Customer-Focused Strategy

Traditional retailers focus on common metrics, such as click-thrus, conversion, and cart abandonment. Often, however, too many retailers focus their efforts around only these metrics, defining elaborate plans to get more traffic, more people into the funnel, and hopefully more converters. So, rather than stepping back and asking some basic questions, such as, “Who is my customer?” and “Why does my site matter to him or her?” they blindly move forward with a strategy focused around metrics rather than customers.

I’ve done a great deal of work with media and entertainment companies, which tend to develop customer- or audience-focused strategies. If your revenue stream is selling advertising, getting more people to your site, getting them to engage more for longer periods of time, and getting them to repeat in a given month, then an audience-focused strategy is key. Although certain metrics such as page views, unique visitors, and video starts still make up a key part of the online marketing plan, growing audience segments that are valuable based on behavior, repeat visits, and time spent on the site are crucial to executing on any effective marketing plan.

So, where to begin?

Attribute Selection

Start with your most engaged visitors and discover which attributes differentiate them from other visitors.

Building a customer-focused strategy means understanding what makes valuable customers valuable. For instance, perhaps an engagement with a particular offer or product category will show value across a customer segment, but that engagement has a sharp correlation with customers under 50; or perhaps different behavioral attributes — such as social likes and tweets — correlate with a strong affinity with a particular product line. Developing customer segments with a focus on the behaviors and metrics that define those segments will help you create customer personas that you can target through specific campaigns, marketing channels, or product offers.

As an analogy, think of a salesperson in a store. A good salesperson can spot a customer and size him or her up pretty quickly. The salesperson can determine whether a person is likely to buy, can make a suggestion about particular product that might be of interest, and can spot people who are just browsing or killing time. We are trying to use digital data to size up the same thing in our online world by picking up on behavioral cues and creating a virtual “gut instinct” based on known correlations.

Correlation and Causation
Don’t assume correlation equals causation.

Segments are segments. They are helpful for understanding a customer set and what might motivate those customers, but in the end, they are just assumptions. You would never assume in a face-to-face sales scenario that you know what a customer wants; you might make an educated guess and ask questions accordingly, but in the end, you are making assumptions and expect the interaction to distill more data.

Perhaps in your data you identify a particular product category that seems to excite a set of customers, so you develop a program to target that product category across that customer segment, but it shows no meaningful lift. Does that mean that the segmentation strategy was wrong? Not necessarily. All that the segmentation shows is commonalities and differentiators. It doesn’t show true customer motivations. Your targeting program made some assumptions about what the data showed, and your assumptions were wrong – not the data.

So, the first thing to do is to implement the segmentation strategy across the entire targeting population and see which segments respond to which offers. This will give you a better understanding of motivators on a segment-by-segment basis.

Target the Past and the Future
Don’t target based only on past behaviors, but also on the next page.

If you knew what customers wanted when they arrived at your site, you would be golden. I want a watch. I land on Zappos, and presto: all its watches appear, already sorted to my interest. But, unfortunately, this intent-based mind reading is not possible, so we have to focus on what we do know. We know from the data what customers have looked at previously. We know how their interests are similar to and different from what others have looked at and converted. So, what is the “tipping point behavior”?

If you are clear about the tipping point behaviors that would put me in the watch-buyer segment, you can identify the behavior and automatically assign me to that segment when I exhibit interest. For watches, perhaps it’s as easy as clicking through from a paid Google search; you see that my referrer includes watches or popular watch brands, and immediately you promote me to a member of the watch-buying segment. Or perhaps I get there by viewing more than three watch brands across two visits. Whatever the determinate behavior, leverage what you know now based on where I am in my visit and target me immediately.

Better understanding your customers, their interests, and their motivations should have a direct impact on your bottom line. Driving conversion, after all, is key, but it may take time to have a meaningful and stabilized impact.

In Consulting, we’ve developed a number of tools and techniques for easily identifying what makes customers valuable and which activities draw them in. Analyzing those valuable customer segments and how they change over time can help you devise a strategy that will supply a wealth of effective, tactical targeting efforts that will ultimately help you achieve that goal.

Where’s My Marketing: Losing Opportunities

Last week I was talking to a colleague who travelled to Salt Lake City for the Adobe Digital Marketing Summit. He was telling me about this “pay-as-you-go” music service he really liked. His initial free minutes ran out just that morning. What we focused on was that when this happened, it just happened. There was no reminder about subscribing to the service or an email asking him to sign up. There was nothing.

Recently, I’ve been traveling to NYC multiple times a month, and I’ve been staying at a small hotel chain local to the city. I’ve always liked one of their 10 hotels better than the others. I, in fact, worked for the parent company 12 years ago and helped run marketing campaigns targeting people who fit the “present me”. So it was easy to pick a place to stay the three or four days every other week I found myself in NYC. Early on in my travels to the city I expected to receive something inviting me to bring my family to the city and stay the weekend. After all, if you are in hospitality in NYC a source of your incremental revenue comes from business people extending their stay, but nothing ever came. I received no email communication, even though I was insistent on leaving my email address. No other direct marketing ever showed up at my home. Like my friend, I was disappointed.

As we chatted about this phenomenon, we were amazed that when we actually wanted to be marketed to, nothing happened. It’s a strange disappointment in and of itself, and also a strange disappointment for us to be talking about this at a digital marketing conference. At Adobe, we spend a lot of time working with customers who want to target messages. They want to get the right message to the right people. Different organizations are concerned about879bfcf2f2e15f04a790385bb2b33ddc all the things that go along with not giving customers that “creepy feeling”, but in the end there are small windows of opportunity everywhere.

This at the heart is the “big data” problem. There are mountains of data and making any sense of it seems overwhelming. Customers often focus on the gaps. We spend hours talking about details without ever acting on a campaign or idea because it won’t be bulletproof. Suppose my favorite hotel just looked at my uptick in bookings and saw that I’ve stayed many times in past years and shot me an email about extending my next stay. It may take a few guesses to assume that I was in town on business, but there are some clues: My business email address, my corporate card, or just an inquisitive front office clerk who could help piece together the right message and right time.

Piecing the data together and making it actionable can make for huge opportunities. In the end, customers want marketing. They want you to remember that they booked at this hotel before, or purchased from this site before, or if they are reading their favorite online blog that they have a particular interest in other products and service that are contextually relevant. Using data to understand things about me isn’t terribly difficult. To understand things about people who are like me is more difficult, but making all of that something that is actionable in real time can seem overwhelming. But accepting that something close to perfect is better than nothing at all is a key first step towards success.

Refocusing on a Solid Foundation

In the world of digital marketing we tend to be myopically focused on the marketing while forgetting the technology that got us here.  Often, we talk a good game about optimization and lift while forgetting that it’s the underlying software automation that is vital. We love gadgets, but in the world of SaaS we too often see the geek who writes the code or deploys the tags as the roadblock. Yet the key to the success of any technology, no matter how simple or complex is a developer who understands the code base, one who can see the underlying software architecture, detect the necessary pattern, and deploy the change.

I love the game of Jenga. It’s a game building a higher and taller structure by stealing from the stability of that very same structure.  Eventually the tower collapses and often its the slightest movement that pushes it over. I think the game of Jenga is a good metaphor for what our customers (and in turn we) create when we push for rapid changes, updates in code or new product installs without fully realizing the scope of the change and what that addition mean to the platform. Obviously we steal resources, but more importantly we add another tag or integration to a platform that sometimes seems cobbled together at best.  We place pressure on resources to deliver and then aren’t surprised when they don’t.

Much has been written over the years about the huge time drain to maintain and enhance large software projects.  In graduate school we talked about the mythic 10 lines of code a developer can write a day.  The agile method is one of the most successful methods of software development.  It focuses on small incremental enhancements. Setting small gains to be made over a short window.

I am often times amazed by Web sites that receive millions of page views and visits per day but lack a good platform for long-term success. Technology is, after all, about repeatability and portability of functionality, yet we rip out and code without much regard. This is not to say that some customers don’t invest in the sustainability of their platform.  Many do, but too many under pressure to move quickly steal from what may already be a tottering foundation.

Any system is always under some stress and its organizations try to mitigate that stress through solid design and thorough testing. Visually it’s easy to tell if something doesn’t work. If I can’t log in or purchase a product, something is clearly wrong. However, if the wrong data is being collected it could lead to bad decisions and lack of confidence in the overall marketing strategy. I believe we need elevate the importance of solid QA around data collection. I like to call it data confidence assurance.
Good data is probably the most valuable asset of organization.  Understanding who my customers are and what they like to buy or the content they like to consume is of utmost importance so why then do we not place a higher investment in making sure that data is accurate?  Using tools like anomaly detection and testing scripts enterprises could move beyond the seat of the pants data confidence paradigm to something more mature and agile.