Using Web Analytics to Create a Content Strategy

Understanding how and why customers make decisions is the fundamental quest of digital analytics—and the most elusive. We often get so focused on how to optimize a piece of the website that we lose sight of how that fits into the bigger picture. Yet if we understand how customers make decisions about our company we will know how to help them.

These four steps have helped me in multiple situations to create a customer-centric content strategy. Armed with this kind of framework you can stop playing with window dressing and instead address the fundamental questions about how to interact with customers.

Customer Needs and Decisions

If we hope to look at customers as people and the selling process as a discussion, we need to understand what customers really want. We aren't trying to trick people into buying what we sell. They aren't going to accidentally buy if we just present them with the right call to action. Our objective is to convince them that doing business with us is in their best interest. And we can't do that without knowing what they need.

We can make this list by looking at what customers actually ask. Do a holistic analysis of all the input they give us: internal and external search terms, survey responses, marketing messages they respond to, etc. List all of the recurring questions and needs. Then compare all of this to the types of content they actually engage with, not just their landing page. You should start to see patterns in what customers are looking for. Realize that customers need to make a set of decisions in some order, so their questions and needs will change as they progress.

The table to the right is an example of what part of this might look like for B2B customers. This will be a variation of a classic customer decision process, but it will be more granular and specific to your business and your customers. Think of it as a purchase funnel—a really big purchase funnel.

In doing this recently for Dell, I also used to uncover this decision process. I created tasks focused on drawing out the questions that users had at different points and where they would expect to find answers. It was an invaluable way to validate some of the trends we saw in the data. Focus groups or other types of usability studies could accomplish the same purpose.

At the end of this process, you should be able to make an ordered list of the main decisions a customer needs to make and what things you can show them to help them make that decision.

Customer Expectations

Once we understand the customer's questions, we need to create resonance. In other words, we need to address their needs where they expect us to.

Sometimes we recognize at some intuitive level what questions our customers might have, and then we pack all the answers into a messy ball and hand it to them. Customers engage with us in order to get information, so they are not averse to content-rich sites. The problem is when we overload them with information that is irrelevant at the moment.

After ordering and grouping the decision process, it's an easy step to find where customers expect to get information. The table to the right is specific to a website, but if you are leveraging big data, the same process can be used to map out all the vehicles you use. The point is to identify what a customer expects to learn at each point. Since the customer's decision process is at least partially unconscious, we need to show them what they are (unconsciously) expecting so that they don't have to start searching for it.

Do another round of usability tests to clearly map this out. Also, group search terms and survey results by the decisions and look at the areas of the site customers reflexively go to. This will probably uncover discrepancies between where they look and where the answers actually are.

Identify KPIs

Without thinking about the data you do or don't have, go step by step and list the things a customer should do if they've answered each question sufficiently. The actions will get more specific as you get further down the decision process.

Repeat the process by making a list of the things a customer might do if their question hasn't been answered well enough. Your framework should look something like this by now:

Now list the metrics that would represent each of those actions. In a true big data environment, you will be able to get data for almost everything you listed. Otherwise, you will need to identify proxy metrics or just leave some of them alone.

The Right Marketing Activities

It bears repeating, though, because it's the foundation of this project: Every interaction with a customer should be measured by whether it meets the customer's need at that point. If it doesn't, it creates dissonance. It frustrates the customer and gives the impression that you're ignoring their interests.

List all the different ways you try to interact with a company. It might be different elements of the site (banners, videos, customer testimonials) or it might be different marketing vehicles (email, Twitter, sky writers). Then use the KPIs you identified to measure which of those is most effective at each point in answering that specific question. If there are six vehicles that effectively address the same question, use the KPIs to determine which addresses the need best and kill the other five.

You will undoubtedly find that certain customer questions aren't being effectively addressed by anything you are doing. When that happens, you can send me flowers. This framework will give you tactical information to help you design a new initiative to meet that need.

And now you will know where to focus your optimization efforts, where to run more tests or try new tactics. You will know what needs to be completely revamped and what just needs some minor tweaks.

No comments:

Post a Comment