I kill dashboards and data noise
Wabi-Sabi. Here are two words you need to remember.
Data is asymmetric, imperfect, impermanent, and incomplete – as is your practice. It will never be perfect, you will never have all the data you need in one place, and you will never be able to achieve perfection, but the secret is to create each dashboard as if it were a final product, complete with all its flaws and faults, then, when it comes to the next one, you can improve your practice.
Every time you work on your dashboard, you are bringing in new insights that you didn’t have before and you are incorporating market changes.
This means that whatever you have now or by the time you are finished, your data will never be quite how you want it, and this is reality, so accept what you have.
We are drowning in data, so much so that we are becoming more and more unaware of our decision tree and what it is that really affects our performance. It’s not an easy thing to drive action; we have all been there.
We have an incredible amount of data within our grasp, but at the end of the day, it doesn't matter if it's a CSV file or a proper data infrastructure, when it comes to deciding where we should invest our next Euro – in a new product feature, improving or killing a problematic feature, or pushing to acquire new users – we still don't have the answers we need.
User score is a KPI that combines a lot of metadata into one number; a single weighted number that tells you the quality of your newly-acquired or returning user. I believe that this is the most important number you could possibly create, and if you haven't by now, you should do so.
It’s the one number that will tell you how effectively your product matches the user and is based on your assumptions about your product and your intentions for it coupled with what the user actually understands your product to be.
Your assumptions about usage fit the users you bring on board. The user score takes into account external information on the one side, like device and location, and the user's app usage funnel steps on the other.
Finding your north star isn't at all what you thought. I love the articles I see from time to time on my Linkedin feed, or just on some random search, promising that we will one day find our north star KPI – the one that will help to solve all our problems – and if you're still unconvinced about this amazing KPI (which is mostly ROI related) you should know it is used by many successful companies within the industry...
But hold up, let's face reality here. If you need someone else to tell you what KPIs you should look at or someone to promise you that these KPIs will result in your overnight success, well, I hate to break it to you, but that's bullshit.
The reality is this: there is no north star – well, there isn't one per se. In fact, when I look at my numbers, the only thing that I would even consider to be a trustworthy KPI – my north star – would be my user score. The trouble is that this KPI will also lead us to make mistakes.
Don’t get me wrong, I'm not against ROI or CPI (cost per install) or any of the common KPIs. I was slammed enough when I predicted that automation would change performance back in 2017!
Now I'm looking to the future, a future in which we're more concerned about creating relevant creatives, talking to the users we want to attract and KPIs that set out quality plus how well a product matches the user.
Apple’s privacy changes are only one small step in a larger industry-wide change and today's users are more aware of what qualities they're searching for in a product. That's why just looking at your income versus your spend makes no sense whatsoever.
Obtaining a set of KPIs that support your north star requires some serious thought. Most of the time, people tend to go with ROI or CPI, but I actually believe that with a strong user score that focuses on the interaction between user and product, opting instead for CPI just won't help simplify decision making.
Let me ask you this: what would your reaction be if your costs were to rise or fall? What we should really be asking ourselves is did we bring the right users on board, and if not, do we need to adapt our creatives or our product?
Now don’t tell me your CEO won’t find this an easy decision to make – where and what should be changed – and no, it's not going to cost you your job. On the contrary, it will help you speak with your clients more directly.
Data strategy won’t help you to solve your problem, but it will help you focus. I once found myself in a workshop after which the CEO said: “Great! Now that we've got our data strategy in place we can grow the business beyond our one million users.
I expect everyone to focus on the numbers we agreed on.” Well, he was right about the second part – we need to stay focused on the numbers we agreed on and understand that there's a big difference between analysis and dashboards.
Analysis is what helps you to explore unknown phenomena, while the dashboard is there to help you make decisions. However, if this is your first dashboard design, I recommend giving the data time to sink in.
Give yourself a period of three months to use the dashboard and note down things in the daily data that aren't working well. Understand what it is you actually need on a daily basis and rebuild your data strategy from there.
Hummus is perfect when it’s consumed fresh… yes, back to my hummus! After years of eating hummus, I discovered that it's best when it's freshly made and has never sat in a fridge.
Data is the same. When you consume fresh data, it means you can build a data infrastructure that enables you to obtain data on a daily basis, refreshed and relevant. You don't need live data, however relevant it is.
Numbers numbers numbers… don’t let them control you! You have so much data, but it doesn’t mean you need to quantify everything and use all the numbers you can.
Instead, keep things simple. Try to restrict yourself to three daily KPIs and a maximum of 23 KPIs per month so you focus more on your users rather than your bank account. Remember, focusing on your users means you are focusing on your bank account!
Just go with what you believe is the right set of KPIs to enable you to make decisions, and remember to find a north star that will give you an indication of user quality, not just user spend and income.
The more your users fit your product, the higher your chances of successfully converting them. Remember that 80% of these users will visit you once or twice, so make sure you have maximum impact!
Previously published at https://www.linkedin.com/pulse/wabi-sabi-your-data-lior-barak/
Create your free account to unlock your custom reading experience.