What you need to know about marketing measurement in 2019

Accurately measuring marketing performance has always been one of the biggest pain points for digital marketers. From proving ROI to creating accurate attribution models and successfully tracking online to offline purchase paths, there is still a lot to be desired when it comes to digital marketing measurement and reporting.

When Facebook announced their new measurement tool Facebook Attribution that was definitely a big step in the right direction. Facebook Attribution came in response to a series of changes in the digital marketing landscape and has the potential of changing the way marketers approach social media marketing measurement.

In light of the promising advances in marketing measurement and the brand new Facebook Attribution, we have a look at the evolution of digital marketing measurement and what the future holds.

The evolution of digital marketing measurement

What were the events and key ideas that shaped the way we think about and measure digital marketing?

The No measurement era

In the early days of digital marketing, a company’s digital presence simply meant their desktop website. Marketers’ main concern was web analytics and the data that was being collected and reported on was very limited. For those of us old enough to remember it, the classic “analytics” product was a hit counter like StatCounter just showing the number of impressions by source. Back in those days digital marketing simply lacked the measurability that allows marketers to determine ROI and make data-driven strategic decisions.

The Last click attribution era

The last click attribution model was born in a desktop only world when Google, Yahoo, Microsoft search and website referrals were pretty much the only performance marketing channels in digital marketing. In those days marketers didn’t need to worry about tracking multiple touchpoints across different channels and devices and last click attribution served them well.

Last click attribution assigns a 100% of the credit for the sale to the last touchpoint a customer has before converting. For a long period of time, last click attribution dominated the way digital marketing is measured and a huge number of companies are still committed to last click attribution to this day. However, in today’s world, last click attribution is becoming more and more problematic.

The problem with this model is that today a person is likely to have multiple touchpoints with your business before they decide to make a purchase and each one of those touchpoints plays an essential part in propelling them down the sales funnel. Focusing solely on the bottom of the funnel activity and ignoring the top and middle of the funnel touchpoints puts you at a risk of making some regretful decisions about your marketing strategy.

Let’s say you are a software company providing accounting software for small and medium-sized businesses and you are looking to get more sign-ups for your software. This is what an example customer journey may look like:

1. Someone searches for the term “small business accounting” in Google and lands on your blog post on “How to manage your accounts as a small business”.
2. That person is then retargeted with a Facebook Lead Ad pointing them to your webinar on “How to save time and money with the right accounting software for small businesses”.  
3. The person watches the webinar and is entered into a lead nurturing email programme
4. The person receives a special offer email and signs up for your product on a special offer landing page on your website.

In this case, with a last click attribution model, the last special offer landing page that they converted on will get all the credit for the sale. This is an overly simplistic way of looking at your marketing performance and can often lead to relying on misleading data. If based on the last click data you decide to focus on creating more special offer email campaigns and reduce your Facebook advertising budget this can cause a domino effect in your marketing and leave you with email campaigns that nobody engages with.

The Multi-touch attribution era

Over the past 4 or 5 years, Google Search has lost its once monopolistic dominance of performance marketing as new channels such as social media platforms take bigger and bigger shares of the market. With more marketing channels, more devices and more demand for information, consumers are more likely to have multiple touch points before making a purchase. This adds to digital marketing attribution complexity with many more trackable touch points along the customer journey, across devices. In response to this measurement challenge, marketers turned to multi-touch attribution.

The multi-touch attribution model is the idea that each of the touchpoints leading up to a conversion plays a role in the customer journey and should be assigned a certain value. Multi-touch attribution was made possible through web cookies and sessions. Cookies allowed marketers to track users’ actions across web pages, remember information about them and use that information for ad targeting through third-party cookies.

Session cookies made it possible for marketers to remember a user’s actions in a single website visit giving them more engagement context. Sessions rather than single visits became the standard in audience analytics. Google Analytics uses Sessions as its default measure and defines a session as a group of interactions within the time frame of 30 minutes.

The GDPR era

When the General Data Protection Regulation (GDPR) came in force in May 2018, it presented to businesses marketing to EU citizens the biggest change in data protection laws in 20 years. Uncertainty around the types of information that marketers were allowed to collect and the methods of collecting it added another spoke in the digital marketing measurement wheel.

Discussions around the use of cookies and their implications on privacy were particularly concerning for digital marketers in the wake of GDPR. All web browser cookies capable of identifying individual users are subject to GDPR regulations. The new GDPR regulations require businesses to not only clearly communicate the use of cookies on websites but to ask for the user’s positive and explicit consent to use them.

Since the time of the earliest web browsers, cookies have been an integral part of making our browsing experience better and more convenient allowing us to log into websites and save our preferences. When the first cookies were used in the early 1990s to track returning visitors to the Netscape web browser, Internet users were blissfully unaware of their existence. Cookies came to the public’s attention in 1996 when media started reporting on the potential threat to privacy. Concerns today are mainly around third-party cookies which are used for ad targeting by third-party networks.

In line with user privacy protection, Apple blocks third-party cookies on its web browser Safari. This has a huge impact on measurement as a marketer with over 50% of third-party cookies now being blocked. In a mad arms race, Google, Facebook and other providers use first-party cookie based setups (meaning your cookies are from your own domain), to get around the blocks.

Quick tip: Make sure you’re on the latest versions of all your website analytics scripts to ensure you are measuring everyone who visits.

The People based measurement era

In a world where consumers spend more time on social media and mobile devices, the ability to track them as they move across a range of devices and touchpoints became mission critical for digital marketers. In this cross-device world, marketers could no longer rely on web cookies for effective tracking and targeting. Many argue that cookies are crumbling as a user targeting method as mobile users spend most of their time in apps rather than browsers and two-thirds of mobile devices don’t accept cookies.

The solution of the cross-device problem came from platforms like Facebook and Google in the form of identity resolution or what is known as people based measurement. Google and social media platforms like Facebook had an inherent advantage - the logged in user. Unlike cookies, people-based marketing tracks real people across devices through registered user data. This allowed marketers to reach real people across web browsers, mobile apps or locations and track offline transactions.

The future of digital marketing measurement

We have definitely come a long way since the early days of the Internet but digital marketing measurement still has a long way to go if it is to thrive in a constantly evolving, multi-channel world. So, what does the digital marketing measurement of the future look like?

Marketing Measurement looks to be evolving into a combination of many of the techniques we’ve described:

Multi-touch

Unlike last click attribution that only focuses on the bottom of the funnel and first click attribution that only looks at the top of the funnel, multi-touch attribution models account for multiple touchpoints and draw a fuller picture of your customer journey. Multi-touch attribution assigns multiple touchpoints a specific value. Here are the most common multi-touch attribution models:  

U-shaped - U-shaped attribution model assigns 40% of the credits to first and last touch and distributes the remaining 20% equally to the rest of the touchpoints in between.

Linear - a linear attribution model assigns each touchpoint an equal percentage of the credits.

Time Decay - this attribution model assigns the majority of the credits to the last touch and assigns diminishing values to each preceding touchpoint.

If we go back to our software company example, the way a U-shaped multi-touch model would look at the same customer journey is as follows:

1. Someone searches for the term “small business accounting” in Google and lands on your blog post on “How to manage your accounts as a small business”.
2. That person is then retargeted with a Facebook Lead Ad pointing them to your webinar on “How to save time and money with the right accounting software for small businesses”.  
3. The person watches the webinar and is entered in a lead nurturing email programme
4. The person receives a special offer email and signs up for your product on a special offer landing page on your website.

With a U-shaped multi-touch attribution model, you might want to assign the blog post that initially generated the lead 40% of the credits and the special promotion landing page that converted them into a paying customer another 40%. The Facebook Lead Ad and the webinar will then split the remaining credits and each get 10% of the credit for the sale. This way you allocate enough attention and resources to the essential top of the funnel activities that generate leads and the bottom of the funnel conversion points without ignoring the role that the other important touchpoints in between play in nurturing and qualifying the lead.

Companies like Fospha take this to the next level. Rather than picking a human-based model, Fospha uses historical data to analyse what touchpoints influence a transaction, and weights each touchpoint accordingly, if it occurs in future actions.

Apply Machine Learning models to this concept, and you have an ever-evolving picture of how marketing influences your sales. As marketers, we’re going to need to give up the control freak in us, and trust in the machines to help us make better decisions on where to put our money!

Cross-device using People Based Measurement

Mobile devices are playing a bigger role in people’s lives and in the way they browse, research services and products and shop. While mobile shopping is affecting more and more industries, research points that people prefer to transact on desktop devices especially when it comes to high ticket items. For example, someone looking to buy a new car might be more than happy to browse car ads and search local dealerships on their mobile phone, they would likely prefer to view car configuration pages and book a test drive on their desktop. It is becoming increasingly important for marketers to adapt to those natural consumer behaviours and account for a fragmented, cross-device customer journey.  

Until recently, web tracking tools tracked devices rather than people. Earlier in 2018 Google Analytics introduced its Cross Device tracking capabilities. Google’s Cross Device tracking gives you access to reports such as Device Overlap, Device Paths and Channels. In order for Cross device tracking to work you need to active Google signals in your Google Analytics settings.

Follow the steps below to get started with Cross device tracking with Google Analytics

1. Login to Google Analytics
2. Click Audiences
3. Click "Cross Device"
4. Follow the wizard.

We can expect companies like Facebook to follow suit and use their people graphs to help map the customer journey across devices.

Quick tip: Downloading Facebook Ads cross device reporting makes for fascinating viewing. One client of ours (a luxury homeware brand) showed 80% of their sales started on mobile and purchased on desktop. This highlights how critical it is to measure the customer across devices.

Online to offline

The online and offline worlds are blending into one and the customer journey is far from linear.

More and more industries are being affected by the digital world and more and more of the touchpoints are happening online for industries that traditionally rely on offline sales. 90% of retail sales happen offline, and 10% online (Deloitte digital divide), however in reality 60% of offline purchases are influenced by digital.

Someone looking for a new coat, for example, is likely to browse an online retailers’ website and mobile app but is then likely to go into a local store in order to try on and purchase the item they have chosen. This behaviour is more complicated and fragmented but thanks to technological advances it’s now easier to track and measure. The relationship between online browsing and offline conversions is becoming increasingly important for businesses to track.

Marketers have been able to track online marketing to offline sales through initiatives such as online coupon campaigns, click and collect and call tracking. In 2014 Google’s conversion tracking in AdWords introduced estimated store visits. The data Google is able to provide is only an estimate and relies on data from smartphone users who are signed in to Google and have location history enabled.

Both Google and Facebook now have Offline Conversion APIs, which allow us to send back offline sales data (email addresses, phone numbers etc), which are mapped to people who have seen the online ads.

This isn’t easy to set up, so solutions like Driftrock’s online to offline conversion sync make tracking individual ads or targeting ads through to sales possible by connecting customer events like an offline purchase, lead or CRM sale and pushing these events to the marketing and analytics platforms for measurement.  

Facebook Attribution and the future of social media marketing measurement

Social media was traditionally considered a brand awareness channel that contributes to top of the funnel metrics such as impressions, views and visits. Social media marketing traditionally didn’t have to deal with conversion metrics.

All that changed when Facebook’s algorithm started favouring organic content from friends and family over content from business pages. This made it increasingly difficult for brands to drive ROI with organic social media promotion and increasingly necessary for them to invest in social media advertising. Social media spend has grown more in 2018 than any prior year and is expected to expand by 66% in the next five years.

With more dollars and pounds spent on social media advertising, marketers expect to see more easily measurable outcomes. When looking at social media marketing performance, marketers today look at the cause of conversion and want answers to questions such as “Where do leads come from?” “What devices do the leads use?” “Which devices are leads more likely to convert on?”

Facebook Attribution pulls together the conversion performance for your ads on Facebook's family of apps, as well as across different platforms that you use in your marketing such as Google Ads, LinkedIn and Twitter (the list is long with 37 available platforms at the moment). This allows marketers to monitor conversions across multiple channels and devices and answer important questions such as - “What is the ultimate channel and device mix for a customer that is more likely to convert?”. For a detailed introduction to Facebook Attribution check out this awesome blog.

As well as allowing for monitoring conversion performance across a number of channels Facebook’s Attribution tool moves away from Ads Manager’s last click attribution and gives you the chance to create an attribution model that better fits your business. Facebook Attribution comes with 7 different attribution models:

• Last click or visit - this attribution model assigns 100% of the credits to the last click or visit in the conversion path
• Last touch - just like the Last click or visit attribution model,Last touch gives a 100% of the credits to the last click or visit in the conversion path. The only difference is that if there was no click or visit, Facebook will credit the last impression
• Positional 30% - if you decide to use this attribution model, Facebook will assign 30% of the credits to the first touch and 30% to the last touch and will distribute the remaining 40% equally across the rest of the touch points
• Positional 40% - this attribution model assigns 40% to the first and last touch and distributed the remaining 20% equally across the rest of the touch points
• Time decay 1-day - this attribution model gives touchpoints an increasing percentage of the credit for the conversion as they get closer in time to the conversion
• Time decay 7-day -  this attribution model gives touchpoints an increasing percentage of the credit for the conversion as they get closer in time to the conversion

Check out Facebook’s guide on Attribution for more information on how to choose the best attribution model for your business.

In a constantly changing digital landscape, it is increasingly important for digital marketers to have a single reliable source of information that allows them to accurately attribute return on investment and properly distribute their budgets. Embracing the future of digital marketing measurement would mean marketers are moving closer to a world of 100% measurable outcomes.