strategiesacrossplaormsto

increaseresponserates.

Enableadvancedfeaturessuch

heatmaps,tes-ngandlivechat

toop-miseconversions.


55/77



56/77


AppleiPhone4


57/77


AppleiPhone4


58/77

>Affinityre-targe/nginac/on


Differenttypeofvisitorsrespondto

differentads.By

usingcategory

affinitytarge-ng,

responseratesare

liedsignificantlyacrossproducts.

MessageCTRByCategoryAffinity

Postpay Prepay Broadb. Business

BlackberryBold

5GBMobileBroadband

BlackberryStorm

12MonthCaps

Google:vodafone

omniturecasestudy

orh;p:bit.lyde70b7


59/77

>Ad-sequencinginac/on


Marke-ngisabouttellingstoriesand

storiesarenotsta-c

butevolveover-me

Adsequencingcanhelpto

evolvestoriesover-methe

moreusersengagewithads


60/77

>Samplesitevisitorcomposi/on


30%exis/ngcustomerswithextensive

profileincludingtransac-onalhistoryofwhichmaybe50%canactuallybe

iden-fiedasindividuals

30%newvisitorswithno

previouswebsitehistory

asidefromcampaignor

referrerdataofwhich

maybe50%isuseful

10%serious

prospectswithlimited

profiledata

30%repeatvisitorswith

referraldataandsome

websitehistoryallowing

50%tobesegmentedby

contentaffinity


61/77

>Searchcalltoac/onforoffline



62/77

>PURLsboos/ngDMresponserates


Text


63/77

>Uniquephonenumbers


2outof3callers

hangupasthey

cannotgettheir

informa-onfast

enough.

Uniquephone

numberscan

helpimprovecallexperience.


64/77

Exercise:Clientdatajourney



65/77

>Exercise:Clientdatajourney


Toreten/onmessagesTotransac/onaldata

Fromsuspectto Tocustomer

Frombehaviouraldata Fromawarenessmessages

TimeTime

prospect


66/77



67/77

Exercise:Targe/ngmatrix



68/77

Purchase

Cycle

Segments:Colour,price,

productaffinity,etc

Media

Channels

Data

Points

Default,awareness

Research,

considera/on

Purchaseintent

Reten/on,upcross-sell

>Exercise:Targe/ngmatrix



69/77

Purchase

Cycle

Segments:Colour,price,

productaffinity,etc

Media

Channels

Data

Points

Default,awareness

HaveyouseenA?

HaveyouseenB?

Display,search,etc

Default

Research,

considera/on

Ahasgreat

features!

Bhasgreat

features!

Search,

website,etc

Adclicks,

prodviews

Purchaseintent

Adeliversgreatvalue!

Bdeliversgreatvalue!

Website,emails,etc

Cartadds,checkouts

Reten/on,upcross-sell

WhynotbuyB?

WhynotbuyA?

Directmails,emails,etc

Emailclicks,logins,etc

>Exercise:Targe/ngmatrix



70/77

>Qualitycontentiskey

AvinashKaushik:

Theprincipleofgarbagein,garbageout

applieshere.[whatmakesabehaviour

targe;ngpla


71/77

>ClickTaletes/ngcasestudy



72/77

Exercise:Tes/ngmatrix



73/77

Test Segment Content KPIs Poten/al Results

>Exercise:Tes/ngmatrix



74/77

Test Segment Content KPIs Poten/al Results

Test#1ANew

prospects

Conversion

formA

Nextstep,

order,etc ? ?

Test#1BNew

prospectsConversionformB

Nextstep,order,etc ? ?

Test#1NNew

prospects

Conversion

formN

Nextstep,

order,etc ? ?

? ? ? ? ? ?

>Exercise:Tes/ngmatrix



75/77

June2010 DataliciousPtyLtd

Dontwait

forbe;erdata,getstartednow.

75


76/77


[email protected]

Learnmoreblog.datalicious.com

Followmetwi;er.comdatalicious


77/77

Data>Insights>Ac/on

SuperIQ Analytics

Documents