Big Data 101: Use Web Data to Understand Customer Intent

“What do our customers really want?” We started off a communications strategy discussion this morning with that question. It’s the same question, frankly, that every marketer should be asking when they wake up, throughout the day, and before every project or marketing campaign is developed. Coming to our aid is the promise of “big data” analytics – that that we can now better understand customer intent, and not just customer behavior.

It had been nearly impossible to gain such insight in the past. However, web data is an excellent starting point to understand the power of using big data analytics to drive business value. It’s one of the most accessible sources of big data and pretty universal across industries.

Many marketers have already integrated detailed, customer-level behavioral data sourced from a website into enterprise analytics environments and marketing automation solutions. Typically, this involves transactional data, and push messaging. It’s not magical that we often see:

• I get a follow-up email message offering suggestions for accessories after buying a little black dress from Talbots online.
• After browsing my bank’s website for refinance options, I get a phone call from an agent in my area offering help and information.
• The newsletters I receive start to have more of the content I like and less of that stuff I usually skim over, along with adverts for things in my area and aligned to what I read most.
• Three days after I downloaded a whitepaper off a corporate website, and then tweeted about it, I get a tweet from the author asking me if I have any questions. The same person also emails me a link to a related whitepaper.

These are great automation strategies, and I encourage everyone to continue them. Certainly, they pull us forward from traditional web analytics, which was aggregated and offered only a rear view mirror perspective. However, these approaches are old fashioned in a big data world. They are only using data from the “last mile” transaction, and ignoring the customer journey information. It’s that journey data that can help us understand intent.

Imagine what an airline could do to improve your yield and pricing optimization if its team knew exactly which types of combinations of airports, flights, and originate-return combinations someone looked at before they selected one. Imagine if you could see if price, or airport, or seat availability was the determining factor in selecting or abandoning each choice.

If you are an airline, you already have that data – it’s all in the weblogs of the transactions for each customer. Yet, every time we visit an airline or travel site, consumers muck through the same steps as if our past actions were a deep mystery. Instead of just seeing the results, marketers now have visibility into the entire buying process. This is a big data source with huge analytics and automation potential.

How can you take advantage of this powerful big data source? Consider these important steps:

1. Collect the right data. Any action taken should be captured if it’s possible to capture. This includes data from kiosks, call centers, and mobile apps, as well as web data. Consider: purchases, product views, shopping basket additions, video views, downloads, help requests, registrations, comments, forwarding a link, reading/writing a review, searching, and more.
2. Don’t be creepy. Please consider the consumer perspective as you use data – any data. Privacy is not just hot for Congress right now (see updates on the current Senate inquires here), but also for your customers. Don’t bury your intent in the privacy policy, although it’s within the letter of the law. Make it clear early and often how you will use data for the benefit of customers. Most analytics work can use “faceless” data – which is not aggregated, but individual, and uses some arbitrary and non-traceable number (not customer PII). For follow-up messaging, ensure you have responsible notice, consent, and opt-out practices in place.
3. Identify a starting point. For transactions and shopping, consider where the experience starts. Is it on the search engine or email message, or is it on the landing page? Often, the path to the source prior to the website visit is crucial to understand intent. Were they on a competitor’s site? Were they reading an industry article? Were they searching for free shipping offers?
4. Investigate patterns. This is what analytics people do best. They help us learn what customers do. Who arrives and leaves, and where? Who reads or writes product reviews? When did they look at shipping information? What products are being “compared”? What are the product bundles reviewed prior to a purchase? Do “recommendation engine” suggestions get added to the cart – and then abandoned or purchased? Use these web data patterns to make more robust “next best offer” decisions, as well as more customer-centric and accurate attribution, attrition, and response models.
5. Do your research in the data. Forget expensive, small-scale surveys where people say what they think they will do. Use your web data to assess what people actually do. Look for patterns in segments as well as feedback behavior – do reviews feature certain product characteristics? Highlight those in your marketing and up-selling. Do certain types of customers always write a review or post a comment? Celebrate them and make them special. Note, of course, that not every influencer will be a “most valuable” customer. Web data can also improve your segmentation strategy, because it lets you segment on how (and potentially why) customers shop, not where they shop or who they are.

Now, the airline can identify ways that customers arrive at their purchase decision, and the factors that go into decision-making. In short, we start to understand intent. Am I a convenience or price shopper? If the cheaper flight gets in two hours later, but saves me only $100, will I go for the price or the convenience? Perhaps that extra two hours of sleep is worth $100 to some convenience customers. Now the airline can make my next purchasing experience more successful and enjoyable by offering options that meet my intent profile. It can also start to email me offers that make sense based on how I like to travel.

The new reality is that our customers’ “now” is always evolving. Any historical view is just that – history. Just when you think you have it figured out, the world, the customer, or the competition changes.

However, in web data we have a pretty cool – and frankly, rare – opportunity. We don’t often have the chance to capture this detailed information at the distinct customer level – and modern analytics and automation technology make it possible to utilize this data in creating better customer experiences (and increasing revenue). Plus, once we set up our analytics machine, we keep this customer view updated over time.

How are you using web data to improve your customer experiences – and increase revenue? Is web data a viable starting point for your marketing automation plans? Please let us know in the comment section below.