In this article we’ll start by discussing the importance of leveraging artificial intelligence and machine learning (AI/ML). Next we’ll review benefits and real-world success stories from global customers using Virtual Analyst and Segment IQ to detect anomalies, avoid outliers, and maximize ROI.

The Value of Artificial Intelligence

Perhaps you recall the time chess champion Garry Kasparov was defeated by IBM’s Deep Blue. Experts were convinced a machine could not possibly beat human decision making and judgment in a game as complex as chess. Yet, it was done and this was key to a fundamental shift in business strategy and technological innovation as the power of artificial intelligence was unearthed.

Adobe Analytics is the core system of intelligence for the experience business, enabling anyone in the enterprise to understand and optimize customer interactions with their brand across all touchpoints in real-time and at massive scale.

Adobe’s AI tools are not here to replace you, but rather to enable you to achieve maximum ROI on your efforts.

Transform Your Analytics

To evolve your analytics, we need to focus on three key considerations:

1. Organization - How to create holistic views of customers, prioritize insight-driven decisions and democratize data
2. Technology - How to make sure data and technology deliver personalization at scale
3. Customer - How to build trust and adapt to change

Analytics is challenging and time-consuming, yet there is a constant need to accelerate time-to-insight. Key issues organizations face include:

• Limited organizational resources - various business goals can limit resource availability
• Limited technical expertise - can knowledge be shared and data democratized
• Customer expectations - can your team react dynamically to customer behaviors changes

Master Customer Intelligence with Virtual Assistant, powered by Adobe Sensei

3 Levels of Customer Intelligence

When it comes to a successful customer intelligence strategy, we need to move through three levels (see Figure 1 above) from: (a) data collection, to (b) data processing, to (c) analytics and machine learning, before we can finally take action and optimize our content and ads.

• (a) Data collection will depend on your organization and may include a variety of channels and mediums, such as OTT devices, video, enterprise, call centers, in-store, social email, web, ads, mobile apps, wearables, IoT, voice assistants, connected cards, and geo / spatial.
• (b) Data processing includes real-time data collection, processing rules, audience syndication, context-aware sessionization, real-time triggers and views, and platform.
• (c) Analytics and machine learning includes Segment IQ, Virtual Analyst, Segmentation, Analysis Workspace.

Leverage your Virtual Analyst

Think of the Virtual Analyst as the rockstar analyst who:

Discover hidden opportunities

• Gain up-to-the-minute visibility into status of marketing KPIs
• Make good marketing investments reliably and predictably
• Keep up with and meeting / exceeding customer expectations

Real World Success

The virtual analyst uncovered the following scenarios for real Adobe customers:

Campaign Termination – $1.7m daily increase in revenue driven primarily by a campaign that had been ended prematurely.

Vendor Bug – 73% increase in cart removals due to a tag manager bug that was auto-removing certain products from the cart.

Browser Issue – 8% increase in cart abandonment associated with Chrome browsers. This fix led to a $1.2m daily incremental lift in revenue.

Coupon Fraud – Identified an 81% spike in orders caused by traffic referred by two major daily deals / coupons sites promoting fraudulent toiletry coupons. These orders were able to be canceled.

Corporate Espionage – 200% increase in visits caused by bot/crawler created by their main competitor to scrape content off their site for reuse. These IPs were able to be blocked.

Adobe Analytics Capabilities

• Anomaly Detection:

  • Leverage built-in predictive algorithms to help identify spikes and dips in your data that you didn’t know existed
  • Make use of 28 unique algorithms are used to identify anomalies, including seasonality, growth, and cyclical models, as well as holiday alignment.
  • Reduce reliance on data scientist and unlock citizen data scientists’ capabilities

• Contribution Analysis:

  • Rapidly identify contributing factors that caused significant changes in your data
  • Save countless hours searching for explanations to changes in metrics
  • Leverage powerful machine learning designed to transform the analyst and marketer into a data scientist

• Intelligent Alerts:

  • Stay informed of anomalies within your data at all times – whether you’re in the office or on the move

    • Create alerts directly from Analysis Workspace
    • Base rules on anomalies (90%, 95%, 99%), % change, and above/below
    • Use Alert Preview to see how often an alert would have triggered
    • Leverage SMS and email support with links to auto-generated Analysis Workspace projects

• Segment IQ:

  • Discover the differences and overlaps between your segments to inform your segmentation strategy
  • Uncover key characteristics of the audience segments that are driving your KPIs
  • Get reports and visualizations within seconds to minutes that display commonalities across dimensions, metrics and other segments
  • Improve engagement with high-value customers

Real-World Success with Segment IQ

Mobile vs Desktop: “We compared hits from one of our sites to another site and quickly found a bunch of tagging inconsistencies.” → Avoid data problems before a product release

Feature Usage: “Customers who used our product comparison feature were 10% more likely to convert. Moving it to the top of the page grew orders.” → 4% increase in conversion

Content Engagement: “ We discovered that visitors to our news section were twice as likely to watch video ads, so we added more video options to that section.” → 7% increase in video ads viewed

Paid Search: “Visitors coming from search engines were 3x more likely to upsell. We upped our spend on specific keywords as a result.” → 56% upsell lift

Product Stock-Out: “People buying Fitbits were 6x more likely to get an ‘out of stock’ than everyone else, so we quickly ordered more Fitbits.” → Stock-outs prevented and more holiday orders completed

For more information watch our webinar here