Measurement is a department, not a project.

You would never hire a CEO for only 3 months, so why would you do that with your data??This is rhetorical question, but many do this unintentionally.

Measurement is a department, not a project.?

What are the attributes of a well-maintained “data department”?

1. Data Dictionary/Center of Excellence?(CoE)

Where do I go look for what has already been established? (this is the first step to stay organized and reduce redundancies). Even if you only have an outdated UTM naming convention spreadsheet, that is better than not having a data dictionary at all.

It’s all about being “good enough to get going” and you can always make it better later on.?

2. What questions are we trying to have answered??

A crucial aspect of a robust data department are the quality of questions you are asking and the actions that can be taken as a result. Just like in other areas of Marketing, the questions will evolve and “level up” over time.?

Some questions aren’t even possible in shorter sprints, depending on the data infrastructure and tooling.?

3. What does the user journey look like? (documentation is key)

An agreed-upon “user journey” is part of our planning process (Planning, Building, Launching, Forcecasting, Optimzing are our 5 stages).?

We’re big fans of LucidChart and use it to not only docuement the user joruney, but map events (all the different types of events) to have an agreed upon “definition” of a funnel, as on example.?

Often times user journeys change on the site. (Who’s idea was it to update the URL sctructure?! ??)?

It's crucial to make sure the existing tracking, and user behaviors we are measuring for, don’t break. But also to plan for historical reporting with this change/comparison periods, etc. Annotations are key and can't be viewed as “optional”.?

4. What are our UTM naming conventions??

Standardization is key. Just like annotations, these aren’t optional and what UTMs you fill in also aren’t optional. If they are treated as such, there will be gaping holes in your data and “direct / none” will be your source / medium report…which we all love ??

One trick here is to ask your future self if “will this be useful information for me to report on”.?

For example, if this was my UTM rednavelconsulting.com:

?utm_source=linkedin&utm_medium=share&utm_term=article&utm_content=data-department&utm_campaign=thought-leadership        

I would pose this question (to myself, very inceptional):

"So if I was to tell you, Ryan, that this click came from LinkedIn (source), it was organic - not paid (what we use 'share' for), it was from your thought leadership campaign (thought leadership as it's a content type for me, can be grouped together or 'rolled up' in reporting), it was a LinkedIn article (article for content meaning it came from that LinkedIn content type), and the individusal article was about data departments (data-department for the content UTM)…

…would that be helpful?”

If not, speak now or hold your peace for the next UTM naming convention update ??

5. Annotations

I'd say about 10% of companies we start to work with are using annotations to a point where they are actually useful. Analysis can be hard, the point of annotating is so that the trends and patterns match up to expectations and are easier to spot.

We advise clients to create a dedicated Slack channel for annotations. Even if it's that alone, it's helpful. But we take it one step further and sync it up to our annotations tool of choice, Crystal Ball (which shows an overlay on GA4, GAds, Looker Studio, syncs with Zapier, etc):

Why is data a department, not a project??

• Your questions evolve?

Think of the questions you are asking when you first start with a company compared to 90 days in...they are so much better and lead to deeper insight, right?!

• Your marketing goals change?

Pivoting is key, especially as your data collection and research methods change/evolve, so should your Marketing plans.

• Data changes?

The data itself rarely changes, but the way it is collected certainly does. Therefore, what is stored changes. Often times for better or worse...but we can always clean it. Let's just hope it's consistently inconsistent ??

• Privacy policies change?

Change is the constant here and probably will be for some time. If you aren't keeping up with this then you really aren't being the data professional you need to in today's climate.

• It’s something you should be looking at weekly and in order for it to be constantly actionable, the reports will have to evolve.?

Making dashboards actionable is hard. But so, so important. If your dashboard isn't actionable within a first initial glance, you probably aren't building them right.

You shouldn't need "more time to do analysis", instead, you should make simpler dashboards.

Why should you hire an agency/consultant instead of in-house??

Data is all we do - every day, all day.

There are so many more “gotchyas” with GA4 than other other version of Google Analytics or any other analytics tool out there, it’s really hard to keep up with and support.

We are 100% of the mindset that for some areas, you should build that "Measurement Muscle” in-house. For other areas, it would be much cheaper to work with Rednavel Consulting (or an equivalent Measurement-Consultant/Agency) instead of hiring that role in-house.

Let's take a look at some example job roles that we almost entirely replace acting as a "Measurement Department”

Here is one for a “Director of Performance Marketing”

Let’s take a look at some of the priorities:?

1. This should all be governed through GTM for the same “definition of a conversion” across all media buying and analytical platforms. The vision should be in the product/service page copy and communicated through the site. If you were trying to measure “how effective” is this messaging then you’d need to introduce qualitative research to properly assess.?

Quantitative analytics tells you the how and what, it will never tell you the “why” and that is what messaging testing/qualitative research is for.

We’re experts in both areas.?

2. Building a testing roadmap is easy, testing itself is hard. But if you build a system to process you can move mountains with your company. We have experience and expertise and true "testing departments”

3. We have long conversations (weekly) on attribution/propensity to buy modeling, etc. High likelihood that we have more experience than this person you’d be hiring. Also across all of our clients, you’d get that “what works for them” insight that I’d share that most companies find quite hard (industry insights that aren’t generic and are actionable).?

4. Our core competency is to run Marketing departments simliar to an accounting department; we’ve been doing that for years and ROAS/ROI/CAC (however you want to measure it) is what we do. And automate so you can see the trends and patterns in a dashboard to easily spot action.?No need for a Director of Performance Marketing to ask the analyists or engineering team to build them a query in the data lake, we do that.?

5. There might be some things here we simply wouldn’t do, but most of it presents itself within the data. (LTV, retention, CAC, etc) - what isn’t listen there is velocity of return (how long does it take to get to LTV. Everyone loves to talk about LTV, no one loves to talk about time to LTV. In the “profitability” focus we are in now-a-days, I think this trend will pick up.?

6. Most people here would simply dive into the quantitative data and action. Nothing wrong with that, but I disagree that only this would solve the issue of why chrun is happening. To solve the root issue, you should be using qualitative research (surveys to start) to better understand motivations and reasons for the churn in the first place.

This will help you form a VOC-driven hypothesis to run a test against.?

Let’s take a look at some of the skills and experience preferred:?

1. I have 10+ years experience. And experience with my own budget being used as well. You learn a lot faster when it's you own money. Most all Director of Performance Marketing aren't going to be pressing the buttons, they are going to be strategizing, let's honest and call it what it is here. We work with media buying agencies weekly in situations like this.?

2. So much of this listed here is about having “clean data” and that should never be taken for granted. We see so many companies “collect it all” and then spend 80 hours cleaning it in a warehouse when it could have been governed and scoped much better in the first place.

I would prioritize a well-defined ROI with the company instead of ROAS only.?

3. Speaking from a place of confidence, there are very few agencies and people out there that are bettter at analytics, tag managment, and bulding a measurement sytem over us.?

4. I would push back agaisnt Excel (that usually means data from desparate system and “ad hoc” stuff) and focus on the data warehouse and using Google Sheets with BigQuery instead of manual spreadsheets. We also use Looker Studio for all our reporting and it’s very easy to “export to sheets” from the dashboard.

Point here is live data and automation, not manual “pulling” of data.?

Instead of comparing this to a tool or your ad agency “bundling” a GA4 setup with your ad spend, you can clearly see why the comparison is at an in-house role instead. And how you almost always get more, for less, when comparing to that of a FTE making $150k/year.

Think Rednavel Consulting might be a good fit for you/your company?

Neat-o.

We'd love to talk ????