CallSource

CallSource

Performance Management for Sales and Marketing

Digital Management

Call Tracking and SEO: Don’t be Afraid

by

Search Engine Optimization – better known as SEO – is top of most marketers’ minds nowadays. To exist on the internet means to exist on page one of Google. And to do that, you must optimize your website for Google’s always-changing algorithms.

As a call tracking and business performance company, CallSource has knowledge of both the service of call tracking to help marketers and how SEO helps marketers as well. As such, there is a statement involving both of these that has to be made.

Call Tracking does not harm your SEO.

Call tracking and SEO working together; the rumor that call tracking harms your SEO is one of many myths about call tracking. Let’s dive into this topic and demystify this rumor to show that call tracking does not harm your SEO.

Call Tracking & SEO: The Beginning

In the earlier days of SEO, until a few years ago, it was always said that call tracking hurt SEO because of Google’s rule of businesses having a consistent NAP (name, address, and phone number) across all listings. Therefore, the consistency of a business’s information was key, and having multiple unique call tracking numbers on different listings was a big SEO no-no.

For example, if your main business line was listed on your website, but then you had different phone numbers in other business listings, Google would crawl all of these and negatively impact your SEO ranking because of the inconsistencies, making your business seem a bit less legitimate. As Google’s crawlers evolved, this practice was quickly stopped. This “fake fact” persists in the minds of folks who do not stay up-to-date in best SEO practices.

Call Tracking & SEO: Enhancements

Call tracking technologies, and SEO best practices have become more robust over the years, making these two work in unison much better than previously.

Dynamic Number Insertion

Call tracking is no longer strictly for static phone numbers – dynamic number insertion (DNI) has increased in popularity and its capabilities in the recent past.

Dynamic number insertion is a form of call tracking that uses a script that leaves your main business phone number in the background for Google to crawl; DNI displays a unique call tracking number to each user landing on the page.

DNI doesn’t conflict with any of SEO best practices since the script masks the actual number tied to the business. DNI provides deeper analytics and insights into your marketing, tying online and offline attribution.

The Evolution of SEO

The older days of SEO practices could be labeled a bit like the Wild West.

Content was being created for search engines and Google’s crawlers, instead of being created for people. To rank higher, marketers were creating things simply to follow the “rules” of what search engines are searching for, rather than the relevance or quality of the content. Things like keywords and backlinks were of highest priority, so the value of these web pages suffered and did not always deliver truly helpful results to users.

This led to a lot of use of applying “black hat” SEO practices to try to get a website to rank higher in Google. HubSpot has a great, succinct definition of black hat SEO:

“Black hat SEO is a practice against search engine guidelines, used to get a site ranking higher in search results. These unethical tactics don’t solve for the searcher and often end in a penalty from search engines. Black hat techniques include keyword stuffing, cloaking, and using private link networks.”

It is because of these bad practices that were running rampant that has caused SEO to evolve to look at the bigger picture, rather than just specific things being ranked as higher priority without much other context. Content that is created for real people, instead of just for search engines, will end up ranking higher organically because of its relevance and usefulness.

Google’s Updated NAP Rules

Google’s most recent guidelines for representing your business on Google states: “Provide a phone number that connects to your individual business location as directly as possible, or provide one website that represents your individual business location.”
Google also notes that a local phone number is preferred over a central call center helpline number whenever possible and that you may use additional phone numbers on Google My Business websites and other local services.

What does this mean?

Add your call tracking numbers (local prefixes are best) into the business phone number listing and then place your NAP (direct phone number) into the additional phone number sections within Google My Business.

Find the full step-by-step process of how to use a call tracking number in Google My Business here.

Google My Business is how businesses manage their local results and the phone numbers displayed in organic search results. An interesting note from Google’s language is that a phone number or a website are required. While this isn’t 100% clear, it could mean that you only need one NAP (Name, Address, Phone) match to be validated (phone or website). Most businesses haven’t even been claimed, verified, or setup on listing sites such as Google My Business correctly, so make sure to take some time to complete the process.

Download the top 7 places to display your static call tracking numbers online.

Google My Business and Call Tracking

Google wants to optimize its search engines for the ultimate customer experience – so, rules aren’t quite as inflexible as they used to be.

Today, companies are using call tracking to improve their online and offline marketing and the customer experience. Call tracking; especially DNI, is imperative for businesses that advertise online and offline.

Get started with call tracking to improve your SEO and customer experience

You no longer have to be afraid of using call tracking and bettering your SEO for your business – the two are not mutually exclusive.

Ready to have someone reach out to you? We’ll have a representative contact you, or feel free to reach us today at 888.788.0123 to learn more.

Start tracking your inbound calls.

Get Call Tracking.

I want to talk to someone about utilizing call tracking at my business.

How to Calculate Expected Customer Lifetime Value (CLV) for Contractual Businesses: A Marketer’s Guide

by

Learning the Customer Lifetime Value (CLV) formula will help marketers predict the future net profit of every customer in a contractual business and take action to improve the value of high-value customers across the business.

The formula to calculate the expected customer lifetime value for your contractual business as a marketer is:

Formula to calculate the expected customer lifetime value

Where EV(t) is the net income or Profit, S(t) is the survival probability or the retention rate, d is the discount rate, and t is the time period interval that you are using. Put another way, the definitional expression for expected customer lifetime value E(CLV) is:

Definitional expression for expected customer lifetime value E(CLV)

Got it? Now good luck!

…Just kidding.

You probably need a bit more background to understand this equation and its underlying principles behind the formula so that you aren’t just plugging in numbers without understanding all of the moving parts.

Want a simple plug-and-play excel template for expected customer lifetime value calculation? Get it here!

Why should I care about calculating the expected customer lifetime value for my business’s customers?

In order to promote your business with marketing campaigns, you must spend resources such as time, effort, and capital to attract customers. You want to ensure that all of these resources spent aren’t for nothing. Wouldn’t it be great to have a crystal ball to see the future for every customer and their potential value, so you know which customers are worth your time and effort by the amount they may spend on your products or services?

While that kind of magical crystal ball doesn’t exist, determining a customer’s expected lifetime value comes in a close second for a similar usefulness. Yet, the calculation of the E(CLV) alone is not the end, since customer value is finite – no business relationship lasts forever. Simply giving every customer a potential value is not enough, since you may or may not gain the benefit of this customer depending on a number of determining factors.

The day you acquire a new customer, their potential value starts decreasing. You can lose potential value at any point if the relationship is burned too quickly, the partnership soured, or the customer leaves you to go to a competitor. While a customer might look very profitable at first glance with really high potential value, it is not always so easy to capture that potential. It’s also not likely that you will ever capture all the potential value of any single customer in the lifetime you have a business relationship with them.

So to answer the question – you should care about calculating the expected lifetime value of your business’s customers because you could be leaving money on the table by not properly reaching customers that have high potential value, and throwing money away by spending frivolous resources on customers with low potential value.

Looking at each individual customer in the marketplace in this potential value fashion is generally called customer centricity. The way you track, measure, and apply customer-centric strategies is a function of using the expected customer lifetime value E(CLV).

Focusing on the Heterogeneous (Truly Unique) Segments of Your Customers

Not all customers have equal potential value let alone the actual value they pay you at the first purchase. Focusing on User Experience (UX), personalization, and dynamic marketing campaigns create a common misconception that you are working with customer centricity. While those put the customer at the center of marketing – it is the function of incorporating the potential value through E(CLV) that truly defines customer centricity.

The focus of customer centricity should begin in your customer relationship management platform (CRM) with the segmenting of your customer information. If you’ve ever sorted coins into the pennies, nickels, dimes, quarters, etc. and counted them up, this is an example of what you will aim to do with the customers in your database (whose potential value is a lot like the different values of coins). You are going to separate your customers by potential value E(CLV) so that your entire company can better understand the different customer centric tiers your company works with.

Segmenting your customers

The segments (like the coins) should be heterogeneous so that every segment is different enough from each other (all dimes together, all pennies together, etc.). Within each group, you might expect to see a lot of aspects of the customers in common other than just similar expected customer lifetime ranges.

The simplest way to get started with segmenting your customers is to group them by three categories: high value (highest potential value), moderate value (average potential value), and low value (lowest potential value). The decision points (what will separate your segments) are highly subjective to you and your business, so pick the initial break points but expect to adjust them in the future.

(Tip: You might also want to segment the customers further, so the groups are even more distinct from each other. See example in the table below)

Here are some additional combinations alongside value tiers that are targetable with advertising:

High-value Web Visitors Moderate-value Interests Low-value Industry
High-value Email Clickers Moderate-value Genders Low-value Age Groups
High-value Asset Downloaders Moderate-value Income Level Low-value Geographic Area
High-value In-market for X Moderate-value Job Title Low-value Products Purchased

After you have your segments, what do you see? Looking at your high potential value customers, do you see accounts that:

  • Have been loyal?
  • Tend to interact with your communications?
  • Come to view you as a partnership?
  • Own more than one of your products or services?
  • Maximize the use of your products or services?
  • Tend to refer your business to others?

You could look at each tier of customers this way and start asking these initial questions. The enlightening part of this journey is that once you start to ask the right questions, the solutions, processes, and marketing campaigns start to reveal themselves. Every business has the constraints of resources and capabilities, but once you see what composes your individual customers, you can efficiently act to capture additional value.

How to get started in order to calculate the expected customer lifetime value

Whether you are a marketer or in some other position, you should not be a lone wolf in attempting to make your company more customer-centric. It’s important that you manage your expectations and realize that you cannot do it alone forever – you need buy-in from the top and from the stakeholders in your organization because customer centricity is not localized to marketing or finance alone.

With that said, here are the basic components you will need to be able to calculate the E(CLV):

  • Your business must be contractual (subscription based), or you must be able to tie purchases to individual customers. If you are not able to, while it’s still possible to figure out their E(CLV), it’s far more theoretical and I will not be tackling that in this article (but feel free to reach out if you want guidance).
  • You must have and use a CRM (like Salesforce or similar) and;
    • The CRM must store customers uniquely (little to no duplication).
    • The CRM must contain the status of the customer’s relationship (Ex: current customer, cancelled customer, or prospective customer) along with dates of those status changes.
    • The CRM (or you must have a way to match it to customers from another system) must contain the revenues and expenditures (think Net Profit or Net Cash Flow) of each of your current and cancelled customers broken down by your desired time periods (so not a single number in aggregate).
    • The heterogeneous segments in your CRM must be as unique and distinct from each other as possible.
    • Finally, the CRM must contain a large enough sample of the status type of cancelled/discontinued customers for you to get a healthy prediction since it will be based on this.
  • You must have access to all of the information above and meet with finance infrequently to obtain a useful discount rate.
  • Lastly, you must have a way to calculate it. Start with Excel, and maybe ask someone for help (like the author of this article!) if you need it.

When calculating the expected customer lifetime value, you are using the present and historical customer data in your CRM to make a calculated prediction about the future and potential value of your customer segments. You are using the cancelled customers in your segments as the baselines for the survival (retention) rates in each segment. This alone is why segments that are not unique enough from each other will abstract the final calculation of E(CLV).

(Tip: To find out if your segments are not unique enough, try segmenting further with data you have in your CRM by experimenting with different combinations such as demographics or behavioral engagements that you have tracked.)

Customers you have newly acquired may also skew results because they have not been customers long enough to give you quality data. For companies that are brand new, it’s more important that you set yourself up for calculating this in the future than it will be to get an accurate calculation in the first year.

Setting up your tables in Excel to calculate the expected customer lifetime value

As long as you are within the row and column limits of your Excel you can calculate the expected customer lifetime value and graph the results. The following example table is how you should organize the data out of your CRM:

Segment Time (t) Profit(EV(t)) Survival/Retention S(t) Discount rate (d) E(CLV)
High 1 $$$$ 0.## 0.## $$
Med 1 $$$ 0.## 0.## $$
Low 1 $$ 0.## 0.## $$

On the top you want to label your columns (I’ve added formulaic representations to help you). On the left are your dimensions with segments listed and the corresponding time periods. You want to make sure that for each segment you have all the periods aligned with that segment. The Profit is the net cash flow for that period of time (Revenues – Expenditures – Taxes). Remember that this value is subjective to your organization and could be computed differently by company. Finally, add your survival probability (retention rate) and the discount rate (you should get this from finance).

Enter the E(CLV) formula that mentioned at the beginning into the final column and make sure that it pulls from the cells in its own row. Once the formulas run, you should see your E(CLV) amounts by segment and time period.

How do I calculate the customer survival probability (retention rate)?

The survival probability or the retention rate is how likely that segment is to survive into the next period. To get the survival probability (retention rate) for each period you will look at the historical information you have and use the following calculations based on the data in your CRM:

Calculating the retention rate of your Client Customers

You can only do this calculation if you have the statuses of who is a prospective customer, a current client, and a cancelled customer.

As you can see, the major decider of a high or low survival rate (retention rate) is the amount of cancelled customers you had in each time period. From this metric alone you might be able to notice seasonal patterns or buying trends depending on how aggregated your time period is.

The life of finance circles around annual periods that are broken up by quarters (Q1, Q2, Q3, and Q4), but could also turn into months (1, 2, 3, …12) or weeks (1, 2, 3, 4, …52). The time period is sensitive to how quality your data is and the volume of how much data you have per segment. The more aggregate you go (annual, quarters), the more context you lose. Yet, the finer you go (days, weeks), the more quality data you need to have. Again, this is one of those areas that is subjective to your business.

I’ve calculated E(CLV) for my segments, now what?

First things first, give yourself a pat on the back, share your success here on the blog or let us know on social media that you’ve accomplished this, because we’d like to applaud your efforts. In terms of businesses who are capable of the feat of calculating customer value – you rank in the top globally.

If you have not already, graph your results of expected lifetime value.

(Tip: I recommend a waterfall chart to start as it gives you a nice little representation of the values over your time periods).

I created simulated information on my own so that I could share the results and talk about the value in obtaining this data. Below you will find the simulation results that I ran for two segments (Cohort A, and Cohort B).

Cohort A is akin to a high-value customer segment. Cohort B is similar to that of a low-value customer segment (These two are sort of like dividing your customer value database into two).
The shocking difference is just how much better the high-value customers are than the low-value customers.

Difference in high-value customer segment - Cohort A

Difference in low-value customer segment - Cohort B

I was shocked to see that in this simulation, the Cohort A (high-value segment) is light years better than the Cohort B (low-value segment). In fact, the low-value segment is actually losing money over time and unable to make up for the initial acquisition cost.

Combined, the customers are net profit $133K per year (12 t periods) – so on the surface, this simulation wouldn’t appear that bad. However, when we segment the simulation into two groups, we see that the Cohort A carries all the weight with $180k per year in profits and Cohort B is dragging all of the profits down. What would you do if you found out this was happening in your customers?

Interested in our excel template for the calculation? Fill out this form so we can send you the simple way of figuring this out!

Now, not so fast… do not fall into the trap of wanting to remove all of the bad customers you have that are unprofitable until you properly examine their contributions.

A good example is when a grocery store finds out that all of its high-value customers are those that buy soda, chips, and diapers. Taking this insight, it decides to turn the entire store into offering only soda, chips, and diapers. The sales plummet, and customers leave in droves. But wait – isn’t that what the best customers wanted? So why didn’t that strategy work?

Customer preferences, loyalties, and behaviors

It does not work because even though customers may have preferences, loyalties, and behaviors that you can measure, they also want the opportunity to do more. Customers may not have been buying products outside of soda, chips, and diapers at high frequency, but they were still buying them. They weren’t coming solely for the soda, chips, and diapers – even if those purchases are most common.

Your least valuable customers might be that way because of your past decisions, alignment, and even the products or services you sell. Instead of sending all your low-value customers away, you should think about ways to reorganize their products, services, and offerings so that you can make them more profitable to your valuable customer. The lesson in your actual data can tell you a lot about customer behavior, customer insights, and how you go to market.

Next Steps?

The next step you should take is to try to go more granular with your segments if possible. You should also try your first experiments to see if you can affect the segments directly.

Finally, I would recommend that you take your expected customer lifetime value and apply backwards to all of the activities in your organization that directly and clearly impact the customer. You might find (for example) that your phone calls (with rich customer data and highly personal communication) have a high potential value across all your customers. If you knew that a phone call was associated with tons of customer value, that would be a sign that your phone is not only important but possibly vital (high correlation) to the success of creating value from your current and future customers.

(Tip: CallSource is the phone call attribution pro so ask us for help if you need it.)

Conclusions

Customers are all created with potential value that you can realize by segmenting your CRM by expected customer lifetime value. Expected customer lifetime value is a financial calculation that forecasts the potential value of a customer based on the present and historical data you have of current and cancelled customers.

Using the formula, you can predict the future value of a customer discounted to the present period. Businesses that focus on customer-centricity are aligned from top to bottom in using the expected customer lifetime value across their entire organization. The day a customer is created starts the clock on your ability to capture that value over time with targeted marketing and sales strategies.

Customer centricity is not a short-term or “get rich quick” proposition and requires your company to take accountability for the CRM and customer data. By using potential customer value, you can efficiently use your company’s valuable resources and capabilities to generate higher returns.

References:

  1. 2018 by Peter S. Fader and Sarah E. Toms, The Customer Centricity Playbook, Wharton University Press
  2. 2007, Estimating CLV using aggregated data: The Tuscan Lifestyles case revisited, Fader P.S., Hardie B.G.S., Jerath K., Journal of Interactive Marketing, 21 (3), pp. 55-71.

What is a Dynamic Number Insertion Script and How Do I Use It?

by

TL;DR (Too Long; Didn’t Read) Version Here

Dynamic Number Insertion, also known as DNI, is a call tracking solution used by businesses that seamlessly ties offline and online attribution.

DNI enables businesses to attribute phone calls generated from PPC ads, SEO campaigns, retargeted ads, and more to measure success for each digital marketing channel. This call tracking technology aids in multi-channel and multi-touch marketing attribution to show the complete customer journey online to offline.

We have a podcast episode dedicated to explaining how dynamic number insertion (DNI) works for online call tracking purposes.

Basically, DNI is an accurate way to track consumers who call your business after discovering your number online. Dynamic numbers allow marketers to discover all the actions a user took before calling. With DNI, you can see which ads, keywords, and online content tie into generating inbound calls and should get credit in marketing attribution. For marketers that are unsure whether or not they are comfortable using call tracking because of SEO, DNI is the perfect solution.

How does dynamic number insertion work?

When a lead comes to your website, DNI will display a unique phone number to each user. When called, this toll-free number routes to the main line of your choosing (usually your main business line).

Dynamic number insertion uses a pool of toll-free numbers in the background. The volume of numbers provided is based on your website volume to ensure that each user or source gets assigned its own specific number to be tracked from digital to offline conversion. A phone number from the pool will display after being dynamically switched based on the user or source. This ensures that each user or source gets its own unique number.

Some number pools (like ours) are managed automatically, so you do not need to know ahead of time how many visitors you are expecting. Most other solutions require you to provision the amount of numbers in your pool ahead of time.

Dynamic Number Insertion Capabilities

Depending on what type of DNI solution you get, the availability of analytics and its capabilities may differ.

For example, in the most simple dynamic number insertion solution, a user may see a distinct phone number based on ads they clicked, keywords they searched, or some other marketing campaign that has been set up. Each of these will have their own unique number to be displayed to the user and will tie back to show you analytics based on these specific campaigns.

Once you get to more robust DNI solutions, you can see more details on the consumer journey based on an actual specific user, rather than just the campaigns you have set up with DNI.

These DNI solutions will assign a unique dynamic number per user, which is usually based on cookies (keep in mind, there are some problems with attribution in cookie-based tracking). By collecting cookies on a device, the same dynamic number will be displayed every time, allowing you to track the actions taken by a person this way.

Although some DNI solutions may slightly differ, the setup for dynamic number insertion is basically the same, and only takes a few steps.

DNI Set Up & Use

  1. Determine your highest daily number of concurrent users on your website. This will be used to determine your dynamic number pool, which must be provisioned ahead of time with a manual DNI solution. With a more robust DNI solution, this step can be fully automated.
  2. Install a one-line snippet of JavaScript on your website.
  3. When a user visits your website, DNI’s JavaScript will detect the user and/or source they came from and swap out the phone number automatically for that unique user or source.
  4. When the user calls the dynamic phone number, you can tie back online actions made before the call to your campaigns, ad sets, ads, keywords, or even consumers – depending on the solution.

A more reliable DNI solution – without cookies

As it currently stands, not all solutions are created equal. A more reliable dynamic number solution does not use cookies, allowing the data delivered to be highly accurate and person-specific instead of “user” specific. There are enough problems with attribution: browsers changing, devices changing, cookie blocking, and even VPN services that get in the way of accurately determine marketing contribution.

How DNI works without cookies

With a dynamic number insertion solution that does not use cookies, the page loads with the DNI script, and users will start to be tracked in two ways:

  • ID (Unique Identifier) as a unique person without using cookies based on:
    • Device ID, IP, Location, and other discrete detection.
    • Cross-referencing a proprietary database and other indiscrete detection.
  • Sessions. These can be grouped to include:
    • Time/Dimensions
    • Activities/Interactions/Engagements
    • Properties (moving from different owned domains)

Information is stored about this unique user that allows not only cross-domain tracking but actual cross-device tracking as well. Instead of relying on cookies for a user, CallSource’s DNI will track a specific person no matter the device they are using at that moment: desktop, phone, tablet, etc. In addition, our solution does track users across browsers and VPNs. No other DNI solution can automatically do all of this – but it only matters if you want accurate and reliable data.

Want to accurately track real people – so you aren’t missing any of your marketing data?

For online to offline attribution, many types of conversion actions are tracked:

  • Chats
  • Emails
  • Calls
  • Form Fills
  • Texts

Automating the pool of numbers can also help so you don’t have to rely on the specific number of dynamic numbers that you first allotted with your DNI solution. Why is this important?

CallSource Visitor Chart

If you end up having more concurrent visitors on your site than originally allotted for, a regular DNI solution would end up having to assign the same number to multiple users – not an ideal situation. The whole point of using DNI is to map and attribute actions back to the sources. If people are getting assigned the same number, you would lose this insight completely.

Dynamic number insertion script

Typically, DNI uses lines of JavaScript code to be enabled. This code is responsible for swapping out the phone number seen to the end-user depending on your DNI setup.

Dynamic number insertion simply needs one script to give you multitudes of data. Compared to static number call tracking where you have to manage every page and every unique number to make sure it’s in the right place, with DNI, you can now just throw the script on and forget about it.

Dynamic number insertion and call tracking

Dynamic numbers are typically used on your website’s landing page – but they aren’t necessary for all of your online marketing efforts.

Sometimes a static call tracking number works best in certain listing areas that your business appears online. Anywhere you cannot place the Javascript code would be where you place a static call tracking number.

Read our comprehensive guide for step-by-step instructions for listing call tracking numbers in the most popular business listings:

Using Call Tracking Online: 7 Popular Listing Services

Can I use both call tracking and DNI?

Yes, you can – and should – utilize call tracking and dynamic number insertion at the same time! As mentioned, both of these solutions have similar, but different, outcomes and reasons for using them – and will be used for different marketing campaigns and attribution.

Call tracking is, of course, the most reliable way to track your offline attribution, and should also be used on various other online sources for people to contact your business from.

Want to start using DNI?

Do you think that your website could benefit from dynamic number insertion? You can begin implementing DNI for your business in no time and with minimal effort.

Click here to have someone reach out to you, or contact a representative today at 888.788.0123 to learn more and compare our solution to anything out there.

Summary

What is dynamic number insertion?

Dynamic number insertion (DNI) is a type of call tracking that ties users’ online actions to their phone call to your business. DNI can track many types of actions such as channels visited, keywords searched, and ads clicked on prior to calling the dynamic tracking number.

How does dynamic number insertion work?

Dynamic number insertion swaps out phone numbers based on a unique user or source where the dynamic phone number appears. When this unique phone number is dialed, it connects to the main phone number you choose (typically your main business line), and you are able to view that individual’s online actions taken before calling the dynamic number.

Most DNI solutions rely on cookies to track unique “users,” but CallSource’s DNI solution uses a cookieless method of tracking for truly reliable data down to each person. We also use a dynamic pool of numbers to ensure that if your website traffic goes above the number of users originally allotted for, the same number will never be displayed to two different people.

How easy is it to install dynamic number insertion?

CallSource’s dynamic number insertion solution requires only one line of script to be enabled.

Why do I need dynamic number insertion?

Dynamic number insertion is a must-have for marketers that want to tie their online and offline attribution together for a more robust look into their customer journeys. It enables you to see what actions led a consumer to call your business.

Learn the complete consumer journey for more reliable marketing ROI.

Get Dynamic Number Insertion.

I want to talk to learn more about dynamic number insertion.

Properly Track the Results of Your Marketing: Google Analytics Channel Groupings Explained

by

Google classifies your website visitor traffic into buckets based on rules – some simple and some complex (and still unknown). Understanding how they are grouped is essential to properly tracking your marketing campaigns.

website-visitors-graphic

Where did my website visitors come from?

Google Analytics classifies website visitors into the Default Channel Groupings which are the most common sources of website visitors to any website. When a website visitor navigates from web page to web page, information about their history is logged and made available to Google Analytics.

By the time someone arrives on your website, you will not get to see all of that visitor’s history but will be able to see the last source that brought that person to your website. You get the ‘last attribution’ for website traffic sources that brought someone to your website.

What are the Default Channel Groupings and what sources of traffic do they contain?

Every channel should be thought of as exclusive. This means that every channel will only contain a visitor once and so no matter the traffic source – the visitors must be placed into a bucket and only one bucket. Even though you might have visited many sources on your journey to a website – Google is only classifying you into the one channel you came from last.

Those exclusive default channel groupings explained:

  • Direct – Someone who visits your website by typing in the full URL into their web browser, clicks a saved bookmark, or when the referrer and source information is not known. Often where visitors who are clients, employees, or visitors have had their URL parameters stripped, some mobile traffic, application traffic, non-secured sources, and visitors who were logged in will show up. Direct is a soft catch-all for when Google Analytics is not certain if there were any other sources.
  • Other or (other) – Mentioning Other after Direct is purposeful because other is another form of catch-all for when a visitor comes to your website but the source is something unfamiliar to Google and it cannot determine what category the source falls into. New search engines, tools, and site discovery tools will end up in Other because Google knows it has a source and a referral; it just does not know how to classify it.
  • (not set) – The final catch-all category is a hard catch-all for when privacy, outdated browsers, pop-up blockers, tracking blocking browsers, poor internet access, firewalls, and other technological constraints limit and obstruct the ability for Google to determine anything about the visitors. The same visitor could come back and tracking could be perfect, but on a specific visit the visitor could fall into the (not set) category. While it is popular to either ignore (not set) or to spread its numbers into the other categories by weighted-averages – a lot of (not set) indicates that someone is wrong. Small numbers in proportions is to be expected and spikes are normal, but consistently high values could indicate an issue with your website or tracking deployment.
  • Email – Sources of visitor traffic where the link bringing the visitor has a parameter indicating that it is from an email. The most common parameter for this is utm_medium and the value is just, ’email.’ Another less common practice is that some email marketing systems actually integrate into Google so it will have the information it needs regardless of parameters that it was from an email. In some cases you might have to tell Google in the channel settings what parameter to look for if you do not use the medium of email.
  • Organic – Sources of visitor traffic originating from non-paid search engine queries. First, the web visitor has to type a query into a known search engine. Second, that visitor has to click a non-paid non-advertisement search result. Third, they have to arrive on your web page. Once all of this has occurred, Google will classify it as organic. Most search engines are known by Google, but there are new ones and changes to old ones that takes place all of the time so these numbers (like all of the numbers) are not 100% perfect.
  • Paid Search – Sources of visitor traffic originating from paid search engine queries. First, the web visitor has to type a query into a known search engine that hosts ads. Second, that visitor has to click on the search-based paid advertisement result that is tracked as ‘paid search.’ Third, that visitor has to arrive on your web page to be tracked. Only after all of this will Google know that visitor’s source is Paid Search. The most common paid search platforms are Google Ads/Adwords and Bing. The medium values Google Analytics looks for are: cpc, ppc, or paidsearch. Google has a paid search ad network in its platform that will automatically place traffic into this category based on the GCLID cookie.
  • Display – Sources of visitor traffic originating from advertisements that are images (not paid search), videos, or contain specific medium values. Display denotes advertisements that are visual and interactive in nature. An advertisement that does contain an image could be thought of as Display. The most common medium values for this are: display, cpm, and banner. Google has a display ad network that will automatically place traffic into this category based on the DCLID cookie.
  • Other Advertising – Sources of visitor traffic originating from unknown advertisement channels or when the specific medium values do not fall into the Paid Search or Display values. The most common medium values that end up in Other Advertising are: cpv, cpa, cpp, and content-text.
  • Affiliates – Sources of visitor traffic originating from exact match medium parameter values. The most common parameter value denoting affiliate in the medium parameter is: affiliate. Any variation of this must be added to the Affiliates setup in Google Analytics so the platform can recognize this type of traffic. Affiliates should be thought of as paid referrals while the actual Referral category is for non-paid referral traffic.
  • Referral – Visitor traffic originating from the exact medium value of referral and websites that are not: known search engines, websites that you own, subdomains of your website, and where the link was not placed by you (this is known by Google). Referral is often hard to explain because it can be grey of what is determined to be a true referral and one that is not. A partner company could be thought of as referral unless you are paying them (in a sense) and then it should be Affiliates. Google does not know who is being paid and not so the catch-all between the two is Referral.
  • Social – Visitor traffic originating from known social networks, when the social source referral parameter is Yes, or when the medium matches: social, social-network, social-media, sm, social network, or social media. The most popular networks that Google Analytics will pick up are Facebook, LinkedIn, Twitter, Instagram, Google Plus, and YouTube. Social is also a catch-all for social media visitors even when those social activities are paid. This is problematic to decipher between Social and what is Display because the visitor originates from social media, but they saw a placement that was a paid for advertisement. Some marketers have solved for this by changing the settings in the Display channel or by creating a Custom Channel called Paid Social. The trouble with custom channels is that they do not play nice with every report in Google Analytics (MCF Channels) and they can have heavily sampled data.

Why does it matter if my visitors are properly grouped into the right channels?

For a long time marketers were not confident about Google’s ability to properly classify website visitors into the correct channels. This lack of confidence spurred many decisions to ignore, joke about, and mock anyone who did use this data. In the past decade (yes that long), this has been improved to the point that marketers now have the tools to resolve issues in their own analytics platform. You now have the ability to ensure that your visitors end up in the right channels.

Completely understanding and feeling confident that your website visitors are being tracked in the right channels allows marketers to better understand their audience, gain deep insights into the customer experience they are delivering on the web, and can dictate decision making with wide reaching implications.

When your data is limited and you have small numbers of website visitors it is even more important to understand your data. A few visitors being misclassified can completely shift a result and throw the context of what happened off its true course. The difference between a conversion coming from Social and Display is huge if you are spending time and effort on one of those channels.

The inability to figure out which channel really contributed to the bottom line is now an inexcusable response for digital marketers. When you aren’t sure – go investigate, add UTM parameters to links, and validate you are getting data so you are sure.

Determining the channels driving traffic to your website is an important way to understand what is driving your business from online sources and where to put money and effort in the future. Break down the data into a story so it makes sense and your company can get behind the journey of what is taking place.

Want to learn more about Google Analytics reporting? Download our free guide: Top Reports to Look at in Google Analytics.

Why Your Brand and a Website Redesign Matter

by Leave a Comment

Is it time to refresh your website’s look and feel? Here’s why you may want to.

CallSource and Our Brand

Great companies are built on strong brands that influence customer choice and build loyalty. A strong brand is a competitive asset, improves market position, and customer loyalty.

So what? A brand is the sum of all the experiences of a customer’s journey with the brand. Every customer engagement with a brand (good and bad) contributes to a company’s reputation.

One of our goals as CallSource re-launched their marketing in 2016 was to communicate CallSource’s brand, mission, vision, and purpose — with a visual voice that is true to our brand values.

As a company, we came together to define our mission and vision. CallSource’s purpose was established many years ago by our founder, and it is still true today—“To enhance performance, accomplishments, and results for individuals and organizations in work and in life.”

Once established, marketing set out to create the brand guidelines and included our brand stance, promise, and attributes. A year later, we launched a website that speaks to who CallSource is as an organization.

Aligning the Brand with Our Website

When you have a great brand, you need to ensure your visual voice matches the brand’s core values. A business’s website is a huge visual representation of who your company is, how it speaks, and what it stands for. Your website should be synonymous with the value you add to customers.

A 2018 goal for the Marketing team was to redesign our website. We needed it not only from a technical point of view—UX, quick uptimes, mobile-friendly navigating, and general user experience—but also for an emotional, interactive experience to reflect the CallSource brand.

Marketing is ever-evolving, and a website is a major part of any content marketing strategy. With growing marketing stacks, a company needs to stay up-to-date with the latest technologies and trends.

New Website = Refreshed Brand

With a new website comes a whole new look and feel for customers who are re-discovering you or consumers who have not interacted with your company before.
When our customers think of CallSource, they think of a company that has integrity, is built on over twenty-five years’ of experience, and treats each client— regardless of size— with respect.

CallSource can transform your business’ performance and make a difference in the lives of your employees.

Setting Up Your Google Analytics Filters

by

How to setup some of the most common Google Analytics filters and how you can use them to customize your own view filters and organize your traffic for marketing success.

Setting up Google Analytics is simple: you add the script to the right place on your website and you are good to go. By default, using Google Analytics will collect data from all website visitors from the day you enable it.

As intended, Google built its analytics platform to be able to adapt to the needs of most businesses both large and small. Marketers have found uses to maximize and clean up the platform for their companies which have made their way into the standard platform we have today.

Filtered Views are best used by marketers for the following:

check-mark Testing how changes affect analytics data.
check-mark Error checking for analytics.

Most Common Google Analytics Filters (that we’ve come across)

  1. When do you not need to use a filter in Google Analytics?
  2. Excluding the Internal IP (so your employees clicks/visits aren’t skewing the results)
  3. Excluding the Development or Staging Sites (so your developers aren’t skewing results)
  4. Including Hostnames & Subdomains in Reports (so you can see cross-domain hostnames in reports and ensures tracking code reliability for subdomains)
  5. Normalizing the Ending Slash (so your pages have a “/” at the end by default)
  6. Lowercasing the Pages (normalizes the page names with case-sensitivity)
  7. Lowercasing the UTM Parameters (so your data is normalized with case-sensitivity)

What is a Filter?

Filters are a way that Google Analytics has provided Marketers with the ability to organize their website traffic. The structure of every Google Analytics instance is as follows:

Account > Property > View

google-accounts-properties-views

The Account

The account is the highest tier in Google Analytics and represents the overall company or entity that governs all of the properties beneath it. Users who have access to Google Analytics accounts can modify the overall settings and control filters that affect every property beneath them. You can have access (per user) to 100 maximum accounts.

The Property

The property is the second tier in Google Analytics and represents the entity tied to a single tracking code. Often this aligns with individual hostnames and domains (Ex: CallSource.com). Users who have access to Google Analytics properties can modify the property settings and how it connects and integrates with other tools. The maximum is 50 per account.

The View

The view is the subset of an individual property which has its own configurations, settings, and definitions of data within the property. Users who have access to Google Analytics views can modify only the view settings, configurations, and definitions within that view. The maximum is 25 per property.

View Filters

Filters are tools provided by Google Analytics to limit, control, and modify the data being collected within the View. Filters are only available within the View settings but you can see the overall Filters at the Account level.

The biggest limitations of using filters is that the actions you take with filters alters data permanently going forward. Adjusting a filter in the future will not apply the new filtering rules retroactively (Filters do not apply to backdated data only data moving forward). Filtering applies to the data after it has been processed (usually 24-72 hours after collection). If you messed up in the past without a raw view(s), then you are in trouble (I’ll show you how to protect yourself).

What is the difference between filters and sampled segments?

  • Filters can only include or exclude based on dimension fields.
  • Filters only apply to the present and future facing data.
  • Filters require a user with higher level (edit access) to create and modify.
  • Filters cannot be shared.
  • Filters apply to near endless timeframes going forward (are not limited to a window of time)
  • Filters are not subject to additional sampling. (This is huge!)

Setting Up and Managing Multiple Views

raw-testing-views

This is important! Before you start creating filters… please setup additional views! At the very least, create two additional views called: Raw and Testing. The Raw View will be the view where you have no filters applied (that’s right, 0!). You need this one because if you ever need to see what change filters had on your data you need a ‘backup.’ The Raw View provides this backup.

The second view you need to make is a testing view where you can play, mess up, change, and alter filters without messing up your primary views. The Testing View is especially important as you add new tools and integrations because filters can sometimes (they do) break down due to filters. Try to keep the Testing View empty after you’ve testing things unless you are running a long term test.

Wait – How Do I Create New Views in Google Analytics?

how-to-create-views

Here are the steps to locate and create Views in Google Analytics:

  1. Login to your Analytics (https://analytics.google.com).
  2. In the lower right of your screen (Desktop) navigate to the Admin Panel (Gear Icon).
  3. From the Admin Panel, the furthest right column is for Views, click the blue (+ Create View) button.
  4. Select what data you are wanting to track (Website or Mobile App, etc).
  5. Name the New View, and set a time zone (Usually to Your Time Zone), then save it.
  6. Select the New View by what you named it, using the dropdown beneath the blue (+ Create View) button from before.

 

Where Do I Find View Filters in Google Analytics?

Now that you have some Views ready for filtering you can use the Testing View to add some filters and make sure that nothing in the primary views will be affected. Adding filters is up to you – don’t just add everything without considering if it will impact something important.

Here are the steps to locate and create a filter:

how-to-create-view-filters

  1. Login to your Analytics (https://analytics.google.com).
  2. In the Lower right of your screen (Desktop) Navigate to the Admin Panel (Gear Icon).
  3. From the Admin Panel, the furthest right column is for Views, make sure you are in the right view, the fifth row item down is called Filters (funnel icon). Click this.
  4. The Filters are ordered into a ranking of what will run in what order (order matters a lot).
  5. In red (currently) click the button (+ ADD FILTER) to create a filter for the view you intend.
  6. You will create a filter, verify it, and save it each time.
  7. If you ever need to change the order, click the text beside the button “Assign Filter Order” to reorder the filters.

What are the Required Fields in a New Filter in Google Analytics?

When you create a brand new Filter in Google Analytics the first field you will always need to fill out is the Filter Name. Next you will be required to select if the Filter Type will use predefined fields and rules or if you are going to go ‘full custom.’ Predefined selections are limited to filtering traffic exclusively from:

  • ISP Domains
  • IP Addresses
  • Subdirectories
  • Hostnames

create-predefined-filters

Getting started the predefined selections might suffice for basic purposes but will not suffice for more complex filtering needs.

While it is not required using the Filter Verification and ‘Verifying this filter’ could save you a lot of time. The Verification will apply the new rule to the past 7 days of data and pulls a small sample of what changed to simulate how it could affect you. Since it does not give you a full picture, it’s more directional if the filter works or did not worked. Sometimes there is not data because what would have been filtered has not occurred yet.

 

When do you not need to use a filter in Google Analytics?

We have found that some cases are not necessary for filtering. Sometimes filtering is either overkill or Google has another feature that negates the use of filtering. Due to sampling, simply creating a segmented view of the data is prohibitive since sampled data is not always reliable (sometimes in rare circumstances the sampling is 99.99% so it’s okay).

The best example of this is excluding URL Query Parameters which can be achieved using a feature of Google Analytics. Navigate to your Google Analytics in the Admin panel:

exclude-query-parameters-built-in

View > Select View > Click View Settings > Scroll Down to “Exclude URL Query Parameters”

Inside the Text Area box you can add your Query parameters you want removed separated or delimitated by commas. We have seen email automation queries in there, language modifiers, and other personalization-based query parameters in this box so they do not add unnecessary pages in reports. If you see a lot of pages in reports with query parameters, consider adding them here.

The Order Matters with Multiple Filters

Each Filter is not applied all at once or independent of other Filters. Instead Filters work in the order they are Ranked, one at a time. This is a big reason why it’s important to build the filters in a Testing View before you move them to your primary property (as mentioned).

Here is a visual example of how multiple filtering works:
Example #1: Assume you need to include two hostnames (two domains) that are ‘firstdomain.com’ and ‘seconddomain.com’ in a filtered view so you can see both of them in reporting.

If you run the filters in the following order what happens?

filtering-order-hostname-1

  1. Include hostname contains firstdomain.com
  2. Include hostname contains seconddomain.com

Result: FAIL only traffic from the first step will be in reporting = data from ‘firstdomain.com’

What about if you flip the filters into the reverse order what happens?

filtering-order-hostname-2

  1. Include hostname contains seconddomain.com
  2. Include hostname contains firstdomain.com

Result: FAIL only traffic from the first step will again be in reporting = data from ‘seconddomain.com’

In order to get data from two values you need to put them into a single REGEX-based filter. The result would look like:

filter-order-hostname-3

  1. Include hostname regex: firstdomain.com|seconddomain.com

Result: SUCCESS traffic from both domains will be in reporting = data from ‘firstdomain.com’ and ‘seconddomain.com’

Example #2: Assume you needed traffic from the page /cars/ but wanted to exclude traffic from the page /car/

If you ran the filters below in the following order what happens?

filter-order-request-uri-1

  1. Exclude Request URI car
  2. Include Request URI cars

Result: FAIL no data from /cars/ would not show up because the word ‘car’ is found in the word ‘cars’ so the first exclusion would remove data from both ‘car’ and ‘cars’ pages.

If you run the filters below in reverse order what will happen?

filter-order-request-uri-2

  1. Include Request URI cars
  2. Exclude Request URI car

Result: FAIL no data from /cars/ would show up because the first step does limit the pages to those with ‘cars’ but the second step, as before, finds ‘car’ in the word ‘cars’ and excludes it.

So how do I create the successful filter?

The example I choose was simplistic to emphasize the importance of the order but the solution is actually pretty simple. The trouble is using words without placement of the word makes it difficult to create a filter. When you create a filter make sure you test it in a regex tester to validate that you have it correct. My favorite RegEx tester is RegEx Pal (https://www.regexpal.com/) which allows you to create an expression at the top and then write different test strings to see if your expression worked.

The correct expression (works in reverse):

filter-order-request-uri-3

  1. Include Request URI /cars/
  2. Exclude Request URI /car/

Result: SUCCESS you will see data only from /cars/ but not from /cars/ because the whole word ‘/car/’ with the forward slashes is not found in /cars/.

How do I test to make sure that my Filters are working as intended?

In many cases you will not be able to stare at the filter logic and know that it will be successful or fail. Sometimes it seems like it has succeeded and you will have to check it against the RAW View to be certain. When creating and testing Filter Views it helps to have a process of how to create and diagnose your filters that saves you time and efforts.

Here are some steps I take to validate filters as I create them:

  1. Always have a RAW View in the property I’m adding Filters to
  2. Always first apply the Filters to the Testing View
  3. Test the results using the Real-Time Data View in Google Analytics by replicating an environment that would should be filtered out (not always possible)
    1. Alternatively… open up the relevant report in Google Analytics and apply my RegEx to the advanced settings of the correct dimension.
  4. If I cannot verify immediately give the Test View a few days or weeks to validate.
  5. Compare the RAW to the Testing View to see if things are working.
  6. If needed Create another Testing View with the opposite (if it’s include, make it exclude) Filter Logic and check the data that way.

Remember… test test test!

 

Setting up each of the Most Common Filters in Google Analytics

filter-internal-ip

Excluding the Internal IP (so your employees clicks/visits aren’t skewing the results)

Why should I use this filter?

Organizations that test their own systems and run through their marketing processes will leave view, clicks, form fills, and more behind. You’ll see landing pages with high clicks, high conversion rates, and emails with amazing delivery. While internal testing can account for seeming negligible data it can still skew results. Unless you plan to adjust dates or remove the data later from reports then adding this filter cleans up your data automatically.

Additional Tip:

Before you create this filter, add a view to your property called “Internal Only” and only include traffic from your internal IP. That way, if you ever wanted to see the internal traffic you can load that view and see it. This might also allow you to see internal pages or internal wiki and media assets to know what your internal team is looking at.

Creating the Filter (Step by Step):

  1. You will need to know the IP Address(es) of your internal organization that it uses to connect to the internet. If you do not know ask your IT professional. You might try searching on Google for it but this might not be your actual IP (https://www.google.com/search?q=whats+my+ip&oq=whats+my+ip)
  2. Name the Filter, you can use, “Exclude Internal IP”
  3. If you have a single IP Address it’s simple to use the Predefined Filter Type:
    1. Set to Exclude
    2. Set Traffic to, “traffic from the IP addresses”
    3. Set Match to, “that are equal to”
    4. Set IP address to your single IP Address
  4. If you have multiple IP Addresses then use the Custom Filter Type:
    1. Select Exclude
    2. Set Filter Field to, “IP Address”
    3. In the Filter Pattern type your IP Addresses using RegEx each separated using ‘|’ between each IP and each number of the IP escaped by a backslash ‘\’:
      1. Example: 100\.100\.100\.100
      2. If you need help with this there is a tool that will convert your IP Address into the RegEx, 3rd party tool here (http://www.analyticsmarket.com/freetools/ipregex)
  5. Save it

analyticsmarket-regex-ip-address

What about the order of this filter?

Due to this being an overarching filter and if there are not multiple filters for IP Address then you could leave this at the top of the rankings so it runs first. Please… test this and run all of your initial filters in a test view to make sure they work as intended.

 

Excluding the Development or Staging Sites (so your developers aren’t skewing results)

filter-staging-development

Why should I use this filter?

For the same reasons that you might exclude Internal Traffic from a view you might want to exclude development and staging sites from your analytics. Undeveloped and ‘Work in Progress’ development environments can easily skew (even in small numbers) the data because of bugs and issues. If you plan on removing it later then you might not need to add this filter.

Additional Tip:

Before you create this filter you need to confirm with your developers what denotes a development or staging environment from the live site. If it’s a subdomain, a query parameter, or specific pages then you need to know this because they have different setups.

Creating the Filter(s) (Step by Step):

  1. Determine what denotes your Development or Staging environments
  2. Name the Filter, you can use, “Exclude Staging Development”
  3. You (most likely) must use the Custom Filter Type.
  4. If you use a subdomain for staging:
    1. Set to Exclude
    2. Set Filter Field to, “Hostname”
    3. In the Filter Pattern add RegEx for all of your staging subdomains:
      1. Example: ^staging\.yoursite\.com$|^development\.yoursite\.com$
      2. Example: (^staging|^development)\.yoursite\.com$
  5. filter-staging-development-uriIf you use a page for staging:
    1. Set to Exclude
    2. Set Filter Field to, “Request URI”
    3. In the Filter Pattern add Regex for each of your known staging subdomains:
      1. Example: /(staging|development)/
  6. Validate and Save

What about the order of this filter?

The order of this filter is determined by the position of other filters that interact with either the Hostname or the Request URI. You might want to place this higher than filters that add or replace elements of either. Always test if you are unsure first in the Testing View.

 

 

 

 

Including Hostnames & Subdomains in Reports (so you can see cross-domain hostnames in reports and ensures tracking code reliability for subdomains)

include-hostnames

Why should I use this filter?

Whether or not to use the include Hostname filter depends on if you can answer yes to two questions: are you confident that your tracking code is only on your active and current domains? And – do you have your property on a single domain with no other subdomains? If you answered yes to both then you likely do not need this Filter. Otherwise, try testing this filter out to see if you can get cleaned up cross-domain data.

Additional Tip:

referral-exclusion-list

Once you have a list of all of your domains, subdomains, and even link shorteners like Bit.Ly that touch your analytics property then you need to add them to your referral exclusion list. Inside Google Analytics > Property > Tracking Info > Referral Exclusion List you need to add each of those domains to the referral exclusions. You can always check your referral domains in your reports to see if you ever miss any.

Creating the Filter (Step by Step):

  1. Gather your list of domains and subdomains
  2. Name the Filter, you can use “Include Hostnames”
  3. You must use the Custom Filter Type.
  4. Select Include
  5. Set Filter Field to, “Hostname”
  6. In the Filter Pattern add RegEx for all of your subdomains and domains, along with some google domains that cache and serve data:
    1. Example: (www|blog|community|subdomains)\.yoursite\.com|yoursite\.com|translate\.googleusercontent\.com|webcache\.googleusercontent\.com
  7. Verify and Save

What about the order of this filter?

The order of this filter is determined by the position of other filters that interact with the Hostname. You might want to place this higher than filters that add or replace elements of the hostname. Always test if you are unsure first in the Testing View.

 

Normalizing the Ending Slash (so your pages have a “/” at the end by default)

normalize-ending-slash

Why should I use this filter?

Did you know, or are you frustrated that if someone visits the page /contact and someone else visits the same page /contact/ that Google Analytics will load both interactions but they will be noted as two separate pages?

When you look up page views and conversions from that page you will have to combine the data outside of the report to get the metrics. However, this Filter normalizes your pages that have and links that do not have the ending forward slash “/”.

Additional Tip:

It takes more than one person to run marketing. For those of you doing it all… hats off to you. For teams with multiple people and links being created by several people – possibly people across the country from you this Filter reduces the micromanagement around having to make everything match.

Creating the Filter (Step by Step):

  1. Name the Filter, you can use “Normalize Ending Slash”
  2. You must use the Custom Filter Type.
  3. Select Advanced
  4. Set Select Field to, “Request URI” in Field A -> Extract A
  5. In the Filter Pattern add RegEx for detecting pages without a “/” forward slash.
    1. Example: ^([a-z0-9/_\-]*[^/])$
    2. Example (if above does not work): ^(/[a-zA-Z0-9/_\-]*[^/])$
  6. Leave the Field B -> Extract B blank and do not Set Anything in the Select Field
  7. Set Select Field in Output To -> Constructor to, “Request URI”
  8. In the Filter Pattern type: $A1/
  9. Make sure that Field A Required & Override Output Field are both checked
  10. Verify and Save

What about the order of this filter?

The order of this filter is determined by the position of other filters that interact with the Request URI. You might want to place this higher than filters that add or replace elements of the page or Request URI. Always test if you are unsure first in the Testing View.

 

Lowercasing the Pages (normalizes the page names with case-sensitivity)

normalize-page-urls

Why should I use this filter?

Normalizing page names keeps the pages consistent in reports and consolidates the duplication of pages that have different case-sensitive names.

Additional Tip:

If you have a use case for some pages to be first letter capitalized, then do not use this Filter. Using this filter makes your data more consistent and reduces needless duplication of pages in your reporting. The page URL is what is being considered here – not the actual name of the page by its title.

Creating the Filter (Step by Step):

  1. Name the Filter, you can use “Normalize Page URLs”
  2. You must use the Custom Filter Type.
  3. Select Lowercase
  4. Set the Filter Field to, “Request URI”
  5. Verify and Save

What about the order of this filter?

This filter does not impact other filters unless they are case-sensitive. Test in the Testing View to be certain.

 

Lowercasing the Query Parameters (so your data is normalized with case-sensitivity)

normalize-utms

Why should I use this filter?

UTM Parameters are easily mixed up, capitalized, and incorrectly typed. If you have one person doing them for Social Media, another for Advertising, and another for Emails you will have a lot of cross-over. This can compound the issues and makes normalizing the UTM parameters to be lowercase (as they should be) a must.

Additional Tip:

UTM Parameters should always be lowercase. Consider making yourself a tool in excel or google sheets that helps you keep track of your UTMs and build them that enforces lowercase. That way everyone can have their own tool but be forced into normalized UTMs at the beginning.

Creating the Filters (Step by Step):

  1. Name the Filters, you can use “Normalize UTM {Insert Parameter Name}”
  2. You must use the Custom Filter Type.
  3. Select Lowercase
  4. Set the Filter Field to each UTM (one at a time so each parameter needs its own filter):
    1. For Campaign, “Campaign Name”
    2. For Source, “Campaign Source”
    3. For Medium, “Campaign Medium”
    4. For Content, “Campaign Content”
    5. For Term, “Campaign Term”
  5. Verify and Save for Each Parameter

What about the order of this filter?

This filter does not impact other filters unless you are using Filters that rely on the Filter Fields (Campaign Name/Source/etc). Test in the Testing View to be certain.

What Filters Should I Not Use?

This is subjective, but from personal experience I can say that you should not use resolution (screen size) based filtering because it will negate any API/Webhooks you have coming into Google Analytics. A webhook of information coming from an integration will not have a screen size so the data it contains will be lost if you do not include it in or if you exclude it out.

Finally, there are circumstances where (if you are not specific enough) the filters do more harm than good. Remember, including means that only the matched filter pattern will be allowed and excluding means that anything besides the excluded filter pattern will be allowed into the final data.

Additional Views to Explore from Filtering

  • Mobile Only (Include Device Type) [Be careful because devices are not always identified well]
  • Blog Only (Include Blog)
  • United States Only (Include Country) [Be careful with this due to VPNs and other elements that can limit this]
  • Specific Language Only (Include Language) [Be careful with this as many people are bi-lingual]
  • Stripped Query (Search and Replace Query Strings)
  • Normalized Site Search (Lowercase Search Terms)
  • Paid Traffic Only (Include Source/Medium/Referral) [Be careful that you can identify this perfectly]
  • Browser Only (Include Browser) [Be careful because browsers are not always identified well]

Additional Custom Filters to Investigate

The community of Google Analytics helps Google by identifying spam, bots, web crawlers, and applications that should be filtered but are not. Currently the community has a comprehensive list of known spam, robots, and crawlers which are known to potentially inflate traffic counts.

Competitors are known to track their competition brands by scanning the changes of their website on a daily even hourly basis. Some of those filters are updated as new ones are discovered or Google lets the community know they have accounted for them.

The current list of spam, robots, and crawlers you can add filters for are found below with how to setup each filter. Please test these in case – you’ve been warned.

Known Spamfilter-spam

  1. Name the Filter, you can use “Block Spam #X”
  2. You must use the Custom Filter Type
  3. Select Exclude
  4. Set the Filter Field to, “Campaign Source”
  5. Type the following Filter Patterns:
    1. .*((darodar|priceg|buttons\-for(\-your)?\-website|makemoneyonline|blackhatworth|hulfingtonpost|o\-o\-6\-o\-o|(social|(simple|free|floating)\-share)\-buttons)\.com|econom\.co|ilovevitaly(\.co(m)?)|(ilovevitaly(\.ru))|(humanorightswatch|guardlink)\.org).*
    2. .*((best(websitesawards|\-seo\-(solution|offer))|get\-free(\-social)?\-traffic(\-now)?|googlsucks)\.com|(domination|torture)\.ml|((rapidgator\-)?(general)?porn(hub(\-)?forum)?|4webmasters)\.(ga|tk|org|uni)|(buy\-cheap\-online)\.info).*
    3. .*((event\-tracking|semalt(media)?|(100dollars|success)\-seo|chinese\-amezon|e\-buyeasy|rankings\-analytics|rednise|video\-\-production|theguardlan|webmaster\-traffic)\.com|traffic(monetize(r)?|2money)\.(org|com)|pops\.foundation|erot\.co).*
    4. .*(((free\-)?(floating|get\-your\-social)\-(share\-)?buttons|hosting\-tracker|alibestsale)\.(com|info)|(justprofit|best\-seo\-software)\.xyz|snip\.to|adf\.ly|copyrightclaims\.org|(black\-friday|cyber\-monday)\.ga).*
    5. .*((monitoring(-your)?-success|uptime|free-video-tool|hdmoviecams)\.com|(monetizationking|popads)\.net|rank-checker\.online|(marketland|dominateforex)\.ml|(ownshop|topquality|easycommerce)\.cf|increasewwwtraffic\.info|(unpredictable|getlamborghini)\.ga).*
    6. .*((eu-cookie-law-enforcement|social-traffic).*\.xyz|teedle\.co).*
    7. .*(semalt(media)?|buttons\-for\-website)\.com.*
    8. uptime(robot|bot|check|\-|\.com)|vitaly|sharebutton|semalt|ranksonic|share\-button|anticrawler|timer4web|free\-video\-tool|responsive\-test|dogsrun|fix\-website\-er|dailyrank|sitevaluation|99seo|top10\-way|seo(\-2\-0|\-analysis)\.
    9. (videos|buttons)\-for\-your|best\-seo\-(solution|offer)|buttons\-for\-website|profit\.xyz|dbutton|keywords\-monitoring|platezhka|7makemoney|forum69|kings\-analytics|checkpagerank|pr\-cy\.ru|\-\-(production|website|sale)\.com
    10. (express|audit|dollars|success|top1|amazon|commerce)\-seo|free\-video\-tool|datract|hacĸer|ɢoogl|slifty\.github|\-liar.ru|3\-letter\-|foxweber|free\-fbook|goodwriterssales|your\-rankings|tourcroatia|spinnerco|justkillingti|suralink|worldtraveler\.w
    11. oldfaithfultaxi|christopherlane|hollywoodweeklymagazine|losangeles\-ads|anniemation|timdreby|pcimforum|yellowstonesafaritours|autoseo|blogarama|for\-placing|brainwizard|casinos4|ḷ\.com|\-backlinks\.com|phoenicx\.co\.uk|be\-escorts|vidyoze
    12. brasseriebread|helvetiiconsulting|johntrapane|cloudsendchef|theautoprofit|:8888|blog1989|incomekey|amazon\-ads\.ovh|krumble\.net|10bestseo|seo\-watch|blog100|seoservices2018|resell\-seo
  6. Verify and Save

Known ISP Provider Crawlersfilter-isp-spam

  1. Name the Filter, you can use “ISP Spam”
  2. You must use the Custom Filter Type
  3. Select Exclude
  4. Set the Filter Field to, “ISP Organization”
  5. Type the following Filter Pattern:
    1. hubspot|^google\sllc$|^google\sinc\.$|alibaba\.com\sllc|ovh\shosting\sinc\.|microsoft\scorp|facebook\sireland\sltd
  6. Verify and Save

 

 

 

Known ISP Domain Robotsfilter-isp-robots

  1. Name the Filter, you can use “ISP Robots”
  2. You must use the Custom Filter Type
  3. Select Exclude
  4. Set the Filter Field to, “ISP Domain”
  5. Type the following Filter Pattern:
    1. paloaltonetworks|scaleway|kcura
  6. Verify and Save

 

 

 

 

Known Language Spamfilter-language-spam

  1. Name the Filter, you can use “Language Spam”
  2. You must use the Custom Filter Type
  3. Select Exclude
  4. Set the Filter Field to, “Language Settings”
  5. Type the following Filter Pattern:
    1. \s[^\s]*\s|.{15,}|\.|,|^c$
  6. Verify and Save

 

 

 

 

Conclusions

I tip I found online that I wanted to share is that if you are testing RegEx you can always bring up the related report in Google Analytics and open their advanced settings where you can enter your RegEx and see if it displays the right data.

Another rule of thumb is to try to only include one filter of each Filter Field Type. If you run out of room (as in the case of the spam/robots/crawlers above), then it makes sense to have multiples of the same type.

CallSource is not affiliated with Google Analytics, but we provide solutions for businesses that want to track and attribute their phone calls online.

If you are interested in learning more, contact a Specialist to find out about our non-cookie based digital attribution platform, AutoID.