Analytics and Measurement for Digital Marketing

November 1, 2024|Kim's Blog, Marketing, Statistics and trends, Web Sites|

Chloe Christine of Legal Marketing UK – TBD Marketing Agency for Lawyers provided a comprehensive and practical workshop for PM Forum’s digital marketing course “Analytics and Measurement for Digital Marketing”. Chloe runs analytics teams across a number of agencies, advises law firms on analytics and is passionate about the subject. The six sections looked at GA4 in detail, helping delegates find and interpret the data. There were practical sessions to help delegates set up their systems and reports. Chloe provided template Looker Studio reports with the resources.

Delegates included marketing executives and managers from legal, actuarial, financial services and pensions firms. Most had a few years’ marketing experience. Some were digital marketing experts and others were in general marketing and business development roles. Details of future PM Forum marketing training sessions – including those by Chloe: PM Forum – PM Forum

1. Analytics and measurement in perspective

A poll revealed that 56% considered themselves beginners and 44% intermediate, so this section laid the foundations. Everyone was using GA4 rather than Facebook Analytics, Adobe or HubSpot.

There was a brief history of web analytics: Page views and hit counters (1990s), session-based tracking (2000), user-centric metrics (2010) and then Google Analytics (from 2020). Whilst GA4 launched in 2020, its use wasn’t obligatory until 2023.

A key difference between former analytics and GA4 is that it is event based (rather than session based). It focuses on engagement (either 10 seconds or complete an action) rather than bounce rate.

Data retention defaults at two months, it offers predictive metrics through machine learning and adheres to privacy rules. The reporting interface is different too.

GA4 still uses cookies but has more advanced features – it is better at tracking users across different devices (desktop and mobile).

Chloe explained consent. This is where users are asked to accept, modify or refuse cookies for analytics tracking. You must obtain analytics consent before firing the Google Analytics tag as it is denied by default. Other categories of consent cover: personalization, functional and security.

She explained that statistics will be under-reported by between 5% and 30% as not all users will accept cookies. And she demonstrated how to check your default consent state. Whilst it is the job of the web developer to set this up, marketers need to know that analytics is tracking properly.

There was a quick look at server-side tracking which is a method of sending data from your website or app to your analytics platform using a server as an intermediary rather from the user’s browser or device. This ensures accurate data.

2. Key terminology and core reports

We considered the key metrics in GA4: sessions, users, engagement rate, conversions and bounce rate. The system changed conversions to key events about six months ago.

Then we took a quick look at the built-in or pre-set reports available including acquisition (where users come from – organic search, paid search, paid social, referral e.g. LinkedIn etc). Chloe demonstrated how to add options to customise your reports. We also saw how to obtain information on what key words your web site ranks for and how to add in Google Ads if you use them.

There was a breakout discussion where delegates discussed the metrics that were most important for their firms. Feedback included: page views for a particular page or (blog) article, where user journeys started and ended and landing pages from MailChimp and other newsletter campaigns.

3. Finding the data

Delegates were equally split between finding and analysing data themselves and requesting it from someone else.

Chloe explained why it is important to wait for 3-5 days after the reporting period to obtain the most accurate data. She shared some other Pro Tips.

Then we looked at customising reports and delegates indicated that their most common needs were: landing pages from specific channels, blog data, social media and email marketing data. Chloe demonstrated how to use social media to create a report on social media including session source medium (total users, new sessions, engaged sessions, views per session, event count) where you can have up to 12 items for the last 28 days. And applying filters to just see organic social traffic.

She showed how to edit report lists to add customised reports. And to arrange to automatically send a PDF report to other people who have access to GA4. Delegates asked about creating reports for the Passle platform Passle for content creation and distribution – Grow your experts (kimtasso.com) which Chloe advised to do through the Looker Studio (google.com) and demonstrated the creation of a heat map.

She continued by showing reports of domains (referrer sites) that sent traffic to your site which also helps monitor backlinks. Noting that you should exclude Microsoft Teams statistics which is usually internal traffic. There were also suggestions for separate reports from AI sites such as ChatGPT, Perplexity and Gemini.

She showed how to look at trends, patterns and peaks in past traffic, track for primary and secondary actions and creating custom events using Google Tag Manager (GTM). She shared a video which showed how to do this.

4. Understanding your data

A delegate poll revealed that the biggest data challenges were: finding the right data (30%), understanding the data (10%), presenting data/extracting insight (40%) and advanced reporting/analysis (20%).

She explained that a rise isn’t always a good thing (e.g. a spike in traffic but low engagement would suggest visitors aren’t finding what they need). Professional service firms must focus on engagement and lead generation metrics not just traffic numbers – to ensure visitors are finding valuable information and can easily take the next step to becoming clients.

Segmentation considered breaking data into smaller groups such as new verses returning users, traffic by source and users in different geographic locations. She mentioned trends such as mobile users struggling to complete purchases.

There was a discussion of attribution modelling to help you understand which channels contribute most to conversions. Chloe talked through the main models: first click, last click, linear, time decay and position based (U-shaped 40% first, 20% middle, 40% last). Noting that GA4 defaults to data-driven attribution model – where machine learning is used to determine how credit for conversions is assigned. There was further discussion of: dynamic attribution, pattern analysis, credit allocation and personalised insights. And Chloe showed how you can determine what attribution modelling you are using.

5. Presenting insights

Delegates were asked how often they used dashboards to present data. 43% said never, 38% said regularly and 13% said always.

We returned to scheduling reports in GA4 where PDFs can be sent automatically when you choose dates and frequency.

Chloe shared her “Let’s make GA4 simple” template to assist with reporting where you can alter: dates, channels, countries and campaigns. Although she urged everyone to get used to using GA4. There was a walk-though creating a simple visualisation using Looker Studio.

Chloe showed how to connect your Analytics account and create reports. And to achieve dynamic chart updates to visualise patterns and trends over time. She demonstrated filters, pie charts, pivot tables and heat maps so you could see, for example, blog posts with declining interest.

6. Advanced GA4 reporting

Delegates were asked about which advanced reporting features they were most keen to learn about: cross channel tracking scored 89% and funnels 11% (no votes for segmentation or attribution).

She then talked through how to set up GA4 alerts in the analysis hub for when certain metrics see significant changes. For example, for particular percentage increases or decreases or for specific campaigns.

She also recommended more than one analytics in your web site to act as a backup.

There was a final break out discussion looking at scenarios for helpful alerts. Ideas included campaign traffic, seasonal content like budgets and weekly or monthly drop offs to prompt new content.

She advised to watch for bot traffic spikes which, at start of 2024, affected about 50% of web sites. She suggested looking for large spikes on certain days/hours from a direct source or unusual locations. And suggested speaking to web developers about preventing spam attacks. With a final mention of UTM (Urchin Tracking Modules) for cross-channel tracking.

Final word

It was interesting for me to reflect back on the session Chloe presented – in conjunction with Simon Marshall – back in September 2023.

My favourite quote then was “GA4 is not inherently bad but it is fundamentally different”. At that session we spent a fair amount of time on the Google Search Console. And there were some insights into using AI – primarily for creative ideas and sourcing of FAQs.

At that session we also looked at maximising PPC and paid social budgets although delegates didn’t want to focus on those at this session. I do remember warnings about vanity statistics at that session. And some neat ideas for monitoring competitor activity.