MeasureCamp Helsinki

This morning, I woke up to see snow falling gently, checking whether my phone picked up that we switched to summer daylight saving time and not miss my flight back home in the afternoon. But there is nothing summery about my present locale; it does not even feel like spring. Snow is falling outside because I am in Helsinki, Finland. Yesterday, our Finnish friends hosted a unique MeasureCamp event. For more on MeasureCamp, please read my other article. MeasureCamp Helsinki was a physical event, not a virtual one. There were over 100 attendees—the first of many surprises that Italy sent the largest foreign delegation. Also, several came from Austria and Germany. I found a few fellow French speakers as well. I believe I was the only one from the UK.

Over a breakfast of Karelian pies, we pondered how the Finns, who are somewhat introverted, might hesitate to give a session, i.e. a presentation, or lead a discussion. Peter O'Neill O'Neil, the founder of MeasureCamp, was even happy to let them run a session in Finnish as long the presenter would indicate clearly that the session would not be in English. After asking how many had planned to run a session, about a dozen hands went up, which confirmed our breakfast consensus. Not for long. As the session board opened, you could have thought you were at MeasureCamp London. The board filled up in 5 minutes flat. I found, to my horror, that the 30-min slot I picked would make me compete with Simo Ahava , Mikko Piippo , and Adam Greco , with no chance to move it until later in the afternoon. Adam managed to move his session, me not so much. But I had a second session on another topic later.

I started with Steen Rasmussen 's session on the rise of Decision Science. Steen revealed that the worldwide investment in Decision Science is half a trillion USD. Perhaps Digital Analytics should focus on helping decision-making and ride on the coattails of that investment. We are facing significant but not insurmountable challenges. First, we may have too much data. Choosing what to track exposes you to not having the data you need when you suddenly realise you need it. The clients and stakeholders would instead track everything, but this exposes you to too much data, most of which is useless, and exposes you to the risk of finding spurious correlations and false causations. It also makes data cherry-picking possible, not exactly what we want or need.

Companies face infamous cultural resistance to data, especially those that challenge their opinions and beliefs. But as Peter Jackson says, an influential CDO in the UK and former teacher of mine, data is like fish in a restaurant. Hoarding fish in the fridge will see it to spoil, spoil everything else in the refrigerator, and you have to throw it all away. Data and the actionable insight the analyst extracted from it require fast action. Between the time a client or stakeholder identifies an issue, and the moment the problem has a solution in place, the solution's business value has degraded over time, and sometimes quite precipitously. The analysts need time to identify how to track the issue, start collecting data and have enough of it before analysing it all. If the stakeholder or client is dragging their feet, the longer they wait, the more business value that actionable insight loses.

We have no flexibility on the time required to identify the tracking needs, the amount of necessary data without risking a lack of statistical relevance, or the time needed to analyse and extract the actionable insight. The only flexibility lies in how long the stakeholders and clients commit to making the changes. Steen proposed using Data Analysis as a navigation tool. When you begin with the end in mind, i.e. the desired outcome, you must focus on how your client or stakeholder will need to decide to implement your recommendations. Data Analysis will help keep track of the direction and the pace of change. Rather than letting an issue drop into insignificance, we should seize it while the opportunity to prove the business value in helping businesses make data-informed decisions is still hot. We need to start connecting the raw data to decisions.

My second session was with Adam Greco, a veteran of our industry who has recently transferred from the US to the Netherlands after joining Amplitude. Adam, an early Omniture employee, published the book on Adobe Analytics, which is still a reference for many of us. This book is out of print in paper form and commands a respectable amount as a second-hand copy. Mine is full of post-its, and a few years ago, Adam was my teacher for the day on his Adobe Analytics Top Gun course. I could not resist and asked him to sign my copy of his book. Adam and I have bumped into each other at various events, such as the first MeasureCamp in Brussels and the Adobe EMEA Summit. Adam's career spanned vendor, consultancy, and client-side, and he ran a session on the pros and cons of these three career options.

1. Client-side can be frustrating to some people because, as a Digital Analytics practitioner, you must use the tools they have chosen. Switching to a competing tool is difficult. There's also a lot of office politics and a lack of career progression. But you will gain deep domain knowledge about the industry in which your employer operates.
2. Agency/consultancy side will expose you to many clients, industries and tools. But going wide may come at the expense of mastery and gaining only superficial knowledge in each. Think of the jack of all trades, master of none situation. There is also a lot of travelling and client-facing, which may not be for everybody. If you have an affinity and the opportunity to focus on the technical side, then you might not need to travel and meet clients so much.
3. Vendor-side allows you to impact product development, which can be hugely satisfying. You will gain exposure to only a few tools, which can be refreshing after coming from an agency, and let you gain deep tool knowledge. But having an impact on the clients may be subdued as they will perceive you as having a one-track mind of selling to them and not having their best interests at heart. You can work for a vendor as a consultant, which would be a hybrid option similar to working for an agency and a vendor

Adam recommended starting one's career in Digital Analytics client-side to gain a solid grounding in how companies apply to Digital Analytics. This surprised me because I felt that there was a consensus about starting with an agency or a consultancy. The broad exposure allows you to discover what you prefer regarding tools stack, analysis or implementation, client-facing or not, and an affinity for travelling. Then, once you have identified what you like, you could seek a new role either at a vendor or client-side. Adam explained how starting agency-side or vendor-side lacks practicality. As an agency consultant, your recommendation may make sense on paper, but your client will know it will not work. As a vendor employee, you may have an idea for a new product feature your clients will never use.

So, which option is best will depend on everyone. Having worked in all three sorts of roles in over a decade in Digital Analytics, this session felt particularly relevant for me. But I am also wondering what the future has in store for me. That's when Adam unveiled the fourth option: evangelist/independent expert. The main pre-requisite is to develop a personal brand and launch our own company. You must blog and be consistent about it, speak at conferences, sell online courses and consultancy services, publish newsletters, or even a book. This is pretty much what I have been doing with speaking at MeasureCamp since 2016, publishing posts like these on LinkedIn for almost a year, nearly every Sunday. I have launched a free Javascript course for Technical Web Analysts every Monday evening on Zoom. I am looking into publishing newsletters, one for the Javascript course and another for my LinkedIn articles. I may launch the latter in multiple languages, so what this space!

After a lunch of delicious bagels from a local bakery, the time for my first session was looming fast. How many would turn up when Simo Ahava and Mikko Pippo spoke simultaneously as me? The session started with confusion as one of the organisers offered to switch rooms. About 20 people poured into the room I had booked, some saying: "oh, it's much better!" but only one was there for my session; the rest came from a smaller room. So, I had a chat with Liina ( Karoliina "Liina" K. ). As an organiser, it may have been the only session she could attend. But however small the audience was, Liina shared some valuable insight with me. I wanted my session to be a platform for other attendees to discover how Scandinavia builds its talent pool of Digital Analysts by having some IHM Business School and/or Medieinstitutet alums talk about their experience to attendees from countries outside Scandinavia or Finland. These two schools are both in Sweden; both have campuses in Stockholm, Gothenburg, and Malm?, i.e. Sweden's three largest cities, the last two on the West Coast of Sweden.

• IHM Business School: 2-year degree
• Medieinstitutet: 18-month degree

Both schools offer a comparable syllabus; both require that their students do an internship each year, called #LIA, so two internships in total at each school where the employer must follow a plan where the students will be able to put into practice the teaching modules they saw at school and validate these modules. Both schools have a steering committee made of members from employers such as digital agencies and client-side companies in Sweden. The result is a curriculum matching employers' needs and enhances students' chances of finding a job after graduation. The students often get their first job where they did their LIA internship. I believe some steering committee members are also alums of these two schools. There is something to say about having a manager who was someone just like you a few years ago. As a prospective student, it shows you what your career could look like. According to the annual Harnham salary surveys, career progression matters a lot, which reveals year after year that the lack of career progression is of the top three reasons for data practitioners to leave their roles.

This brings me to my discussion with Liina. I asked her whether Sweden is an exception in the Nordic countries. Are there other similar schools in Finland and other Nordic countries? There is an emerging educational effort to teach Digital Analytics in Finland. A few schools have started teaching the subject in Finland, but none with the reputation of IHM or Medieinstitutet. Linna has also made a few overtures to see which schools would be interested in collaborating or expanding beyond Sweden. There is interest, but it is too early to say much more about it.

Due to both Swedish schools having campuses in Gothenburg and Malm?, Denmark and perhaps Norway can benefit from that advantageous geographic position on Sweden's West Coast. Gothenburg is relatively close to Oslo, Norway's capital, but Norway is very quiet on the Digital Analytics scene. Malm?, on the other hand, is one-half of a large metropolis as it lies directly across from Copenhagen, connected by a bridge and train. Creating another Digital Marketing school in Denmark, especially in Copenhagen, may not make much business sense. But it could happen in ?rhus, Denmark's second-largest city in Jutland, Denmark's region connected to mainland Europe, as it is several hours' drive from Copenhagen. I have no information on any such efforts in ?rhus, though. In any case, IHM and Medieinstitutet allow their students to do their LIA internship abroad, not just in Sweden. And, of course, nothing prevents their graduates from taking a position in another Nordic country outside Sweden.

My next session was with Peter Meyer sharing his results with two automated testing solutions: Cypress and Microsoft Playwright. Both tools allow you to run Javascript code on a website to simulate user behaviour. You can compare both with Digital Analytics governance tools like ObservePoint:

Cypress

• Can run off a cloud platform or as a software install
• Chromeless or not
• Can generate screenshots and videos
• Relies on a third-party tool to reproduce the actions you want to repeat – it is similar to recording macros in Microsoft Office products

?

Microsoft Playwright

• Software install only
• Chromeless only
• Can generate screenshots and videos, like Cypress above
• Has its own code generator; Cypress above requires a third-party tool
• Supports several browsers, unlike Cypress
• Excellent integration with Microsoft Visual Code with extensions

?

ObservePoint

• SaaS tool, so no install required
• Chromeless only, based on Node and Puppeteer, so it can only emulate Chrome incognito
• Generates screenshots, no videos, to my knowledge
• Has a Web Journey Manager browser extension – for Chrome only, I think
• Recognises tags from many vendors; can add support for new ones free of change
• Scheduling and email alerts included

Peter's session was relevant to me because I use ObservePoint every day. But I also developed a pet project called BananaFarm which does something similar on top of a browser extension called Tampermonkey. The whole BananaFarm saga started when Simo Ahava asked me a few years ago to deliver a course on the browser for his Technical Marketer Minidegree programme. I dug deep into DevTools and learned how the Sources tab lets you save scripts for future use. I created one snippet for automatically populating the many pages of forms under our remit. I work for an insurance company. But I am too lazy to open DevTools, select the Sources tab, go to my snippets folder, right-click and run the scripts. I wanted to assign shortcut keys to each of my scripts. This is how I found Tampermonkey and ported all my snippets. My little project has grown a lot since. Peter started with Cypress but prefers Microsoft Playwright now. I can't wait to try them out.

Then the time came for my second session. This time, I wanted to talk about Javascript Senpai, my free live Javascript course for Technical Web Analysts, every Monday at 5.20 pm London time. The course is on Zoom and runs for 40 minutes. I ask the students to propose a lesson topic, usually not vendor-specific, since I am on the Adobe stack and barely know the Google stack. I look into the lesson topics, write the code for them, and take the students through the code with a demo. I do not provide the code, but I make two exceptions:

• You submitted the topic, so as a token of my appreciation, you get the code for free – it's always nice when I am running out of ideas, and it addresses a need for at least one of you
• You published a shout on LinkedIn or Twitter about the Javascript Senpai lesson you have attended

Right now, I have about 5 or 6 students every Monday. I often have to pick a topic, and I am running out of ideas quickly. So, the purpose of my session was to drum up interest for more people, typically Web and Digital Analysts, who lack coding experience but now need to do tagging on top of analysing data. After going around the table, we discovered that only two attendees out of the dozen who attended my session had previous work experience as a Web Developer, so it felt relevant to many in the room. I wonder if some of them will join tomorrow's Javascript Senpai lesson. I still have no topic for tomorrow, though! For more on this subject, I published an article on LinkedIn last year about the Technical Web Analyst concept.

There were two more session slots after this, but none felt relevant, so I decided to start networking early and not wait for the after-party drinks. I met Alexander Kirtzel at Ebbwalker from Hamburg, Germany. Alexander has a framework to help implement Digital Analytics which resembles an older pet project of mine called Kermit. You can leverage HTML5 data- attributes covered in a previous Javascript Senpai lesson- to declare all your page names, custom dimensions, etc. I have tagged my entire personal website using Kermit, and it works particularly well for tagging single-page applications. I have a demo with Angular 1 using two different routing modules on my personal website. You can use DOM mutation observers in Javascript to catch any value changes to HTML5 data- attributes. If one of them contains the page name, then you know that the screen has updated, and you can fire a custom Javascript event, which provides a helpful hook for anything to respond to it. Here are a few advantages of such a tagging method:

• No need to speak to your developers to get any single-page application tagged anymore – this is framework-agnostic
• Should you need to migrate to a new Digital Analytics stack, half of your migration is already done; you can keep all your HTML5 data- attributes; just replace the module that listens to the custom Javascript events and transforms them into tracking network requests
• Inspect the element and know immediately whether a link is tracked
• Generate an SHA checksum that will let you spot immediately whether the devs have messed up with your Digital Analytics tagging

As I am nearly home in Wales, I want to thank everybody I saw yesterday. But special thanks to Liina and all the organisers who have organised a not-so-chaotic MeasureCamp; you have pulled off something unique! And also, I would like to thank the sponsors, without whom none of this would have been possible.

For more articles like this, please follow me on LinkedIn or Twitter: @albangerome

#MeasureCampHEL #MeasureCamp #CBUSWAW #WAWCPH #DigitalAnalytics #DigitalAnalyticsAssociation