Chapter 07: Applying Website Analytics to Your Digital Marketing

What is your emotional reaction when you hear the word server? 

Stress? Fear? Frustration? 

Fortunately, the data and the analysis don't need to be scary. In fact, it can be a lot of fun (or, at the very least interesting, if you know how to turn off all these numbers and reports, to any intel that you can use to grow your business. 

In this chapter, you will learn the techniques to make the analysis, the data of your company, the stats that matter the most, the language you use to talk about it, and the groups or roles that should be in charge of it. 

But first, let's talk about the why of the data and the analyses are so important to the success of the company. 

Why Data Matters 

The Data is available in two versions: too little and too much. 

The challenge is that most people are grappling with how to turn the number into a meaningful decision. The static figures, in and of themselves, are meaningless. 

So, why would you want to do analytics? 

In order to understand the answer, let's look at some examples: 

Oakland A's athletics club 

Billy Beane took over as the General Manager of the Oakland A’s in 1997.

There, he applied statistical analysis (known as sabermetrics) to players, radically changing the way they acquired players.

Beane’s approach was to focus on specific metrics, such as batting and runs, to find undervalued players no one else was noticing. This approached made the Oakland A’s one of the most cost-effective teams in baseball and carried them through 20 consecutive wins, playoffs, and even the world series.

Essentially, data made the A’s competitive with much bigger clubs while working on a budget a third of the size.

Netflix

Netflix's core belief is that the adjustment will win customer loyalty, faith, place data at the center of their corporate strategy. 

When they were still in a DVD rental company, Netflix has invested heavily in data-mining techniques in order to develop a movie recommendation algorithm, the global leader in the use of the data in order to provide a good customer experience. And it worked. 

Recommendations and drove in 50% of their traffic. 

Following the adoption of a streaming model, and the first method is continued, and which has made them one of the top streaming, video-on-demand services that are available. 

At present, it is to provide them with the knowledge of the customers ' needs as they need to in order to create hugely successful original content like Daredevil, House of Cards, Orange is the New Black.

None of this would have been possible without data.

Dream Digital

We're no strangers to the data, either. I'm going to get into more detail later on in this chapter, but here, in a DM, we will have to rely on this data to help us in-to make the business decisions that are all but guaranteed to work. 

Is our faith? Gut instincts can be a good thing, but the data never lies. 

The challenge is, of course, of your practice. How do you go from a spreadsheet to a strategic decision to grow your business? Let's take a look at it. 

The Methods of a Well-Executed Analytics & Data Strategy

Master the analysis and the data, you'll need to master the 3 guiding principles: 

• Assign tasks to a job. This is the basis of the analysis of the data. Each and every piece of data you collect will help you answer the questions, and make smart decisions. 

• Use of hypothesis testing, in order to convert the queries into strategies. That's what makes the data meaningful. It is the process of transforming raw data into decision-making. 

• Apply to the context, to explain the unfathomable. Some things are hard to measure. For this situation, you will need to in order to contextualize the data. 

Analytics and data shouldn’t be stressful. But it’s easy to feel that way when there are so many sources to draw from, each formatting the data differently, sometimes even giving different numbers for the same metric.

Where do you put your attention? How to compare the data from different sources?

To start, give your data a job.

Principle #1: Give your Data a Job

One of the easiest ways to understand data is to think of the marketing funnel.

This is one of the fundamental concepts that make it easy to visualize your customer acquisition marketing. Your brand is marketing to create awareness and attract new visitors to your website. Some of those new leads will be interested enough in order to evaluate your business and the products and some of them will go on to become customers. 

For sale, this is a good model, but we can adjust the ductwork just as well for the analysis of data. 

In this model, a funnel, statistics, process flow chart, which is not only the maps of the stages in the customer journey, and also lists the metrics that will be measured at each step.

With this approach, we use the same 3 phases of the funnel but rename them like this:

1. TOFU, or top of the funnel, is the awareness phase

2. MOFU, or middle of the funnel, is the evaluation phase

3. BOFU, or bottom of the funnel, is the conversion phase

But we don't stop there. We also need to know what happens after someone becomes a customer. 

So we're adding one more to the stage after conversion, the phase of the process, focusing on how they can be transformed into a regular customer, the lifetime of the users, and as an advocate for your business. 

This is how you'll be able to give your data to a job. You do not need to look at all of your data all at once. 

You need to choose the various values for each stage of the funnel. 

Rather than measuring your company's health just by its bottom-line numbers are, and you measure the health of each of the phase-detection of leaks in your funnel, to find strategic ways to wire them up and make it easier to convert. 

To begin, let's identify the instant of the values you need for each step of the customer journey.

Categorizing Data by The Stage of the Funnel

TOFU (Top of Funnel)

Your goal for this stage? New visitors.

The key question when choosing metrics for this stage is this: Does this metric give me insight into brand-new visitors?

A good example of a TOFU metric: Direct new visitors.

Direct visitors, the number of people who enter your site's URL directly into the Program! It is capable of measuring the efficacy of online and offline advertising 

If you have been driving awareness ads, you should be seeing them from the people who are trying to learn more about you. 

If you have the signs of which contain a URL, you should see a spike in the geographical areas of us billboard when they visit your website. 

MOFU (Middle of Funnel)

Your goal at this stage? Converting new visitors into leads.

Your guiding question when deciding whether a metric is right for the middle of the funnel is this: Does this metric give me insight into how well I’m getting visitors to commit?

“Commitment” can be defined as:

• People subscribing

• People filling out a webform

• People following you in social media

It’s about people giving you permission to reach out to them and offer more value.

A good example of a MOFU metric: CTA clicks.

If you have a blog on a banner to learn more about one of our products. You'll need to know how many clicks that banner, which is the proportion of your blog visitors to click on, so that you can evaluate how well your content is to convert your visitors into leads. 

BOFU (Bottom of Funnel)

Your goal? Converting prospects into customers.

Your guiding question is when you select the dimensions of this step is to get this measurement gives me insight into how well prospects to convert into customers? 

This point is especially important because when someone buys something from you, something small and cheap, the likelihood that they will have to buy it again increases to 10 times, and their desire to invest in a relationship increase significantly. 

A good example of the BOFU metric conversion percentage. 

How many people clicked or purchased from a brand communication? This tells you which offers are working and what kind of offers you should make to new customers.

Retention & Monetization (After Conversion)

Your goal for this stage? Customer satisfaction. You want to increase membership, traffic ROI, retention, and customer lifetime value.

The guiding question is, when you look at the post-transformation: this measurement gives me the insight into the level of satisfaction of our customers is it? 

You are looking for the statistic, which describes the real-world result from the use of the product. 

Something like this:

This information is not to look for it on Google, so it is difficult to come by but satisfied customers are often willing to share some of them. Here, in a DM, we will have to look for positive reviews are from the people in our membership area. That will tell us how well we are helping people to achieve their goals. 

Keep in mind that these are vanity metrics. They help us to know what it takes to keep the people of the DM in the Lab and encourage them to tell their friends about the Lab.

Example of a post-conversion: the Members On/Off report

This report tracks how many people we’re adding to a subscription product, how many people we’re losing, and how we’re losing them.

These metrics tell us how healthy the product is, and when they’re combined with other reports, we can see the things we do that drive cancellation or boost membership. This information is critical for a membership product because member retention drives profits.

Categorizing Data by Type

We’ve just reviewed TOFU, MOFU, and BOFU metrics, which is a way of categorizing metrics by the stages of your funnel. But there are other ways to categorize metrics, and that’s by the type of information they provide.

There are 2 types of metrics:

Key performance indicators to dictate the public's health. These metrics are like a thermostat for your needs. In order for a metric to be important, as you need to in order to be able to look at it and know immediately whether your company is a good thing or not. 

The Drill-down metrics answer the question. These measures are more detailed to help you understand what is going on in specific areas of your business. 

Typically, you use both types of statistics are together, not just one or the other. Regarding the methods of measurement in order to tell us things are going well, you have to use the drill-down metrics to help you understand the why so that you can replicate your success.

For example:

Improving On-site Banner Clickthrough

On Average, the Banner of the Click is a key variable. It will tell you on the whole site, how likely it is for a visitor to click on an ad in one of your articles. 

In order to help us understand why the click rate is 3.25%, and how we can improve this issue, we need to have a drill-down measure of How likely it is that a visitor will click on an ad for a particular item? 

It was not so long ago we made it to the DM's blog. After a review of all of our posts in the blog and drill-down to specific metrics, we were able to identify the factors that impact the rate. For our results, we have improved the banner ad click-through rate by about 2 percentage points that of, the blog. 

Improving Search Traffic

In this example, it is the most important variable is for New Visitors: the number of first-time visitors who land on your website? 

To find out more, the drill-down of measurement would be the Percentage of the Find: how much of the traffic is owned by the person, or a brand name? 

In here, would you want to compare how much traffic you are getting for a keyword that no matter how much your competitors are getting from with the same keyword? This would make it possible for you to find out the areas in which you can compete with much larger companies, because of your keyword search. Alternatively, it is going to tell you exactly what keywords you need to add a little bit of work to it. 

Improving Ecommerce

In this example, the key variable is the Average Of the Values. To drill down, you would need to look at the actual order. 

Do you want to know the location of the main portion of the proceeds from the Average value of the order is coming? This will reveal that you need to move things around in your sales funnels, or you will need to run a campaign on the basis of the average value of the order is much higher. 

As you can see, there are 2 ways that you can provide your measurements for a job. 

• You can assign them to a specific stage of the funnel: TOFU, MOFU OR BOFU. 

• You can use it to measure the health of the various areas of your business, and then answer the deeper questions of whether, how, and why. 

Once you understand the overall health of your business and where things are working (or not), you can start to use the statistical data in order to solve the problem.

Principle #2: Using Metrics to Solve Problems

Data is collected on a dashboard, right? But on the dashboard, it’s raw data. 

Your job, as a data analyst, is to turn raw data into actionable data. For that, you use the analytic decision-making process.

This process works a lot like the scientific method, except that it is based on statistical data. 

In the scientific method, you have to begin with the questions and hypotheses, and then you can make predictions about what will happen if you are testing a variety of hypotheses. 

It's the same thing with data and analytics. You can view your data, and begin to ask questions about it. You are making assumptions about what might happen to you if you would be able to influence some of these people. And then, you can display a test to see if you are allowed. 

Quite simply, by examining the results, you are able to clearly see what needs to be done in order to improve your business. Decision-making is no longer on your gut instinct, it's about what the data is telling you. 

That's the theory, anyway. 

But the reality is that we often don't know enough to know what questions we should be asking ourselves. In these situations, it often helps to take the data from the dives.

Reviewing Key Metrics to Inspire Questions

When you don't know enough to know what you should be using, your data will often provide you with the insight you'll need. 

Step 1: Start by reviewing Your metrics. To identify the places where your results are better than expected, or it might be a downward trend. In many cases, this is going to inspire more questions. 

•  This blog post traffic is the dual of the other posts in the blog. What did it do better? 

• We have a new subscriber every day, but the subscription, there is all the same. What's going on here? That's where we are losing subscribers? Why is that? 

• Each time the author of the CHAMPIONSHIP writes in a blog post, traffic, and the number of shares is higher than normal. What makes her blog posts are better than someone else's? 

As an example, when you review the Relevant On/Off report: report of the DigitalMarketer Lab, we have seen a few strange inconsistencies in the Members which are Added, the column 

In a few more weeks, we could put 131, 112, or 106 of the new member states. In the second weeks, and we might just add it to 11, and 21. After seeing these figures, some of the questions that came to my mind: 

What is the cause of these spikes in the subscriber's conversions? 

• What are the causes of these periods, the lower conversion? 

Step 2. To generate a Hypothesis as to What's going on. Make a few predictions about what's going on. For the user being Added to inconsistency, we made three conjectures: 

• $1-this This is a better deal, as it will convert more website visitors into all of the sources. 

• It is a $1 trial, customer churn is higher than that of the full-pay of the members, which means that it does not contribute to the, our goal is to grow in the lab. 

• Start-up for $1 trial offer for their time runs out, the blow-up conversion rates. 

Hint: don't be satisfied with just a hypothesis. It is best to consider a number of factors—ideally 5 to 7, hypotheses, and test them all. Otherwise, you may limit your ability to learn what's going on. 

As you can see, in most cases, it is not the only cause of the problem, as you can see. A number of factors can contribute to the success or failure of the, you want to understand. The more assumptions you have, the better your chances of identifying all of the factors that are involved in it. 

Step 3. Use the Drill-Down Methods, in order to Test the Hypotheses. For this, you will have to use more detailed, in-depth data in order to figure out what is causing the problem that you are trying to understand. 

That's not usually over on a day-to-day basis, but there are, and you'll know where to find it. There are also data to help you answer these types of questions. 

For this case, we have used a cohort analysis in order to test the 3 hypotheses. 

We developed the 7 cohorts, or the number of ways you are able to group the people in the DM Lab, including the cancellation date, the average of the percentage to be paid, the length of time they were active, and much more. 

After reviewing all of this information, we know that the $1 trial, be a better front-end of the offer other than the full-pay offer. 

Step-4. Take Action Based on the Results. The overall conclusion of the analysis of the data was that the trial offer will generate more paying customers than the national average. For every 100 people that we put up in the DM Lab, 21, and came to the trial, while 20 were from the full-pay offer. 

At the same time, there is not a significant change over a period of time, it can add up. So, we have to change our front-end offer. The $1 trial offer today is our first, and so far, it has been given an additional 1,000 of the Lab members.

Principle #3: Contextualizing Data to Account for the Unmeasurable

No matter how good your data is, it sometimes doesn't tell you everything you need to know.

For example, suppose you review your data and see trends. Why is this trend beginning to take shape? Maybe you did a campaign at that time.

Maybe your competitors have done something different. Or maybe you have a technical problem that has disrupted the data.

If you ignore these things when checking your data, you may be making assumptions based on a set of false data. Your conclusion will not work.

In these cases, the context helps you to address the diversity of your data. And there are 4 situations you need to consider.

Historical Contexts

What is the meaning of history is to tell you to expect? By examining the data through a historical lens, you can understand the trends and the typical behavior of your customers. 

For example, at DigitalMarketer, we have seen sales dip in the summer. To date. Each year in the summer. 

So, instead of worrying about the lower numbers, we have come up with strategies in order to increase the sales at the end of the spring. We have also reduced the ad spend in the summer because we all know that the ROI is not going to be just fine. 

External Contexts

As for the changes, which are beyond our control, have had an impact on our metrics? Maybe a new competitor has entered the market. Or maybe the technology has changed, which will require major changes in the way that you do. 

Think of Google's algorithm updates. The external factors may be outside of your control, but you'll need to keep them in mind in the evaluation of the results.

Internal Contexts

If you have made changes to your strategy in order to improve the performance? Have you made any changes to the website or the launch of a campaign? 

This is a private room. Consider some of the changes that you have made internally, which may have influenced the number. 

Contextual Contexts

This has nothing to do with the way you are pulling the data. Are you comparing the raw numbers or percentages? Is your data are affected by extreme values? Do you have data that doesn't make sense as an internal or external factor? 

Together, these contextual factors help to explain the order of the star of the things, the things you are not able to predict or explain your data. And they will help you possibly to evaluate the validity of the data.

Making Data Actionable

As you can see, the 3 principles to carry out the analysis, and the data can help you to turn on or off the random numbers to any data of your company. 

You'll need to assign roles to your data, so that you will know the stage of the funnel they are related, and whether they will help you to know something (which is the key of birds and fish), or to provide you with the information in order to respond to a question (the drill-down data). 

You will also need to use data to make smarter decisions for your business. Use it to test your ideas about what works and what doesn't, and how you can improve your outcomes. When you examine the numbers, in order to answer a question, do you know what it is you are trying to prove or disprove? 

Finally, you need to put your data in context by examining the factors that drive the numbers up or down. By linking the data to the real world, data becomes more meaningful, and it is going to be easier for you to use in your business in order to drive business growth.

The Lingo Analysts Use

There are 5 terms you must understand to talk intelligently about analytics and data.

Analytical Decision Making

This refers to the data scientist’s scientific method. It’s the process you’ll use to identify the questions you should be asking and the best methods for answering them.

Analyst’s Toolkit

These are the tools, templates, and resources you’ll use to turn concepts and ideas into data and reports. Your toolkit will help you ask the right questions and develop a process that makes data analysis easier.

UTM Parameter

This refers to the code you can append to a URL to give you more information about where your traffic is coming from.

In this example, yellow indicates the actual link. Everything after that is a UTM parameter. This is additional information that will help you keep track of your traffic sources. 

• Green highlights the source, who says to you, the audience, or referring the website (house list). 

• The pink highlights on the medium, which is to talk about how this room was referred to in the e-mail. 

• The blue highlight to the content, which is your ad, the content identifier, and data CAC and the launch of an e-mail to 1). 

• Gray is the highlight of your campaign, which is to identify the promotion and/or strategy of the traffic came from (data-cc 1-1-2016). 

When you add UTM parameters to your links and the like, click on those links will have to be ready, and you're able to track those tags are in the Google Analytics service. 

Take a look at the sources of, and communication is to provide you with the best traffic.

Key Performance Indicator (KPI)

KPI is another way to refer to a metric in general, and it’s usually used to talk about a metric that someone thinks is driving their business. KPI is another way to talk about a key metric.

Dashboard

A dashboard is a web page that collates your metrics from a particular source.

You’ll likely have a dashboard for each data source: Google Analytics, your email service provider, your social media platforms, and more.

Most dashboards also provide graphs that turn your data into visuals, making it easy to see how well you’re performing.

Dashboards should be available to everyone on a team. They help you easily understand what’s going on in a business, which helps team members see how their work is impacting the success of your business.

The Roles: Who Needs to Be in the Know?

Who should own your business’s data? Where in the company does data analysis live?

Data & Analytics

The analytics team (or individual) should have primary responsibility for gathering, vetting, and interpreting your data and analytics.

Larger teams may also have a data implementation manager, who aggregates all this information and turns it into a beautiful dashboard that’s easy to understand.

Marketing

Every marketer worth their salt needs to know a little about analytics and data.

Whether you’re running Facebook campaigns, tweeting 50 times a day, or posting articles to your blog, you need to know what’s working and what’s not.

Conversion Rate Optimization (CRO)

The people who run tests to optimize your marketing rely heavily on data to develop their hypotheses, set up tests, and measure performance.

Summing Up

The Analytics don't have to be scary or overwhelming—even if you're not a number. You need a process to deal with the numbers, this is a way to figure out which ones will help you to identify the opportunities, and what you wish to do so. 

Once you have your process in place, you may actually find that you enjoy it. There is no better feeling than knowing without a doubt that your marketing plan is working, and what are the metrics will give you confidence. 

We are nearing the end of our digital marketing dashboard. The next (and final) lesson is, conversion rate optimization, which is a simple process that will improve your marketing results over a period of time. 

Honestly, conversion rate optimization, or CRO, as it is often called, is the secret sauce of your digital marketing strategy, and you're going to love how it focuses your energy on the tasks that matter the most.