What data do you have?

A good first task is to take inventory of the customer data you already have. Even the smallest business has data, but it might be residing in a surprisingly wide variety of business systems. Customer contact information is essential, of course, but you might be able to make use of other kinds of data, such as purchase history.

Cast a wide net when listing your possible data sources. Don't limit yourself at this point; you pare the scope of the data later. Don't forget to consider the following locations:

• Customer relationship management (CRM) system: For example, an automotive service shop probably has contact information, information about the vehicle, and a history of services.
• Point-of-sale or billing system: If you collect information from your customers as part of the sale, your point-of-sale or billing system can be a rich resource of customer data.
• Existing communications methods: For example, if you've ever set up a form for customers to sign up for a newsletter or request more information, that form has been collecting valuable data you can use.
• Loyalty program: If you offer rewards to your customers for their continued business, don't forget to mine the system you use to administer it.
• Customer preference center: Your existing content publishing processes might have already inspired you to set up a website where customers can indicate preferences, such as what topics they're interested in and how often they want to receive messages from you.
• Website analytics: The analytics tools you install on your corporate website can provide an amazing amount of information about the people who visit your site, even if they don't make a purchase. Be especially careful with how you use this data, though, because it's easy to jump to the wrong conclusions about what your visitors want and end up upsetting prospective clients before they even have a chance to buy.
• Spreadsheets, file cabinets, and stacks of business cards: Think about the life cycle of a customer interacting with your business and check out all the locations where the interactions might be stored.

Figure 5-1 illustrates the information you might have about clients and how you could store that information in Marketing Cloud.

You must have explicit permission from your customers before you send them marketing messages. If you've already gathered that permission through any of these channels, be sure to make a note of it. If you haven't, you'll have the opportunity to ask for that permission when you begin your campaign.

What data do you wish you had?

As a marketer and a businessperson, of course you would like to know everything about your customers. But the question of “What data do you wish you had?” isn't about curiosity. It's about the specific data you need to power the marketing campaigns you plan to execute.

“What data do you wish you had?” sounds like a straightforward question, but it can quickly turn in to a chicken-and-egg situation. You need to predict what data your future online marketing will rely upon, but the needs of your future campaigns will depend in part on the data you have.

Regardless of what other data you need, you're definitely going to need data to segment your subscriber lists into groups to which you send your different messages. After you have that, you can use more data to choose the content in the message on a subscriber-by-subscriber basis.

Data for list segmentation

You use some of the data that you collect specifically to choose which subscribers to send messages to. Figure 5-2 shows a sample data set of subscribers you might send to.

Logical groupings depend on your business (we know we say “it depends” a lot, but it’s just the reality of online marketing). Natural ones we've used in the past include the following:

• Geography: This can be by ZIP code, state, or country, or even hemisphere! You'll need the data to support your audience groupings, whatever they are.
• Class of customer: You might have a loyalty program that has different tiers, like an airline has gold and platinum members. You might want to send a specific email to people who are basic members, so you would use your membership-level data to send to just this audience.
• Active subscribers: You can leverage the click and open data within Marketing Cloud to target your most active subscribers, such as those who have opened or clicked in the last 60 days.
• Specific interest: If your preference center has interests that people can select, and you have content specific to that interest, you can use that information to target the audience.
• Recent purchases: You can leverage your sales data to follow up with your customers, encouraging them to complete website feedback about the items they purchased or write a review about your business. Customers are more likely to give positive feedback if you remind them.

One phrase we hear too often in email marketing is “email blast.” Compared to most other types of marketing, email marketing is precise. It gives you enough customer data to target both the subscriber and the content. If you're taking advantage of your tools and data, your email marketing could never be described as a randomly scattered blast.

Data to choose message content

One of the most underutilized features of Marketing Cloud is the capability to change what content you include in an email based on data about the specific subscriber.

Simple personalization, such as including the subscriber's first name, takes little effort. With more elbow grease and the right data, though, you can create a single email send that selects content tailored for each individual subscriber. One example appears in Figure 5-3.

For example, you could use this capability for the following sample cases:

• Geography: Let's say you have an offer that is going to become available at the same time across the country. You can use the subscriber's time zone to include the local time when the offer will go live.
• Class of customer: You might want to show subscribers how many more points they need to get to the next tier in your loyalty program.
• Upsell: If your products have a natural progression, you could use the subscriber's purchase history to suggest the next item to buy. For example, if a subscriber just bought your printer, you could send an email with a link to buy toner.
• Weather: Imagine there's a big snowstorm in the Atlantic Northeast. You could send an email that advertises snowshoes to subscribers in that area but other kinds of shoes to the rest of the country.

This level of customization is possible through Marketing Cloud's proprietary scripting language, AMPscript, covered in Chapter 11. This list only scratches the surface of what you can do with AMPscript and the right data.

Defining an initial data set

The worst thing you can do to yourself is to try to build all these scenarios at once. We recommend starting small and building toward more complexity over time. You'll have the opportunity to think of new uses for the data you have and to ask subscribers for new data that you can use to craft a delightful user experience.

It's tempting to store data just because you can. Despite your best intentions, though, you aren't going to find a use for every piece of information you have about your customers, products, or email interactions. Be objective and narrow the data you store in Marketing Cloud to what you will really use.

You can also run into trouble trying to store data that isn't easy to get or keep up to date. For example, data that is available in real-time thanks to an integration between Marketing Cloud and another business system is better than data you have to load into Marketing Cloud manually every month.

We recommend that your initial data look like what you would expect from a basic direct mail database:

• Salutation (Mr., Mrs., Ms., and so on)
• First name
• Last name
• City
• State
• ZIP code
• Country (if you send internationally)
• Birthday

If you have more than one list type (newsletter, discounts, random promotions), you should also capture which list each subscriber is on.

The data in the preceding list is the base level of information you need to start sending effective email marketing that is both personalized and potentially customized to fit your customer’s needs and expectations.

Dealing with a data shortfall

What if you don't have the data to fill in all the fields for all your subscribers? Three options are available to you:

• Proceed with what you have.
• Change how you collect data.
• Purchase supplemental data from third-party sources.

Let’s examine each option.

Change how you collect data

People usually use the second option — changing how they collect data — with the first one. While leveraging the data you have available, you improve your process for collecting data.

Preference centers and surveys are tools you can use to ask your subscribers for data. As you build trust with your customers, customers will be willing to provide you with more information.

For example, you might send a periodic, dedicated email that asks subscribers to go to a survey tool and answer some questions. Another approach is a simple question at the bottom of every email you send asking whether the information was helpful. Figure 5-4 gives an example of such a question. Clicking yes or no shows the subscriber a “Thank You!” landing page while showing you what kind of content the subscriber likes to see from you.

Beware of becoming too extreme in your data collection efforts. Requiring all fields on a sign-up page, for example, will inhibit list growth substantially. Instead, think about all the processes you have in place that interact with your subscribers and make minor adjustments. From call center operations to physical point-of-sale installations to your email programs, you have numerous opportunities to enhance the quantity and quality of your data without scaring off your subscriber.

Consider quick actions like the following to gather data without offending your subscriber base:

• Send a “spring cleaning” email to encourage your subscribers to update their information in your preference center. This can be used with an incentive, as in Figure 5-5.
• In your newsletters, insert links that ask subscribers if they want to see more of this type of content. By tracking the clicks on the links, you can build a profile of the interests of each subscriber.
• Publish incentives for people to update their profile and interests. These incentives don't have to cost you money. Just a statement such as “Update your profile today to receive more of the information you like” reminds subscribers that you want to send them relevant content. Just make sure you have enough content to support a statement like this!

Data after a send

Don't forget that you'll receive a wealth of new data from each message you send. This continual influx of data includes useful facts like the following:

• Bounces: Emails that were returned to you undelivered. Subscribers who bounce might not be worth continuing to maintain information about.
• Open date and time: The date and time the subscriber opened the email. You can use this information to get an idea about how prompt certain subscribers are about opening their email. In aggregate, you can even use this data to make decision about what day of the week or time of day to send future emails.
• URLs clicked: The date and time the subscriber clicked a link in your email. You might be able to infer the success of particular kinds of links (images versus text), the subject matter of the link, or the position of the link in the page from which ones your subscribers click.
• Unsubscribes: Recipients who no longer want to receive your emails. Obviously, it's important to know which subscribers to continue targeting.

Combine this information with your other useful data, and you have a powerful base of data to make immediate, intelligent decisions about the people you market to and how you market to them. It also gives you a solid foundation for testing.

Marketing objectives

You decide the marketing objectives that make sense for you and your business. Just like professional development goals, your marketing objectives should be

• Specific so that everyone understands what needs to be accomplished
• Measurable so you know if your campaign met the objective
• Defined by time so you have a date that you plan to complete it by
• Pertinent so that the objective helps business performance
• Actionable so that there are clear steps to take to achieve the goal

Don't forget to be realistic. Too many objectives, timelines that are too short and goals that are too high can do more harm than good. Start slow and remember that you can always add more objectives later.

Data to fulfill your objectives

After you've decided on your objectives, you can start figuring out what data to gather to determine whether you've achieved those objectives.

Let’s look at a sample goal. You have a newsletter that reaches 100,000 people. On average, 25 percent of the subscribers open the email and 5 percent click a link (click through rate, or CTR). One goal you might have is to increase the CTR to 10 percent over the next 12 months.

A lot of variables could be at play in determining the CTR of your email newsletter, and you can use data to experiment with each of them. For example:

• How many emails you send per week. You could experiment with sending more or fewer emails to see the effect on CTR. See information about A/B testing in Chapter 11.
• Whether a subscriber opens the email. A subscriber has to open an email to click a link, and adding personalization to the subject line is a proven way to improve open rates. Find information about personalization in Chapter 11.
• How many subscribers on your list are inactive — never open or click at all. You could remove inactive subscribers to increase the density of clicks among the subscribers who remain.
• Number and placement of links in the newsletter. You could experiment with repeating a link at the beginning and end of a newsletter to raise the visibility of the link and improve CTR.
• Time of day you sent the newsletter. You can evaluate what time of day you get the most opens and clicks and focus on sending at that time.

Data objective exercise

Let’s wrap up this chapter by using data to solve a real problem:

1. Take out a piece of paper (or open your favorite note-taking app on your device of choice). Write down a problem you're facing and your objective to improve the situation.

   Is that objective specific, measurable, time-bound, pertinent to business performance, and actionable? If not, take the time now to mold it so it meets all those criteria.

2. Brainstorm the factors that might be contributing to the problem.

   Don't limit yourself to the obvious factors. You can even include the factors that influence the factors!

3. Think about what data you could use to investigate each of the factors.

   Is the data readily available? How can you access it?

Getting used to using this approach to problem-solving makes the data in your Marketing Cloud account immensely useful to you and your company.