UTM tags and their role in business intelligence

UTM tags are a basic element in the analysis of advertising campaigns. Without them, you can’t find out which ad sent a visitor to your site, can’t associate sessions with ad costs, and can’t build business intelligence.

In this article, we tell you what UTM tags are, why they’re used, and how they combine data from different sources. We also look at what errors to avoid when creating UTM tags and how to simplify work with dynamic parameters.

If you want to set up business intelligence and get full data in the correct format, try OWOX BI. OWOX recognizes dynamic parameters and checks UTM tags in your advertising campaigns. It then puts data in the correct format, monitors its relevance, and updates it retrospectively.

Table of contents

What are UTM tags and why are they needed?

A UTM tag is a special parameter that’s added to a URL after the question mark.


http://www. site. com/?utm_source=google&utm_medium=cpc&utm_campaign=TV&utm_term=TV&utm_content=samsung

The acronym UTM stands for Urchin Tracking Module. A bit of history: in 2005, Google bought Urchin Software because of its analytical system Urchin on Demand. This system subsequently formed the basis of Google Analytics. Following the release of Google Analytics, UTM tags have become a standard that marketers use to track transitions across various advertising campaigns, including offline.

There are five main types of UTM tags. Three of them are obligatory:

  1. utm_source points to the user’s transition source — that is, the site on which the ad is displayed
  2. utm_medium indicates the channel the user came from
  3. utm_campaign identifies the ad campaign from which the transition was made

And two tags are optional:

  1. utm_term shows the key phrase from the ad campaign
  2. utm_content identifies the advertising content item on which the user clicked

Learn more in our article on what UTM tags are and how to use them.

Why is proper UTM tagging essential for business intelligence?

To answer this question, you need to understand how business intelligence works. There are many ways to implement it. In this article, we discuss in detail the method that OWOX BI uses.

In short, the essence of business intelligence is to combine data from different sources: your site or mobile application, advertising sources, email and call tracking services, CRM/ERP systems. All this data is uploaded to a single repository (in our case, Google BigQuery) and then combined using some keys. You can then use this information in reports and analyze it in different segments.

How data is combined with OWOX BI:

how data is combined with OWOX BI

OWOX BI has a free trial period. Sign up and configure your business intelligence system.


Benefits of business intelligence

  • Analyze data in a single system
  • Accurately evaluate the effectiveness of advertising campaigns
  • See the user path from the first visit to purchase
  • Make decisions based on data, not intuition

Challenges and features of data consolidation

In theory, everything sounds simple enough, but in practice, you can face the following problems when setting up business intelligence:

  • Facebook, and other ad platforms don’t give advertising statistics in terms of UTM tags. We’ll explain why this is a problem a little later.
  • You can combine user session data with advertising cost data only using UTM tags. There’s no other way.
  • If there are no UTM tags or they aren’t set correctly, you can’t identify a campaign or ad. This means you won’t be able to correctly attribute costs to a session, meaning you can’t find out the cost of each session. Knowing the cost of each session is the basis for building business intelligence.

Questions UTM tags can answer

There are quite a lot of questions UTM tags can answer, including:

  • Which ad was responsible for the transition?
  • What region is the user from?
  • Which sites showed the ad from which the transition occurred? (This mostly concerns contextual advertising.)
  • On which type of devices is the conversion rate better?
  • What gender or age are the users who generate the most income?

Yes, you can answer some of these questions in other ways, but you can also use UTM tags. You can collect this data thanks to dynamic parameters.

Dynamic parameters and their use in advertising services

Marketers use dynamic parameters to transmit meaningful information that characterizes the user and the conditions in which the ad was displayed. These parameters are specified as UTM tag values in curly braces {}. Dynamic settings are established when you set up a campaign. When the ad is displayed, the advertising service places the parameter value in the braces.

Examples of Facebook dynamic parameters

  • ad_id={{ad. id}} is the ID of the ad
  • adset_id={{adset. id}} is the ID of the ad set
  • campaign_id={{campaign. id}} is the id of the advertising campaign
  • ad_name={{ad. name}} is the name of the ad
  • adset_name={{adset. name}} is the name of the ad set
  • campaign_name={{campaign. name}} is the name of the advertising campaign

We’ve cited only one advertising service, but dynamic parameters are used by many large ad services including Google, and Bing.

How OWOX BI works with UTM tags

OWOX BI helps you automatically upload cost data from advertising services to Google Analytics and Google BigQuery as well as transfer data about user behavior on your site to BigQuery.

Using OWOX BI, you get complete and high-quality data:

  • To link session and cost data, OWOX BI recognizes UTM tags in your ads. No advertising service API gives tags with dynamic parameters in their final form. That is, instead of the value of the parameter in the link, there’s only the parameter name. OWOX knows how to recognize these dynamic parameters. When it uploads cost data from an advertising service and encounters a link with dynamic parameters, it can determine their value. This allows you to view cost data in UTM tags in Google Analytics reports. See our Help Center for a list of all dynamic parameters that OWOX BI supports.
  • When importing cost data, OWOX BI checks the UTM tags in your campaigns and reports any errors. Examples of errors are discussed below.
  • OWOX BI converts imported data to the correct format. For example, Google Analytics has its own schema for storing data and uses parameters rather than UTM tags: for example, ga: source instead of the utm_source tag. OWOX converts data into the format used by the service to which it’s sending that data.
  • OWOX BI updates data uploaded to Google Analytics if it has changed in the advertising service. For example, if advertising service analyzes your traffic and determines that your ads were passed by bots, they’re likely to return money to your balance. OWOX BI tracks these things and keeps data in Google Analytics relevant.
  • If necessary, OWOX BI can upload your historical data to Google Analytics. With paid packages, you can upload data for the past six months. With free packages, you can upload data for the past two months. We’re now working on an updated version of OWOX BI that will allow you to download cost data for any past period if that data is available in the advertising service.
  • In addition, OWOX BI is able to collect raw data about Google Ads campaigns in Google BigQuery using auto-tagging, getting that data from the gclid and yclid parameters. The fact is that UTM tags are not used in the link when displaying AutoLabel ads. This isn’t a problem if you’re analyzing advertising costs only in Google Analytics. But if you try to upload cost data from the same Google Ads and combine it with session data, you have to attribute advertising costs not by tags but by gclid. OWOX knows how to do that.
  • When importing data with OWOX BI, the currency of the advertising service is converted into the currency of the Google Analytics property.

Algorithm for importing Google Analytics cost data using OWOX BI

  1. Using official advertising service APIs, OWOX BI receives ad display statistics.
  2. For each announcement, OWOX gets UTM tags. In most cases, they come as a link.
  3. From this link, OWOX extracts UTM tags if they contain dynamic parameters.
  4. OWOX BI analyzes these tags for errors and replaces dynamic parameters with their values.
  5. After getting all UTM tag values, OWOX forms a CSV file to upload to Google Analytics. In addition to tags, this file contains data on costs, screenings, clicks, and dates.
  6. Finally, OWOX BI uploads this file to Google Analytics.

Errors in tagging that OWOX BI defines:

  • No obligatory UTM tags.
  • Unsupported dynamic parameters are used that are not available in the advertising service API. For example, advertising service has a dynamic parameter that can be used to track the exact position of an ad in the search results. But in the API in the ad section, it’s impossible to get this data. In such cases, we recommend using Google Analytics user settings instead of dynamic settings.
  • Can’t parse the UTM tag. This can happen with some types of ads when advertising platforms don’t have the technical ability to define tagging — for example, smart banners.
  • Syntax errors in UTM tags.

Algorithm for importing cost data into BigQuery via OWOX BI

The next step in building business intelligence is to import cost data into Google BigQuery. Doing so is very similar to importing cost data into Google Analytics. The only difference is that whereas we transfer only information on UTM tags and costs to Google Analytics, we transfer additional parameters (more than 200 metrics for Facebook) to Google BigQuery.

This allows you to build detailed reports for in-depth analysis, create remarketing lists, effectively manage ad rates, and train your machine learning model for more accurate planning.

Collect user behavior data

OWOX BI has its own counter (something like the Google Analytics counter) that’s installed on the customer’s website. With its help, OWOX records data on user behavior in the form of hits in Google BigQuery. This data is available with a delay of just a couple of minutes. After 24 hours, tables are formed with user sessions from tables with data on hits.

We then work to fill in the two fields in these tables: attributedAdCost and AdCost.

  1. To begin, OWOX BI gets data from Google Analytics on ads with UTM tags. In most cases, this is just the cost data loaded using OWOX BI.
  2. OWOX BI forms a table of advertising costs from Google Analytics data.
  3. Using tags from the session tables and tags from the cost tables, OWOX defines the cost for each session.
  4. As a result, in the session streaming table, all costs are allocated to user sessions (the attributedAdCost field).
streaming table

This information helps you analyze data in different segments. The easiest thing is to calculate the cost per order (CPO) for each of your orders. Just take the attributedAdCost of the user who converted, sum up the costs if there were a few paid transitions, and get the cost of your order.

You can also group expenses and revenues by user, cohort, or landing page. This helps to evaluate the effectiveness of campaigns aimed at returning old users or attracting new ones as well as with planning the budget for different categories, regions, etc.

Read more about what tasks can be solved by knowing the cost of a session in our article on how to assess the effectiveness of product categories, client segments, and landing pages.

Mistakes in UTM tags

These are the most common errors in UTM tagging in our experience:

  • Not using tags at all. For example, you run ads on Facebook and have a Facebook company page with links to your site. If you don’t use UTM tagging, you’ll never understand whether there were transitions from your page or your advertising.
  • Not using all the required tags. For example, you specify a utm_source but don’t specify a utm_medium.
  • Using different registers to name a single tag in a single campaign. For example, CPC and cpc Google Analytics will count as two different campaigns.
  • Using anchors in links. An anchor is a lattice followed by a certain value. Anchors need to be put after UTM tags.
  • Making tag values too long (more than 8 KB). In such cases, tags will be trimmed.

Syntax errors in manual tagging:

  • Including gaps.
  • Reusing the “?” symbol.
  • Using “&” in UTM tag values. The ampersand separates one tag from another. If you use it in tag values, it will cause confusion.
  • Using brackets {} in UTM tag values. The OWOX BI algorithm takes values within brackets as unknown dynamic parameters.

Short conclusions

  1. UTM tagging is an indispensable practice for analyzing the efficiency of advertising campaigns.
  2. Design campaigns and UTM tagging based on business logic and tasks you want to solve. Determine for yourself which data segments you need to analyze. For example, if you want to analyze brand and non-brand traffic, it’s desirable to have different campaigns in advertising sources for this purpose.
  3. Use the potential of dynamic parameters to the fullest. In Google Analytics’s Cost Analysis report, you can easily filter your data using advanced options. For example, you can view costs by location, display networks of your contextual advertising, etc.

P. S. If you need help creating UTM tags and setting up business intelligence, we’re here for you.