Customer Journey Analytics

Customer Journey AnalyticsSentiment Symposium2017 2017 Beyond the Arc, Inc.

The main points Make a journey map (an actual picture) Validate it with business metrics and analytics Is each step real? (Customer point of view)Is each step measureable? (Business point of view)Use predictive analytics to find causes and message points Get customer feedback at each point SurveysPredictive analytics shows key drivers tied to desired outcomes Track business metrics While keeping in mind that customer perceptions may not alignwith internal business statements and objectives2

Why link journey with analytics?The customer journey is important, but not enough Customers have always had experiencesWhat steps make up the experience? What steps lead to positive emotions?What steps do we think lead to greater lifetime value?Business analytics are important, but not enough Businesses have always had operations and analyticsCounts and sums are simple analyticsCustomer journey analytics are strategic and tactical Detailed personal data is more widely availableBusinesses can create experience platforms that tie together the journey and the analyticsEach customer’s individual pathway can now be addressedPredictive analytics identify personal targeting and important message levers3

Big data makes it personal As the data gets bigger, itcontains a higher volumeof more specificcustomer interactions General advertisingSegmented marketingPersonalized contactIndividually targeted andcustom messaged4

Not always linearAlthough the map may be drawn in a line The customer journey is not always linear Has multiple points of interaction Has multiple ways of interacting with the customer5

Many paths lead to the same endpoint6

3-D customer journey analyticsCustomerJourneyMappingName the steps inthe journey andmake a pictureCustomerJourneyAnalyticsUse data tounderstand causesof behaviorBusinessMetricsBridge business andcustomer viewsusing analytics7

It’s the customer’s journeythat creates the customer’s experienceThe cumulative effect of all touchpoints in the journey isthe experience, and it matters to the customers To maintain this, the company needs to keep track of The customer point of viewThe customer journey touch pointsAnalytics that uncover causes of behaviorThe business operations that support the touchpointsAnd what it takes to make the business happen8

Organizational support required to iterate Using analytics to support business requires strongorganizational supportCreate repeatable processDefine actions and trackresponsesKeep models and data up-todateStay linked to the customerexperienceAnalyticsto findsegmentsUpdatedata inthesystemMonitorCraftcustommessagesA/BTesting9

Maximizing the value of the customer journey What are the biggest points of pain for the customeralong the journey? Lots of complaints, frustration, frictionVery emotional conversationsAnger What are the risky spots for the business? Points of greatest profitPoints of greatest lossLoss of lifetime value10

Choosing where to startFocus first oncustomers Customer Pain Involving more- Resolving higher pain Resolving higher lossPain and Profit-Business Profit 11

Gather feedback widely to avoid blind spotsAsk early, ask often At beginning During the journey How well did that step work foryou?At completion of a path in thejourney How did you hear about us?Use differing methodsEmail surveys To everyone To a random sample Response boxes on websites Response links on store receipts Phone calls How likely is it you wouldrecommend us to a friend?Why did you leave us?Ask continually And track trends over time12

You don’t know what you don’t know Ask open ended questions Let customers share theexperience from their perspective Ask “Why” questions Let customers tell you what they thinkhappened Ask more than one kind of question Let customersdescribe different aspects of the experience Don’t get caught up in a single metric like NPSTrack it over timeAlso ask “Why?” Ask continually because As the world turns Experienceschange 13

Use analytics to tie friction to value The squeaky wheel gets the grease. The one whoraises a hand gets answered. But most wheels aren’t squeaking. Most people don’traise a hand.And they might be perfectly fine. Use sentiment analysis combined with businessmetrics to put customer comments in context. Emerging trend?Bad apple?Valuable segment?Suggestion for improvement?TrollingCompliment?14

Sentiment Analysis API classification Custom models WatsonIBM BluemixAmazonGooglePython NLTKStanford NLPHybrid systems15

Analyticscan insert new insightsat every step of the journey16

A customer journey is an analytics journeyCustomer journey mapBusiness analytics mapSee online advertisementCreate demand, which segments to targetPurchase productAcquire customers, cost per customer,Track best campaigns, best segmentsReceive welcome letterOnboarding customers30-day / 90-day expected behavior modelGet weekly email, visit websiteSend and track communicationsBehavioral trackingMake more purchasesTrack salesKPI dashboardPredict and measure customer lifetimevaluePurchase in response to an ad about arelated productCross-sell models target related itemsUpsell models target more profitable items17

A customer journey is an analytics journeyCustomer journey mapBusiness analytics mapInteract with customer serviceCustomer service staffing, hot topics, timeto get to the right answerReceive satisfaction feedback surveyCustomer satisfaction measurementExperiences with happy employeesEmployee satisfaction and managementRefer a friendCustomer loyaltyMake more purchasesTrack salesKPI dashboardPredict and measure customer lifetimevalueSwitch to a competitor, or reduce spendwith usCustomer retention models, messaging,actionsReceive email with coupon and switch backWin-back models and targeted campaigns18

Analytics inspire strategy, inform tacticsStrategic AnalyticsTactical AnalyticsWhich touchpoints does the currentcampaign address?Which messages got the best openrates?What factors predicted success atthis step?How many people responded to thelast campaign?What were the top reasons for eachcustomer action?Was the campaign profitable?What did we learn that we can useto improve the business?19

Acquire customers: be careful with incentives Use the customer journey map to verify the acquisition plan Descriptive data is used to buildmodels showing best prospects Initial targetingIncentives may not be needed,or may not need to be as highRetargeting if not acquired 2nd-best prospects May need more incentiveLift models take into account cost of acquisition and expectedprofitability21

Acquire customers:by finding patternsMarried peoplewho do nothave our creditcardSingle people who do not haveour credit card and who havepaid greater than 20% of theirloanPeople who have 2000 or less on ourcredit card, who also come to thebranch (rather than banking primarilyonline)22

Acquire customers: by targeting them A campaign manager might select likelihood values from80 to 100, thus taking the top 20% of prospects for thecampaign. Or a campaign manager might target by segment.23

Acquire customers: by testing messagesA/B TestingSimple: Compare your current message (A)ABto a new one (B) Use the most popular oneMore sophisticated: Vary the messaging by ChannelChannel CombinationSequenceSeason or timeframeAB24

Onboarding customers Map the customer journey What steps lead up to acquisition?What steps are critical to onboarding?What have focus groups and surveys revealed? Find the causes behind purchases in the first 28 days (4 weeks) Based on acquisition dataIf they don’t behave as expected, use more aggressive engagement strategy Find the causes behind purchases in the first 91 days (13 weeks)Find the causes of dissatisfaction in the first 91 days Based on acquisition data, 30-day prediction, and first 3 months of behaviorIf they don’t behave as expected, use more aggressive communications strategyBase communications on causes identified in the model and from the mapAddress the gaps in their behavior25

Routine touchpoints Monitor each point in the journey map for How recently? How often? How many per month? Monetary value Emotional value Data gathered at routine touch points feedsanalytics to Describe the customers we havePredict the customers we might getFind gaps between what our journeymap suggests and what people really do Track customer participation in the business.How often do they Use the servicePay on time, and when did they payPay late, by how many days, at what point in themonthRenew subscriptionsMake new purchasesInteract with customer service or sales Keep historical snapshots of the data26

Sales Verify that the journey map reflects real behavior Allow for looping and skipping steps Get feedback from customers about their experiences What was the experience of making a purchase like in their own words?Were some things more difficult than they thought they should be? Use tactical analytics for dashboards Assess your current strengths against goals and prior performance Use predictive analytics to Find best prospectsFigure out who should have purchased, but didn’tProvide custom messaging based on suspected causes27

Cross-sell, up-sell Increase customer lifetime value by broadening thescope of their purchases Business analytics find higher margin products Predictive analytics find Products that customers buy togetherCustomers who have one product but don’t yet have the secondRanking which customers are most likely to get the secondproductMost important variable predicting the behavior that can beused as levers in messaging.28

Customer lifetime value Identify your best customers, then treat them better Decile them to find the top performers1234Value of decile5678910All deciles treated equally29

Customer serviceNo business is perfect Treat valuable customers well Track what customers are talking about Shorten the time it takes to get to the right answer Use text analytics to automatically extract the hot topics Shorten response time to problems in the field30

Employee satisfaction Grumpy employees create bitter experiencesAnalyze internal satisfaction the same way you measureyour customer journeyThis could be a big blind spot in your company, thatcould lead to a lot of friction in your customer journey So don’t shy away from it31

Analytics on the employee side32

Customer retention and churn Make a plan in the customer journey map for how tohandle customer departure How do you know when they’re gone?Canceling a subscription is obvious, but what about retail? Predictive models can tell who is likely to leave in thenext 28 days, and why Prepare retention offers for those likely to use a call centerPrepare incentives for customers who have “expired” Use win-back predictions; customer acquisition modelsthat also know your past behavior33

Add analytic insights along the journey Validate the journey map with business metrics and analytics Get customer feedback at each point Is each step real? (Customer point of view)Is each step measureable? (Business point of view)Use predictive analytics to find causes and message pointsUse open ended questions as well as structured metricsDon’t get caught up on using a single metricCapture the “why”Use analytics to Systematically quantify what customers say in their own wordsFind trending topicsUnderstand cause and effect34

Thank youSteven J. Ramirez, CEOBeyond the Arc, om/beyondthearcSlideshare.net/beyondthearc36