Digital analytics: How to maximize its value for your company

Ian Scheel

More than a decade of experience collecting high-quality data for midsize and large companies across North America and Europe, founder of More by Ian Scheel

Tons of new regulations and privacy features in browsers and mobile apps have made digital analytics increasingly difficult. Additionally, stakeholders have become more demanding, and automation and AI require a new level of data quality. So here is what you need to do:

Achieve your objectives

Enable use cases

  • Digital Analyst
  • Measurement Specialist
  • Head of Digital Analytics
  • (show more)
  • Marketing analytics
  • Customer analytics
  • Product analytics
  • (show more)

Avoid data issues

Avoid data compromises

  • Missing data
  • Breaking data
  • Contradictory data
  • (show more)
  • Measure everything
  • Measure cross-device
  • Measure without consent
  • (show more)

Avoid tool issues

Mitigate costs and risks

  • Basic vs advanced features
  • Mismatching numbers
  • Legal non-compliance
  • (show more)
  • Data collection costs
  • Downstream data costs
  • Legal compliance
  • (show more)
  • Google Analytics®
  • Adobe Analytics®
  • Segment®
  • (show more)
  • Snowplow Analytics®
  • Mixpanel®
  • Rudderstack®
  • (show more)

Focus on your objectives and use cases, avoid issues by feeding our data into your analytics tools.

Digital analytics: Achieve your objectives

In order to maximize the value of digital analytics for your company, you need to start with your personal objectives to deliver tangible results:

  • Reports and dashboards that don’t break.
  • Insights and analyses that make sense.
  • Results that your stakeholders trust.
  • Numbers that decision makers can rely on.
  • Audiences that marketing teams can work with.
  • A stable digital analytics tool that nobody questions.
  • Numbers matching across different tools.
  • Holistic analytics data that covers any use case.

Achieving these objectives is easier said than done. Many struggle with it but the reason is not what you may think; most have the skills and they put in the effort, but they are usually limited by issues with the underlying data:

Digital Analyst / Digital Data Analyst

When your data has blank spots, you are not able to conduct your analyses. When your data is not reliable, you are not able to answer questions from stakeholders with confidence. When your data is incorrect, you risk providing bad advice and leading decision makers in the wrong direction.

Measurement / Implementation Specialist

When the data measurement architecture is not robust, you are constantly putting out fires instead of making necessary upgrades to your implementation to reflect the latest technological developments in browsers and mobile operating systems. The pressure is on you to keep the data collection infrastructure up to date and keep up with the latest technological developments. Also, when data points stop populating or contain incorrect data, nobody is going to blame the developers, this is your responsibility.

Director / Head of Digital Analytics

Other departments rely on your team’s data. If that data isn’t sound, your team’s and your purpose come into question. When your team is primarily fighting data issues instead of providing valuable insights, even in the most lenient organizations, you are going to be held accountable eventually. As the person carrying the overall responsibility, you need to lead the team in the right direction. From a technical perspective, this includes using a modern and forward-thinking a modern approach to data measurement. It’s usually not enough to apply outdated concepts in an ever-evolving environment.

Next steps

Regardless of your specific role, you need to get your data right. Based on more than a decade of experience, we teach you how to achieve quality data in our articles. If you need results quickly, we can take care of everything and stream data of guaranteed quality into your analytics tools of choice (click here for a list).

Digital analytics: Enable use cases

In order to maximize the value of digital analytics for your company, you must make sure to support the many use cases that depend on digital analytics data:

  • Reports and dashboards
  • Marketing analytics
  • User behavior analytics
  • Customer analytics
  • User experience analytics
  • Product analytics
  • Customer retention
  • Content personalization
  • Conversion rate optimization

Implementing the above use cases successfully requires high-quality data. If your data is not good, you are going to run into a lot of issues, for example:

If data points are missing

Your analytics abilities will be limited, i.e. you may not be able to answer certain questions, you can make wrong decisions, e.g. where to allocate your marketing budget or which new product feature to prioritize, or you come to wrong conclusions because of blank spots. Examples:

  • No behavior and/or purchase data when user opted out of tracking
  • Certain interactions with website elements not covered
  • ID not captured when a user logged in
  • Parts of website without tracking
  • A/B test or personalization variant not provided
  • New feature not covered by tracking implementation

If data is unreliable

Your insights will be partially or completely wrong. Conclusions derived from unreliable data are likely just as unreliable. And that can make you look unreliable in the eyes of stakeholders and peers. And this can get really bad if use cases of others fail due to your data. Examples:

  • Some button clicks included, some excluded from data
  • Only some or no Single Page Application pages measured
  • Wrong purchase price when voucher was applied
  • Return visits not recognized due to poor cookie setup
  • Only some marketing campaigns and channels are measured

If data is breaking

Even if your data is complete and usually reliable, all hell can break loose, if the data that worked well suddenly breaks. Events like that typically break dashboards and reports, which can have management attention or that of other teams, or even worse, entire use cases can come to a screeching halt. Examples:

  • Purchase information stops populating
  • Page title changes suddenly due to content changes
  • New CSP header blocking tracking requests
  • New product build or deployment stops entire tracking

Next steps

Regardless of your specific use case, you need to get your data right. Based on more than a decade of experience, we teach you how to achieve quality data in our articles. If you need results quickly, we can take care of everything and stream data of guaranteed quality into your analytics tools of choice (click here for a list).

Digital analytics: Avoid data issues

The capabilities of data tools, use cases and consumers directly linked to the quality of their data. If you’re experiencing any of the following issues, those capabilities will be severely limited and can never reach their full potential:

  • Data that doesn’t make sense and contradicts other data sources or itself.
  • Data values that abruptly change.
  • Data points that suddenly stop populating.
  • Data missing for certain features you want to use in your tools.
  • No adjustments to constant changes in browsers and mobile operating systems.
  • No consent-awareness leading to legal non-compliance.

Your digital analytics tools and everything built upon its data downstream don’t work well if you have any of the above data issues. Here are the basics you need to get right:

Task the right people

Even though data issues at the source can have huge ripple effects downstream, most companies underestimate how important it is to have knowledgeable experts collecting user event data for the entire organization. Additionally, if data collection isn’t prioritized the way it should be, there are often budget constraints which further worsen the situation. The truth is that data collection requires knowledge in a multitude of fields due to it being at the intersection of business, data architecture, software development, technology, and corporate horse-trading. The “perfect candidate” needs to excel in all aspects in order to be successful.

Invest in a custom implementation

When vendors list features and abilities of their tools, they usually don’t point out that in order to use their solution to its fullest extent, a custom implementation is required. Even worse, when asked about implementation efforts, most sales people will make it sound super easy. The truth is that most tools can be implemented very easily, but basic implementations only enable basic features. What most fail to tell you is that for all the great features, an advanced implementation is required. Additionally, when your application spans multiple domains or you have a Single Page Application using frameworks like React, Angular, Vue, etc., even basic implementations don’t work out of the box, not to mention that most tools are not managing user consent automatically, so the odds that your basic implementation is not legally compliant is very high.

Choose a modern measurement architecture

There is no official definition what is considered modern, but based on our more than a decade of experience, we recommend aiming for the following characteristics:

  • Tool-agnostic: Stop creating data for one particular solution, create generic data that can be used by many tools.
  • Generic data model: Don’t adjust your data to a downstream consumer or use case when you create it, keep it generic and reusable.
  • Embedded / deeply integrated: Make sure that extensions and new features of website, web applications, and mobile application, are automatically covered.

Build close relations with your developers

Because data collection happens at the intersection of business and technology, good cross-functional working relationships are essential. For best results, you should try to establish a reciprocal understanding of the other party’s goals and objectives. This requires personal investment, mutual respect, and will make everyone’s lives easier. From the business side, you especially need to understand software engineering best practices and limitations to make sure you don’t become a burden or are perceived as one. In order to do so, you have to have basic understanding of all the technologies involved and how programming and application development work. For example, you need to understand what’s feasible in their application in terms of tracking. Otherwise you won’t be able to find the sweet spot between robust data measurement and endless back and forth with developers.

Next steps

As you can see, you need to make sure your data is issue-free. Based on more than a decade of experience, we teach you how to achieve quality data in our articles. If you need results quickly, we can take care of everything and stream data of guaranteed quality into your analytics tools of choice (click here for a list).

Digital analytics: Avoid data compromises

A lot of companies make unnecessary compromises when it comes to their data foundation. Here is what leads to issues all the time:

  • Basic measurement producing only basic data
  • Incomplete user journey with lots of missing data points
  • No cross-device measurement with lots of imaginary users
  • No “unconsented” measurement with lots of missing data
  • Unnecessarily omitting data due to legal concerns
  • Not utilizing advanced tool features due to expertise and implementation effort required

Downstream issues are preventable if your measurement strategy includes the following:

Measure everything

Many companies only measure basic information. There are various reasons for that but they usually don’t make a lot of sense:

  • Limited developer resources: Picking a few data points to measure is usually not less work for developers than a generic approach.
  • Budget constraints: Doing an implementation step by step is usually more expensive than doing it all at once, especially using a generic approach.
  • Lack of time: Ambitious deadlines often result in time constraints for measurement so that only a basic implementation can be done, and the rest is never added later as initially planned.
  • Avoidance: Because coordinating measurement involves a lot of people and special knowledge, teams try to do “too much” of it.
  • Many tools sharing resources leading to only basic implementations creating data silos instead of one that covers everything.

All of these things lead to incomplete and low-quality data that your entire organization will suffer from downstream. We covered this in a section above. That’s why you should always try to track everything, including offline touch points, so that you have data on the full customer journey. With the right approach, you can work your way around the above constraints.

Measure without user consent

No, we don’t suggest breaking the law! You can measure everything and still be compliant with the most strict privacy laws (currently in Europe), it’s just a bit more complex from a technical point of view. Basic implementations, however, won’t be able to collect data of users that haven’t consented, at least not legally. Instead, what you need to do is remove any PII (personally identifiable information) and add as many attributes to include anonymized users in cohorts for extrapolation and approximation based on the portion of the user data that was collected with consent and all PII attached.

Measure cross-device / login-based

If you want to track users cross-device, for example based on a cross-device identifier like a login / user ID, this has to be implemented manually. We are not aware of tools that pick up cross-device IDs automatically on web, mobile, and other devices.

Next steps

As you can see, you can’t make too many compromises in data. Based on more than a decade of experience, we teach you how to achieve quality data in our articles. If you need results quickly, we can take care of everything and stream data of guaranteed quality into your analytics tools of choice (click here for a list).

Digital analytics: Avoid tool issues

Most tools don’t work out of the box to their fullest extent, neither do they work well when the data they use is of low quality.

Basic vs advanced features

Tools that measure user behavior, for example Google Analytics, Adobe Analytics, Snowplow Analytics, Mixpanel, Segment, Rudderstack, Matomo / Piwik, are easy to implement. But that’s only true for very basic implementations that enable only very basic features. All these tools only work to their fullest extent when you use their advanced features and put in the effort to feed them very specific data, for example:

  • Extensive product and user meta information sent with a purchase or other conversion
  • User IDs to enable user-centric and cross-device tracking
  • Consent-status to ensure legal compliance
  • Anonymized, pseudonymized, and cohort-based data so that at least some data can be collected legally without consent, see below

Mismatching numbers

A lot of tools collect similar data, and often times, the numbers don’t match. You may hear excuses like “oh, Google Analytics and Adobe Analytics count differently”, or “well, they exclude different traffic as bot traffic”, or “the technologies are different”, but that’s all complete BS.

Your numbers don’t match because your tools are not implemented correctly, meaning not similar enough to produce matching numbers. The different tools are much more similar than you may think:

  • Typically define metrics based on industry standards, e.g. of page views, and ad views
  • Usually the same 3rd-party lists of IPs and user-agents to detect bots
  • Exact same set of technologies, no secret sauce
  • Similar data processing, e.g. to remove duplicates

Legal non-compliance

In many parts of the world, you can’t just collect any data without the user’s consent. That makes most basic implementations unlawful because in order for them to be consent-aware, they need to be made aware of the user’s consent which usually doesn’t make it into the tool automatically. In the worst case scenario, if you can’t distinguish between “consented” and “unconsented” data, you may have to stop using the tool or delete all the data in it in order to not risk fines.

Next steps

If you want to be able to do your job, you need to make sure that your tools don’t get in the way, either through high-quality implementations that are aligned across your tools or by feeding all of them the same data. Based on more than a decade of experience, we teach you how to achieve quality data in our articles. If you need results quickly, we can take care of everything and stream data of guaranteed quality into your analytics tools of choice (click here for a list).

Digital analytics: Mitigate costs and risks

Unfortunately, a lot of companies lose money with their data initiatives. There are various reasons for that:

Shiny objects instead of foundations

A common mistake is to jump into ambitious data projects without creating the necessary data foundation first. It’s a myth that data can be fixed downstream. There are very few things you can do about data that is created incorrectly at the source. “We fix it later” and not spending much time on the foundation are approaches that have huge downstream ripple effects. The consequence: Snowballing downstream costs sink the project even if everyone involved did their very best. Though not free, it is much cheaper to fix data at the source.

Costs for initial data foundation

You definitely need a budget for the initial data collection setup. We recommend going for quality over cost because any money you think you saved on foundation, you’ll pay for exponentially downstream.

Costs for updates and maintenance

Here is what a lot of companies forget: After the initial setup is done, data collection pipeline requires constant maintenance and up-keeping. Websites, mobile apps, products, browser and mobile app technologies, legal requirements, everything is constantly evolving and data collection needs to be adjusted accordingly.

Legal risks and compliance

Legal compliance is constantly getting more complicated, regulations are becoming more strict and fines are increasing. But there are also positive aspects to this: Legal standards are converging on a global level, which can decrease the efforts required to get your data collection “up to code”, if you approach it globally.

USACaliforniaCCPA (California)
EU / EEAAustria, Belgium, Bulgaria, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, SwedenGDPR (region-wide)Up to 2% of global revenue
CanadaAlberta, British Columbia, Manitoba, New Brunswick, Newfoundland and Labrador, Northwest Territories, Nova Scotia, Nunavut, Ontario, Prince Edward Island, Quebec, Saskatchewan, YukonPIPEDA (nationwide), Law 25 (Quebec)

Next steps

We can provide the data foundation you need and we can also assure its legal compliance. Based on more than a decade of experience, we teach you how to handle costs and compliance in our articles. If you need results quickly, we can take care of everything and stream data of guaranteed quality into your analytics tools of choice (click here for a list).

Digital analytics: Data as a Service

With you never have to worry about accuracy of your analytics tools because you’ll be using our high quality data. That’s right, analytics tools don’t need to collect their own data, they can be fed third-party data with no change in functionality.

  • Segment®
  • Rudderstack®
  • Matomo® / Piwik®
  • PostHog®
  • Google Analytics®
  • Adobe Analytics®
  • Snowplow Analytics®
  • Mixpanel®

So no matter your role or use case, we’ll provide you with a hands-off, low-risk, solid foundation without issues or compromises. Our Data as a Service supports the most popular digital analytics tools. If you don’t see your tool, please reach out so that we can build an integration.

Jump-start your data initiatives

Start working on your objectives right away to fast-track achieving your goals and creating the desired business value in record-time while minimizing risk and uncertainty.

With the click of a button

If you want to confidently rely on the data within your tools, connect them to our Data as a Service with the click of a button. This enables you to focus on what matters: consistent reporting and actionable insights.

We take care of the rest

  • We collect the data, and maintain it.
  • We stream it into your analytics tools for web and mobile.
  • We continuously ensure and guarantee its quality.
  • We constantly make technical improvements.

Get started today

Stream our data into your analytics tools of choice. Please reach out if you don’t find a tool you would like to use on our list.