I’ve been experimenting for the past year or so with several proactive assistant apps to guide my day — they remind me to get on conference calls with clients, offer to text participants if I'm running late to an in-person lunch, and keep me in touch with friends and colleagues. Some of these apps also integrate Salesforce, Yammer, and BaseCamp for job-specific context and assistance.
Among the most popular apps, Google Now personalizes recommendations and assistance by applying predictive analytics to data stored in email, contacts, calendar, social, docs, and other types of online services users opt in. Other examples include Tipbit applying predictive analytics to make a more intelligent inbox, and EasilyDo using the notification system to recommend ways to automate common everyday tasks. Expect Labs is tackling this space from the other end of the spectrum, offering an intelligent assistance engine for enterprises to plug into and add proactive features to their own apps.
Here’s what we think:
• Vendors will experience burnouts and early customer frustration, much like in voice recognition. In the music industry, it’s said that an artist is only as good as her last hit. We saw that analogy apply to voice recognition when users got frustrated at Siri as soon as she failed once on them. Expect a similar dynamic with all types of predictive apps.
We attended the recent Glimpse Conference 2013, where members of New York's tech scene came together at Bloomberg headquarters to talk about social discovery, predictive analytics, and customer engagement.
Our key takeaway from the event: small, real-time data coming from very personal apps like email, calendar, social, and other online services will fuel next-level predictive apps and services. Specifically:
• Better insight doesn’t require more data; it needs the right data. Amassing large databases of customer profiles, purchase history, and web browser activity only goes so far, and is costing companies millions, if not billions of dollars every year. Mikael Berner from EasilyDo sees a new opportunity in better utilizing data scattered across personal email indices, calendars, social networks, and file and content repositories that directly indicate customers’ plans, interests, and motivations.
• Email, calendar, and location data is a goldmine for predictive analytics. Expedia or TripAdvisor can track web activities to recall a user searched for hotels last November and is likely to travel again this year, but a flight confirmation sitting in email or vacation time logged in calendar is a much stronger indicator of travel plans.
Last week I had the privilege of participating on the Advisory Board for the Retail Marketing Analytics Program (ReMAP) at the University of Minnesota, Duluth (UMD). Perhaps the best part of these sessions is the opportunity to meet with the students, many of which will be tomorrow’s marketing scientists.
During a few conversations on this visit, I was asked how to secure an entry-level position that would involve lots of cool predictive analytics. I want to focus on one of the answers I shared — don’t tell anyone you’re doing predictive analytics. What do I mean? Imagine you’re a freshly minted analyst in the following situation:
Your manager asks you to quickly evaluate who responded to a promotion.
You have many factors to investigate (because you have lots of data).
You have very limited time to find a great answer and build a deliverable.
The required deliverable needs to be simple and free of analytic jargon.
SAP today announced plans to acquire KXEN, a provider of predictive analytics technology. The terms of the deal are not known. This is an interesting development for both companies and highlights the focus on the democratization of predictive analytics, especially for marketers. The proposed deal puts the spotlight on two shifts in the analytics landscape:
Expert user to casual user. Our research shows that finding top analytics talent is a key inhibitor to greater customer analytics adoption. As a result, users expect analytical tools to cater to nontechnical, nonstatistician business and marketing users.
Developers And Their Business Counterparts Are Caught In A Trap
They swim in game-changing new technologies that can access more than a billion hyperconnected customers, but they struggle to design and develop applications that delight customers and dazzle shareholders with annuity-like streams of revenue. The challenge isn’t application development; app developers can ingest and use new technologies as fast as they come. The challenge is that developers are stuck in a design paradigm that reduces app design to making functionality and content decisions based on a few defined customer personas or segments.
Personas Are Sorely Insufficient
How could there be anything wrong with this conventional design paradigm? Functionality? Check. Content? Check. Customer personas? Ah — herein lies the problem. These aggregate representations of your customers can prove valuable when designing apps and are supposedly the state of the art when it comes to customer experience and app design, but personas are blind to the needs of the individual user. Personas were fine in 1999 and maybe even in 2009 — but no longer, because we live in a world of 7 billion “me”s. Customers increasingly expect and deserve to a have a personal relationship with the hundreds of brands in their lives. Companies that increasingly ratchet up individual experience will succeed. Those that don’t will increasingly become strangers to their customers.
Buy analytics software, hire marketing scientists, and engage analytics consultants. Now wait for the magic of customer analytics to happen. Right?
Wrong. Building a successful customer analytics capability involves careful orchestration of several capabilities and requires customer insights (CI) professionals to answer some key questions about their current state of customer analytics:
What is the level of importance given to customer analytics in your organization?
Have you clearly defined where you will use the output of customer analytics?
How is your analytics team structured and supported?
How do you manage and process your customer data?
Do you have clear line of sight between analytics efforts and business outcomes?
What is the process of sharing insights from analytics projects?
What type of technology do you need to produce, consume and activate analytics?
The Obama 2012 campaign famously used big data predictive analytics to influence individual voters. They hired more than 50 analytics experts, including data scientists, to predict which voters will be positively persuaded by political campaign contact such as a call, door knock, flyer, or TV ad. Uplift modeling (aka persuasion modeling) is one of the hottest forms of predictive analytics, for obvious reasons — most organizations wish to persuade people to to do something such as buy! In this special episode of Forrester TechnoPolitics, Mike interviews Eric Siegel, Ph.D., author of Predictive Analytics, to find out: 1) What exactly is uplift modeling? and 2) How did the Obama 2012 campaign use it to persuade voters? (< 4 minutes)
The deluge of customer data shows no signs of abating. The perpetually-connected customer leaves data footprints in every interaction with a brand. This presents tremendous opportunities for customer insights professionals and analytics practitioners tasked with analyzing this data, to not only get smarter about customers but ensure that the insights get appropriately used at the point of customer interaction.
When we asked customer analytics users about the challenges and drivers of customer analytics adoption, we found that data integration and data quality continue to inhibit better adoption of customer analytics while users still want to use analytics to improve the data-driven focus of the organization and drive satisfaction and customer retention.
Forrester’s Customer Analytics Playbook guides customer insights professionals, marketing scientists and customer analytics practitioners into this new reality of customer data and helps discover analytics opportunities, plan for greater sophistication, take steps towards building a customer analytics capability and continually monitor progress of analytics initiatives. It will include 12 chapters (and an executive overview) that cover different aspects of customer analytics.
Why? What organization couldn’t benefit from making better decisions? Just ask the Obama campaign, which used sophisticated uplift modeling to target and influence swing voters. Or telecom firms that use predictive analytics to help prevent customer churn. Or police departments that use it to reduce crime. The list goes on and on and on. Virtually every organization could benefit from predictive analytics. Don’t confuse traditional business intelligence (BI) with predictive analytics. BI is about reports, dashboards, and advanced visualizations (which are still essential to every organization). Predictive is different. Predictive analytics uses machine learning algorithms on large and small data sets alike to predict outcomes. But predictive is not about absolutes; it doesn’t gaurentee an outcome. Rather, it’s about probabilities. For example, there is a 76% chance that this person will click on this display ad. Or there is a 63% chance that this customer will buy at a certain price. Or there is an 89% chance that this part will fail. Good stuff, but it’s hard to understand and harder to do. It’s worth it, though: Organizations that employ predictive analytics can dramatically reduce risk, disrupt competitors, and save tons of dough. Many are doing it now. More want to.
Few understand the what, why, and how of predictive analytics. Here’s a short, ordered reading list designed to get you up to speed super fast: