Post on 06-Aug-2015
Ernest Oppetit
Product Manager Qubit
Robin Trickett
Senior Strategist Qubit ‣ The acquisition challenge
‣ Using data to focus acquisition efforts on predicted ‘quality’ customers
‣ Demonstration using Qubit’s tools
‣ Q&A
Agenda
Qubit is a global leader in digital optimisation
• Segmentation
• Analytics
• Personalisation
• Testing
Profitability is driven by focusing acquisition investments on those customers most likely to generate high lifetime value
Acquiring customers can require significant investment1
Some customers will go on to renew/increase share of wallet, others will churn shortly afterwards
2
Acquisition success should be measured on lifetime value, as opposed to initial conversion value
3
Two challenges: i) How should ‘predicted high lifetime value’ customers be identified
before they’ve even converted? ii) How is their revenue opportunity then maximised?
The revenue opportunity of ‘Predicted High Lifetime Value’ customers can be increased using 5 levers
Pre-siteA On-siteB Post-purchaseC
PURCHASE CYCLE
REVENUE LEVERS
QUBIT APPLICAT-IONS
Increase volume arriving on site
1 Increase conversion rates
2
Increase conversion values
3
Increase re-purchase / renewal rates
4
5 Increase re-purchase / renewal values
• Attribution modelling
• LTV modelling • Audience
segmentation • Personalisation
& testing
• ESP integration • Audience
segmentation • Personalisation
& testing
Pre-siteAIncrease volume arriving on site1
‣ The objective is to increase the volume of visitors arriving on site who are most likely to go on and have the greatest lifetime value
‣ This requires attribution modelling to identify those channels which fulfil this requirement
This attribution model needs to allocate a behavioural lifetime value to each channel, and if appropriate, keyword
1. Identify those channels, campaigns and keywords which generate greatest lifetime value - Value should be
measured using a behavioural model as opposed to first click or last click
2.Optimise marketing spend accordinglySEM SEO Affiliate Social
LTV (behavioural) First clickLast click
CHANNELS
KEYWORDS
PRE-SITEA
On-siteBIncrease conversion rates2
‣ Step one is to map out a profile which defines what a prospect looks like who is likely to go on to have a high LTV
‣ Step two is to identify these visitors when they arrive on-site
‣ Step three is to deliver personalisations to these visitors to increase their conversion rates and values
Increase conversion values3
Firstly, a profile needs to be defined which can be used to identify predicted high LTV customers when they arrive on-site
ON-SITEB
Identify pre-conversion characteristics of high LTV customers
Create a profile to enable the probabilistic identification of these visitors
Arrival data Site footprint Provided data
• Channel • Keyword • Device • Location • ……..
• # sessions to purchase
• # pageviews to purchase
• Page types viewed
• ……..
• Quote value • Marital
status • Age • Occupation • ……..
• Attribute of a potential high LTV customer
VISITOR PROFILE: Predicted high LTV customer
+ +
Decipher demo -
http://www.qubit.com/webinar/focus-acquisition-on-quality-over-quantity
On-siteBIncrease conversion rates2
‣ Step one is to map out a profile which defines what a prospect looks like who is likely to go on to have a high LTV
‣ Step two is to identify these visitors when they arrive on-site
‣ Step three is to deliver personalisations to these visitors to increase their conversion rates and values
Increase conversion values3
Secondly, once a predicted high LTV visitor is identified on site, a range of personalisations can be leveraged to increase conversion rate & value
ON-SITEB
50% match
Levers to increase conversion rate
Pre-populate form fields
Offer discounts
Levers to increase conversion rate
Re-order high value items
Pre-populate upsell items
‣ The accuracy at which a visitor can be matched increases with the number of sessions
‣ The closer the match, the more levers available to increase conversion rate & value
• Attribute of a potential high LTV customer
…….
…….
Secondly, once a predicted high LTV visitor is identified on site, a range of personalisations can be leveraged to increase conversion rate & value
ON-SITEB
50% match
Levers to increase conversion rate
70% match
Pre-populate form fields
Offer discounts
Levers to increase conversion rate
Re-order high value items
Pre-populate upsell items
‣ The accuracy at which a visitor can be matched increases with the number of sessions
‣ The closer the match, the more levers available to increase conversion rate & value
• Attribute of a potential high LTV customer
…….
…….
Secondly, once a predicted high LTV visitor is identified on site, a range of personalisations can be leveraged to increase conversion rate & value
ON-SITEB
50% match
Levers to increase conversion rate
70% match >90% match
Pre-populate form fields
Offer discounts
Levers to increase conversion rate
Re-order high value items
Pre-populate upsell items
‣ The accuracy at which a visitor can be matched increases with the number of sessions
‣ The closer the match, the more levers available to increase conversion rate & value
• Attribute of a potential high LTV customer
…….
…….
Deliver demo
http://www.qubit.com/webinar/focus-acquisition-on-quality-over-quantity
Post-purchaseCIncrease re-purchase/renewal rates4
‣ Once the initial investment has been made in acquiring a predicted high LTV customer, further efforts should be employed to ensure they realise this potential
‣ This can be in the form of encouraging them to re-purchase/renew their subscription, or in the form of upsell and/or cross-sell
Increase re-purchase/renewal values5
Subsequent website visits provide an opportunity to target these high LTV potentials to drive further value from them; reinforce through email
POST-PURCHASEC
Target on-site during return visits Follow-up with targeted emails
Campaign mirroring
•Increase renewal / re-purchase rate ‣ Content recommendations ‣ Renewal offers1)
1) See churn reduction webinar for more details
•Increase renewal / re-purchase value ‣ Upsell and cross-sell recommendations,
personalisations and offers
•Highly targeted emails can be sent to pre-defined segments, reflecting the exact content they were viewing on-site
•If customers then return through email, their immediate on-site experience should reflect the content and nature of the email
i ii
iii
In summary, strategies should be in place across the purchase cycle of predicted high LTV customers, in order to maximise their value
Pre-siteA On-siteB Post-purchaseC
PURCHASE CYCLE
REVENUE LEVERS
STRATEGIES
Increase volume arriving on site
1 Increase conversion rates
2
Increase conversion values
3
Increase re-purchase / renewal rates
4
5 Increase re-purchase / renewal values
• Use behavioural attribution models to determine which channels are the best source of high LTV customers
• Identify the pre-conversion characteristics of high LTV customers in order to target this segment with personalisations and offers
• Once these customers have been acquired, ensure their value is maximised through further site and email personalisations