Where Do the Data Come From?

To navigate this complex world of overwhelming amounts of data, you need to know where they come from and the nature of the information. To put it in simple terms, the data start with people using your website or app. Their actions are captured, processed, and reported by a machine that sees everything in 1s and 0s. Literally, a computer is attempting to capture, understand, and communicate human behavior on a digital interface (website/app). Just think about how ridiculous that sounds. Yet, that's not the end of it! We now, as humans, are attempting to decode those charts, graphs, and reports that are given to us by these machines. See Figure 2.1.

Your job is to decode the data into meaningful information that explains what the humans (the visitors) were attempting to do on your websites and how that translates into business objectives! If that sounds overwhelming, it is! You are attempting to understand (on a machine) what data machines gather about human behavior (while they are on a machine), and there is nothing simple about that!

Gathering the Data

The first step is to capture and collect visitor activity on your website. In the past, the server where a website was hosted would maintain a log of every page request. Of course, this was called the log file. Every request would be collected and timestamped. A page request, while sounding simple, could be highly complex. An HTML page is simply a file that contains a set of instructions and text. These instructions also provide the links to find the other necessary elements to assemble the page. Images, scripts, forms, and other elements are kept in separate files but assembled in the browser when someone requests the page. The HTML acts as the base canvas with the instructions of where to find all the necessary elements, how to assemble them, and how the page will function. Each request to the server is called a “hit.” So, if a page is requested (one request) that has 10 images, 7 scripts, and 2 forms, it generates an additional 19 requests (or hits) to the server. Each of those requests is logged and timestamped in sequential order.

As you could imagine, once you start getting thousands of visits a day, the file would grow substantially. As the Internet became busier and more people going online, maintaining these log files on servers required more space than hosts were willing to provide. The response was to create a hosted version of the log file and provide analytics from the data. Now, the hosted service needed a method to gather the data because it wasn't connected to a website's hosting server. So began JavaScript tagging.

Just a warning that this next paragraph is going to get technical. There are three methods of implementing JavaScript tracking:

• The first is through an inline HTML tag. It is used to track a single click or a download, and it is not recommended. It increases download time for the page, which can negatively impact your rankings.

  <a OnClick="ga('send', 'social', 'Linkedin', 'Share',window.location.href);">…</a>
  
• Next is embedded JavaScript, which runs on the visitor's browser. It is identifiable by the <script> tag (the <script> tag does not include src):

  <script>
  Window.dataLayer = window.datalayer ||[];
  Function gtag(){dataLayer.push(arguments);}
  Gtag("js",new Date());
  Gtag("config", "UA-123456");
  </script>
  
• Finally, External JavaScript uses the <script> tag and includes src, which enables it to run in the background, apart from other scripting commands. This is used by Google Tag Manager to collect visitor data:

  <script async src=https://www.googletagmanager.com/gtag/js?id=UA-123456-xx></script
  

What Data Does Analytics Script Collect?

That's the billion‐dollar question. Google realized that the data they collect was their biggest asset because they use it to match advertisers to visitors based on the visitor's browsing history. Google's annual revenue from selling advertising will be more than $300 billion by the time this book is published.

For starters, when a page is requested, there is an exchange of information called the HTTP protocol. (Hey, recognize that!?!) Before the page is loaded into the browser, your computer engages in a very formal exchange of introductions. This is first initiated by the HTTP header, which transmits this data:

• Cookies (Google): For Google Analytics to track the user between sites
• IP address
• Device type
• Operating system and version
• Browser and browser version
• Referrer: URL of the current page with the link
• Language: Language and region of the requested website
• Date and time: Every request is timestamped
• Host: Domain name of the requested file

Other data are also collected:

• Screen resolution
• Screen size
• Screen colors
• Client size
• Viewport size
• Performance data
• Plugins
• Top sites (bookmarks)

Other data are also utilized in Google Analytics. Some may be captured in analytics, some by other methods, as Google also owns the Chrome browser, which also transmits user data to Google. You know that these tags collect a massive amount of data, and Google does not offer a list of the full amount or types of data collected by its analytics script.

The development of cloud computing provided a perfect method for capturing and analyzing all this data. To be usable, these data have to be collected and processed, and it takes tremendous computing power to process millions of lines of activity to organize, synthesize, and report the data. Analytics providers employ data sampling to balance the computing time with the need for immediate access to standard reports. For large sites, it would simply take too long to crunch 100 percent of the data and provide it in an on‐demand format. Sampling the data extrapolates the trends from the sample to provide the standard reports. Analytics owners can change the settings to choose a higher amount of data to be sampled. Obviously, if you see that 10 percent or less of your data are being used to create the reports, you know that there will be some major inaccuracies. For example, changing the preferences to greater precision rather than a faster response will create a larger dataset for sampling, but also take a bit more time in creating the report.

How Accurate Are the Data?

Let's just get this out there. There is no such thing as 100 percent accuracy in analytics.

While the data collection methods look impressive and capture a lot of data, there is still a lot of processing and interpretation that must be done with the raw data. Remember, you're dealing with humans and human behavior. How many times have you switched devices when researching a vendor or shopping? Let's say you start shopping while at work, maybe looking for a gift for someone. You cover a lot of ground by searching on Google for gift ideas and browse four or five websites. Later, at lunch, you think about one of the ideas, and you take out your phone and type in one of the websites that you remember. Of course, the ad for that brand is the first result, so rather than scrolling down, you click the ad, go to the site, find the product and spend time reading more about it, maybe even looking at a few of the videos about the product. Now, later that night, you are at home, and you decide to purchase. You go to your home computer, because you might feel more comfortable with the checkout procedure on the big screen, as most people of a certain age do. You type in the website URL, go directly to the website, then to the product, and you've completed the checkout in only a few minutes.

To you, this seems normal and nothing out of the ordinary. That's because you're thinking of this in human terms. In machine terms, here is a typical interpretation of your behavior in analytics:

• New visitor: Source: Google (Organic search), IP address xxx.xx.xx.xxx, device:desktop, pages viewed=12, time on site=6:25
• New visitor: Source Google (Paid Search), IP address: yyy.yy.yy.yyy, device:mobile/iPhone, pages viewed=3, time on site=3:20
• New visitor: Source: Direct, IP address: zzz.zz.zz.zzz, device:desktop, pages viewed=3, time on site=2:45, GOAL: Checkout purchase

In machine terms, you appear to be three separate visitors. This is because analytics is tied to the device. By tracking the device type, IP address (location), browser types, and identifying information this separates activities when using multiple devices. Now this can be overcome by personalizing each device for tracking. This is exactly what Google attempts to do when you login to your Google account on different devices. Google (and Facebook) attempts to unify your devices to know what you use, when you use it, and combine that data into a unified capture of the person rather than the devices. So, how do you share everything that you do to provide a complete record of your online activities to Google? Then use Google's Chrome browser on your devices, while logged into your Google account, and using websites with Google Analytics installed. Google will record everything that you do and track as much data as they can from your browsing behavior! If you aren't comfortable with this, you can mitigate the data collection by using different browsers while logged out of your accounts.

Analytics programs run on algorithms, just like search engines. The algorithms are simply rules applied to the interpretation of the data. Primarily, these rules define how to interpret certain events or data and classify it accordingly. As an example of this, have you ever left your computer overnight while your browser and a dozen or so tabs were open? Well, the analytics algorithm must figure out when exactly you left the website. This is the timeout. Some timeouts are after 20 minutes of inactivity (inactivity is defined as the time elapsed since the last request made to the server). How do you come up with a time‐on‐site calculation when there is no definitive exit action? All time‐on‐site calculations are based on estimates of behavior and the algorithm's rules to calculate this metric.

But what about the next morning? You awaken your computer, click on the browser, and the tabs reload all the web pages? According to the analytics algorithm, you've started another session! Even though you might only reload the tab and then not get to that website for another hour, you've since timed out again. Once you begin interacting with the website in that tab, you'll start another session, which will count the pages you request and the time on site. If you finally convert by purchasing something or filling out a form, the analytics will suggest that you did this on the third or fourth visit, which were generated by a direct visit to the website. But in your mind, it's been the same visit, with a few hours or days in between each interaction that was originally generated from a search query.

This is the issue with accuracy. It's machines and algorithms attempting to classify and report human behavior on websites, which is completely unpredictable. Every analytics program has its own method of addressing these behavioral question marks, which is why you'll never find agreement among different analytics providers. They all have different algorithms that interpret the behavior according to their rules.

Accuracy in Attribution

Here's the final way that accuracy is up for interpretation. Let's take the earlier example of multiple visits to a website. But let's also assume that the analytics has tracked these visits as being from the same visitor. It would look like this:

• Visit 1: Source: Google (Organic search), IP address xxx.xx.xx.xxx, device: PC, pages viewed =12, time on site=6:25
• Visit 2: Source: Google (Paid Search), IP address: yyy.yy.yy.yyy, device: iPhone, pages viewed=3, time on site=3:20
• Visit 3: Source: Direct, IP address: zzz.zz.z.zzz, Device: PC, Pages viewed=3, time on site=2:45, GOAL: Checkout purchase

The default setting in just about every analytics program is called last‐click attribution. This means that the source of the visitor who met the goal in that session gets 100 percent of the credit for the conversion or the goal completion. Because of this, Google Analytics 4 changed drastically to report first touch, or first user medium, which favors search and paid search channels.

In this example, the goal completion was the last session, which the visitor initiated by going to the website directly. This means that a direct visit gets 100 percent of the credit for the sale or conversion. Do you think that is right?

The visitor first found the website through organic search, then performed a brand name search and clicked on the ad in the second visit, and then decided to purchase in the last visit. So, which source should have the attribution? Should the first since it was responsible for getting the visitor there in the first place? Should the second, because it showed that the visitor remembered the brand name? Or are you correct in giving the final source the full credit?

The answer will depend on your question. If you are evaluating the effectiveness of your marketing in driving customers, then you will want to focus on those first‐touch sources and measure their effectiveness in ultimately driving conversions. In this case, you would change to a first‐click attribution model. This would change the data completely to emphasize the sources and channels that are responsible for the first interaction with those visitors that would later become customers.

However, if you are evaluating multi‐channel effectiveness in attracting, nurturing, and converting, then you would want to see all the channels that influence the outcome—from beginning to end. You would accomplish this by using a linear attribution model that gives each source equal credit for the conversion. This would change the data again, enabling you to see which channels contribute the most toward engaging visitors and nurturing them to become customers or leads. Without an understanding of the attribution models and how they affect the data, most beginners to analytics are making decisions based on last‐click data, which neglects the first‐touch channels that are instrumental in gaining new visits or influencing their decisions.

Additional Data Sources and Tracking

When marketing through third‐party platforms, such as Facebook, Instagram, LinkedIn, or even email, you may not receive all of the data you need. There are additional methods of capturing tracking, such as UTM code, which appends to a link and provides clearer data about the visit. In addition, your analytics data may need to be supplemented with additional platform data, such as Google Ads, Facebook, and others.

https://mydomain.com/landingpage?utm_campaign=freetrialoffer&utm_medium=social&utm_source=linkedin

The ABCs of Analytics

Now let's get into the basics of understanding data. To build your foundation of understanding any analytics program or report, just follow the ABCs.

This is analytics and the purpose of analytics at its core. You are answering questions of WHAT happened, WHY it happened, and HOW to make decisions from what happened. You can only get to the data by breaking it into the essential elements of what you want to know.

A: Acquisition  Acquisition tracks where people come from. Here are the primary channels visitors use:

• Search (organic and paid). These are people who made a search and clicked on your website in the results. They either clicked on the organic results or on a paid ad if you are running a campaign.
• Direct. This means that the visitor typed in your website address in their browser or has it bookmarked in their favorites. With autocomplete in the browser, it makes this method much easier for people to remember the websites they visited.
• Referral. These visitors clicked on a link to your website while on another website. If a blog or a news site mentioned you and had a link alongside your business name, people who clicked that link are called referrals. Think of it like a word‐of‐mouth referral. Similarly, someone recommended your business by linking to it.
• Social. This category covers nearly every social media platform that is known in analytics.
• Email. Any email campaigns you send with analytics tracking will show the subsequent websites visits generated by the email.
• Other. Sometimes, referrer data aren't passed to analytics. If the referrer can't be determined, it gets dumped into the catch‐all of Other.
• UTM campaigns. If you set up UTM links, they will show up in your acquisition reports alongside these categories.

The measurements associated with acquisition are shown in the first columns: channel type and number of visits.

B: Behavior  Behavior measures what people do on the website. However, I find it much more interesting to base any judgment of behavior on what people saw and how it affected their behavior. Evaluating behavior without knowing what people were looking for or what they saw is only half of the equation, because this helps us find intent—the reason for the visit in the first place!

Figures 2.2, 2.3 and 2.4 are examples of different analytics programs and their reporting of behavior. The terminology of each is slightly different, but centers on the same idea:

• Actions/Events. These are actions that the visitor performs, such as page views, link clicks, downloads, goals, and out clicks.
• Bounce Rate. A bounce is a single page visit. It means that there was only one page viewed and the visitor left the site. (In strict analytics terms, it means that they did not request another page or file and timed out.) The bounce rate tells you how many visitors out of 100 only had a single page visit. So, if the bounce rate is 33 percent, then (on average) 33 out of 100 visitors only saw a single page.
• Pages per Session. This is another average based on the number of pages that have been requested by the visitor during an active session. Pages per session is being phased out of newer analytics applications and being replaced with actions or events, as many interactions can happen in a single page.
• Average Session Time/Average Engagement Time/Average Time on Website. This is an interesting metric. As mentioned, it is based on the timestamped requests made from your browser. When you request a new page, it is timestamped, and so on for each additional page. When you stop making requests, the session times out, and the calculation is made to come up with length of the session. In the Average Engagement Time column in Figure 2.3, the averages are shown based on the source of the visit. You can compare the behavior of visitors based on their source. When comparing many different types of websites, there are some generalizations that can be made, such as visitors from direct and email tend to have longer sessions. Visitors from social and display ads tend to have the shortest sessions. Of course, this isn't a rule; it's a generalization that is based on the behaviors of visitors from each source. People tend to behave differently based on their source, and it is something to find and evaluate for your business.

Actions, or events, are replacing the page view metric, and analytics programs are adapting to allow businesses to create their own reports and events based on their unique website, app, or purpose. In this example in Figure 2.5, Piwik Pro provides a list of the events on the right that can be added to a custom report. This way, site owners or managers can adapt the event to their specific purpose.

C: Conversion  Conversions are the goals that have been set for visitors to reach. I like to define conversion as getting the visitor to do what you want them to do! Conversions encompass many actions, and they are based on what you define for your business. You'll find that the terms conversions and goals get used interchangeably. Google Analytics attempts to make a differentiation by labeling conversion as the main category, which is made up of ecommerce and goals. In Google Analytics, ecommerce conversions are tracked when integrating your ecommerce website and the Google Analytics ecommerce tracking code. All other conversions in Google Analytics are considered goals. Conversion reporting provides the goal‐based conversions such as Total Conversions, Conversion Rates, Conversion Value, and Revenue Reporting.

Establish Goal Values  A goal here can be an ecommerce checkout. Google Analytics won't track a checkout action as a goal unless it is set up manually. In fact, all goals are set up manually in the Admin section. In some cases, you may find that your ecommerce software or customer relationship management (CRM) software are tracking the same information. Both types of systems are including more analytics into their tracking, which might reduce dependence on Google Analytics in the coming years.

Goal completions are the manually created goals for your business website. These are based on important tasks that you decide are worth tracking, as they show a contribution toward business value. These include email subscribers, form completions, video views, contact forms, downloads, etc. Anything that shows or can be tracked to business value can and should be tracked as a goal.

What's the Impact of the Goal Completion Value?  Finally, goal value. This is one of the more difficult functions to define, and it is typically neglected by most companies—at their own loss. Here is why this column is so valuable.

As an example, how much is a new email subscriber worth to your organization? There are a few methods that you could use to determine this.

First, what would it cost you to purchase a list of emails? (Not that I would ever recommend this.) Of course, any purchased list would be trouble. Those names have not opted‐in to receive emails from you, and they most likely do not know your brand or business. But buying a low‐quality list like this might run you about $0.10 per email.

Second, what is the average value of a subscriber? This is easier to figure out because it is information that should be measured and calculated within your organization. When sending your emails, do you track the revenue produced? If so, then you should have an average revenue per subscriber calculation, and it should be well known!

Third, start somewhere! For many in the industry, they recommend a value for a new subscriber somewhere around $20 or more.

What this goal value reflects is the economic value to an organization when a visitor completes a task‐based action. It does not mean realized revenue, as an ecommerce report would; rather, it is the estimated value produced. B2B organizations track leads, and they usually have the estimated value of a lead in their calculations. When a visitor completes a lead form on the website, they can attribute a value to that lead. To use this example, if you have an estimated value of $10 of revenue per lead, it will accumulate a running total in the Goal Value column.

Referring to Figure 2.6, the source of Organic Search has generated 132 goal completions. This is the sum of all goals defined by the organization, each of which may have a different estimated value. For example, Organic Search generated $6,744.00 of Goal Value from 132 goal completions, while Direct generated $33,140.00 from 478 goal completions.

Viewing goal value or the revenue created enables a complete view. Rather than basing judgments on the number of visitors, the conversion rate, or numbers of conversions, value provides feedback about the impact of the conversion because not all goals may have the same value. For example, gaining a subscriber may have a value of $20, whereas a visitor who completes a lead form may have a value of $150 per lead. While these are reported as two goal conversions, they have very different values to the organization.

Segmentation Is the Key

Using the ABCs gently brings you into the application of segmentation. Why? Because your visitors are not a herd of cattle. They don't all enter at the same page, they don't all visit the same pages, and they don't all want the same things. Looking at the big numbers, such as the number of visits or the number of sessions views your visitors as a large herd of cattle because it simply lumps those distinguishing factors into one number. What you need is segmentation.

Segmentation is the separation of your visitors based on common factors. For example, where did they come from? Google? Bing? Email? Facebook? Another website? Did they type the site into their browser? Simply breaking down how people came to your website is one of the first levels of segmentation. By breaking apart the sources of visits, you can start to develop an understanding of how people find the website.

Another method of segmentation is evaluating the landing page—or entry page of the visitor. This is where you really learn that people are looking for very different information, as they land on the page that they hope answers their question. You can also segment by new visitors or returning visitors, customers or noncustomers. Basically, any distinguishing factor can be used to segment people so that you can examine those factors to find insights into that group, your marketing, or your processes. See Figure 2.7.

This report automatically segments visitors by how they first found the website. Rather than simply looking at reports of aggregate data, you can find more value by separating visitors. You can explore this even further by adding more dimensions.

A dimension is simply adding another segmentation factor. How would you narrow down the segment of organic search? By the source of that search—the search engine that brought those visitors. By adding Source, you will see the segmentation progress to a secondary level of detailed information. The Primary dimension is Channel, and the Secondary Dimension is Source. You can now see which search engines generated visitors to the website and which contributed to conversions. You can measurably count which search engines are sending valuable visitors that act and contribute to the goals of your business.

KPIs and Metrics

One of the most misunderstood terms in analytics is the key performance indicator, or KPI. I've found this in almost every organization, regardless of size.

A KPI is the measurement that influences the next step in the process, or tracks toward the outcome. It is not a final measurement, as it indirectly influences the outcome. For example, click‐through rate is a great KPI. Click‐through rate measures the percentage of people who see an ad and click on it to go to the landing page, giving you a quick measurement of the appeal of the content you've published and its influence in moving people to the next stage. It does not directly impact the conversion rate or the conversion itself; it only measures the rate that people are clicking on an ad to a landing page. It is a measurement along the pathway and holds indirect influence on the outcomes.

A metric is a number without context, or without a story. If I talk about having 2,000 visits to a website, that's a metric. When a number is used without the context of segment, source, behavior, outcome, or influence, then it is a metric. A metric is a measurement that is independent of an actionable outcome. The difference between these terms is the action that you can take to influence the outcome.

Stage 1: Impressions

At the top of the funnel is the method of acquiring visitors to a website (see Figure 2.8). In this case, it could be any type of paid media campaign. In most cases, when starting a campaign, you have to decide on the size and type of audience you will target, along with a corresponding budget to reach that audience.

CPM, or cost per mille (mille is Latin for 1,000), is the standard method of buying “impressions” or “views.”

When you see a campaign that is marked as $40 CPM, that means that you will be paying $40 for every 1,000 impressions or views. If a campaign reaches 100,000 impressions, you will be charged $4,000.

Now is also when I stress the importance of a baseline, average, or a comparative metric: 100,000 impressions and $40 CPM might look okay, but you need to have a comparison to know for sure!

When you have a baseline to compare with, you'll immediately know if a campaign is on track. By asking how many impressions are expected or what a typical campaign averages, you might be able to catch if the impressions are significantly higher or lower than expected. If you do find that the expected number of impressions is very different than the actual performance, what can you do?

Stage 2: Engagement

I don't like the word engagement. It has become a meaningless, catch‐all term for rationalizing campaign activity, and it rarely provides any insight into actual economic benefit to an organization. It is made even worse by the so‐called engagement metrics that lump separate actions into one measurement that is supposed to provide meaning, which it doesn't. It actually obscures meaning. See Figure 2.9.

The problem with engagement and engagement rate is that it combines multiple metrics into a single metric, thereby stripping it of any meaning. Typically, engagement rate measures the number of likes, comments, and shares and any other trackable action into a single measurement. Even worse is when a cost per engagement is applied, making an even more meaningless metric.

The issue is that these actions are not equal! Given a choice, I am sure that any marketer would choose to have more shares than likes or comments. Shares mean that people are presenting your content to their network, friends, and followers. Likes simply show that someone took the least committal action to support the post—you don't know why; you only know that they did. A share shows loyalty, enthusiasm, and an active recommendation of your content and your business. Plus, it has economic value!

Rather than relying on the deceptive “engagement rate,” focus on each individual metric and learn from it.

Stage 3: The Click

This is where you finally reach a trackable clickable action with direct impact on a business objective (see Figure 2.10). The click‐through on an ad, a post, or link is not only a KPI for the campaign, but also the measurement of the effectiveness of the ad or post in influencing someone to act! This is the summary measurement of the campaign's ability to persuade action and movement to the next stage.

The performance side of this stage is the click‐through rate, which is self‐explanatory. It is the rate at which people click the ad compared to the number of impressions or views. To reach this number, you divide the number of clicks by impressions and multiply by 100. This gives you the rate of clicks. If you have a 2 percent click‐through rate, then it means that about 2 out of 100 people who see the ad or post click‐through.

The causal factors provide immediate feedback for a lower than average click‐through rate. If people are seeing the ad, interacting with it, and not clicking on it, then it is simply entertainment. You've not provided a compelling need or reason for them to act. This culprit is your offer and call to action, but it may also go back to the picture and headline. Evaluate the message of the ad or post. Is it clear? Are you presenting a clear solution to the audience's need? If not, the click‐through rate will suffer.

The financial side of the click provides another layer of feedback. The cost‐per‐click (CPC) is found by dividing the number of clicks by the campaign cost. For example, if you spend $10,000 on a campaign that generates 1,200 clicks, then your average cost‐per‐click is $8.33. Most times, this number ends up on a report right next to the click‐through rate and is not applied properly. To apply the CPC properly, it needs to be compared to another CPC calculation, the CPC generated by other channels.

Take the average CPC and compare it to the average CPC of similar campaigns on other channels. It may look like this example:

By comparing the average CPC generated by other campaigns, you can start to see the trend and effectiveness of your campaign. If maximizing the efficiency of your spend is an important part of the campaign or objective, then this measurement is a critical factor in how you may decide to spend your budget. The CPC is a comparative measurement for the financial aspects of a campaign, allowing a quick assessment of the efficiency of the campaign. However, it is not the ultimate measurement of success! It is a KPI. This KPI is a step along the way to the goal of the campaign—the conversion. It only informs you of the factors that influence this stage and this milestone.

Stage 4: Conversion

This is where the sales pitch is made—on the landing page. It is also where the causal factors multiply exponentially. You've probably noticed that the closer that the visitor gets to a decision point, the causal factors that influence each stage increase. The landing page is no exception. Of course, if you need to improve the conversion rate of the landing page, you'll see that I've only listed a dozen factors that influence the conversion. There are many other factors, which is why conversion rate optimization is a skill unto itself and addressed in another chapter of this book. See Figure 2.11.

The click was simply getting them to the page that does the heavy lifting of persuasion. The conversion is the intended action for the visitor, a registration, subscription, account creation, lead form, or a sale. All your work to this point was simply getting people's attention and arousing their curiosity and interest. Now you make the case for commitment.

The conversion rate is the overall judgment of your campaign's ability to attract and convert leads, subscribers, or customers. Divide the number of conversions from the amount of clicks or page visits and multiply by 100. For example, 100 conversions generated by 2,000 clicks yields a 5 percent conversion rate. About 5 people out of 100 decide to convert. But is that good or bad? It depends. You might find some industry trends or averages, but these numbers are so unique to your company and your approach that it is difficult to find a reliable comparison. The best conversion rate comparison is your own. Are you doing better today than yesterday? This quarter compared to the same quarter last year? This campaign compared to another campaign? There you go!

Comparing your campaign to other campaigns is a great way to apply this measurement to compare your effectiveness across a variety of channels. It becomes even more compelling when you factor in the average cost per conversion, be it a registration, lead, or sale. This now provides you with tangible monetary evidence of the campaign, and it is most effective when measuring the cost of the conversion against the other channels you are using.

You can compare your cost per conversion to previous campaigns, and you can compare it to other campaign sources:

Your analysis and decisions of this comparison would be based on your business objectives.

This is vitally important if you are tracking the efficiency of a campaign. If your goal is to maximize your budget, then these numbers are important because they show you how you can adjust your budget to accommodate a lower cost per conversion. However, if you are only concerned with generating leads or customers, and budget or cost per lead is not a concern, then you would be primarily concerned with the numbers of leads or conversions produced by each channel and make your decision based on the production of the channel, not the economic efficiency. The data you need and how you will use it ties back to the organizational objectives and goals for the campaign.

Stage 5: Nurture

Regardless of capturing leads or making sales, nurturing is process of building and encouraging a relationship. Either you are persuading a new customer to be a regular customer or qualifying a lead through additional communications. The nurture process is very similar, and of course, hundreds of factors now influence this relationship‐building or lead vetting process.

This stage measures the follow‐up and development of the new conversion (see Figure 2.12). Depending on your organization, it will look different from others and is unique to your process. Creating measurements, associating them with action, and showing the economic impact are all vitally important at this stage.

Unfortunately, this is where I find many B2B marketers stretched to their limit and being asked to measure things that don't exist. For many B2B marketers, their primary task is to gain leads. For them, the financial metrics don't come into play because they are simply generating leads and handing them off to the sales team. They can track the source of the lead, but once it is handed off, the lead enters another system that the marketer will never see, and the tracking ends.

What puts these marketers in a bind is when management wants them to start measuring the quality of the leads rather than just the number of leads generated, and that's a big problem. First, no one typically defines quality—they just want it measured. Second, the conversion trail was broken. If an organization has multiple data silos that do not communicate—there is no centralized data source—then that question is impossible to answer. For B2B marketers to be able to measure the quality of a lead (according to the source or campaign that generated the lead), the marketer must be able to track that lead through the nurturing and qualification process (Market Qualified Lead—MQL, Sales Qualified Lead—SQL), all the way to the ultimate sales and agreement and the value of the sale produced. If marketers are not able to see that information, then they cannot produce a simple quality measurement.

Tracking at this level also poses an issue of the “chain of custody.” If you watch any crime dramas, then this term will be familiar to you. For the B2B marketer who hands off the lead, the custody of that lead and the associated data now belong to the sales department. Any assessment of “quality” must include the quality of the sales content, the frequency of contact, the salesperson responsible for follow‐up, and any other interactions. This leads down a path that many people don't want to deal with! However, CRM systems are answering this data dilemma as they become implemented across organizations and unify the data.

The Final Stage: Customer Lifetime Value or Lifetime Customer Value

Regardless of what you call it, customer lifetime value (CLV) or lifetime customer value (LCV), this measurement is the sum of the customer experience with your company. It is based on three primary customer measurements (see Figure 2.13):

• The average value of a purchase
• The number of purchases per year
• The average length (years) of customer retention

Capturing this number and the data involved provides a complete end‐to‐end view of the life cycle of the customer. As CRM systems are able to capture more and more data, the connection from account‐level analysis all the way back to the acquisition, campaign, or other attributable sources is becoming easier. This provides the data necessary to close the loop on the true value and efficiency of channels and campaigns. If tracking to this level is the desired objective of your organization, then utilizing the analytics combined with a CRM enables a clearer picture of the efficiency and profitability of campaigns, channels, and offers.