Making data-driven business decisions is hard. Companies are overwhelmed with all types of numbers, sets, variables, and approaches. We try to build complicated models, while the 80/20 method is sometimes the best thing we need.

The following post will tell you:

1. How to differentiate between “owned” and “gathered” data?
2. What is the Net Promoter Score?
3. How can you quickly apply NPS to your daily operations and use it?
4. Why you should always “Aim for the 100!”

Data in business

Modern business is a measuring contest. We gather data, try to calculate them, and hope for analytics-driven decisions. However, it’s not that easy to crunch the infinite sets of data and do something valuable with them. There are just too many options now.

‘Owned’ metrics in business

If it was a regular blog post, I would go with quotations that world doubled data volume in the last year or something like that. However, I have simply typed in Google: “metrics in startups’ and got 24.5 million (!) of search queries

It’s a lot to handle. Some of them are supercritical, and you can’t cope without them (e.g., MRR, ARR, or CLTV), but if you do everything right from scratch, they’re already in the process. Usually, we have them embedded in the systems, and use pulling them out is not a problem. Let’s call them “owned” type of data. We’re not going into them today (and Philip is much more of an expert in it than I am btw.). What you need to remember is that they are numerical, quantitative, and it’s easy to calculate them, but much harder to interpret.

How to conduct marketing research? The assessment of “gathered” data.

The real challenge starts when you want to perform more “informative” analysis, e.g., surveys or customer feedback comments, customer interviews, focus groups. To simplify, let’s name them as “gathered.” These are hard to gather and analyze, but relatively easy to interpret.

From my experience, the majority of people don’t know how to create a proper survey and conduct research. There is a common misconception that survey is easy-to-do, and pretty much everyone can create it. I’ve seen hundreds of badly-written questions. Some of them suggest an answer, others mess up the scale, or asks the wrong thing to the responder.

Surveys are the tip of the iceberg. For instance, we also have string variables with vast amounts of text data, which are hard to assess, and at least they’re time-consuming. Natural language programming is still developing, so we can’t rely on any type of software as well (trust me, I’ve tried them).

If I am not knowledgeable about research, then I’m doomed in performing the right type of analysis? The answer is: yes, but there is a turnaround here. Top-tier consulting firms call it an 80/20 approach. You may think of it as Pareto rule, but here it pretty much tells you that you need 20% of data to have 80% of results.

There are multiple ways of how-to-handle quick analyses, but today let’s focus on the one I’ve recently developed, which is a combination of Net Promoter Score and KCD questions (What to Keep/Change/Delete)?

How to apply easy Net Promoter Score to daily decision-making?

You’ve probably heard of Net Promoter Score. If yes, move to the next paragraph, if not, here you go.

The NPS is a 0-10 scale question you can ask your customers:

„How likely is that you would recommend X to a friend or colleague?”.

Then answers:

• 0-6 are „Detractors”;
• 7-8 are „Passives”;
• 9-10 are „Promoters”.

You calculate it in an equation: % of “Promoters” in the sample - % of “Detractors”; you need to skip the “Passives.”

Promoters” are the ones you need to grow -they are the loyal ones, will remain customers, and you may cross/up-sell them.

“Passives” you need to convert into loyal, boost their satisfaction, and try to satisfy them more as they can smoothly go to the competition.

“Detractors” are unhappy customers who have high churn and definitely will tell a lot of lousy word-of-mouth stuff, you can try to fix it, but it will be hard.

The score goes from -100 to +100. Everything positive is good, but everything +50 is incredible. You can check top companies NPS for free here.

Two-thirds of Fortune 1000 companies applied it in their daily operations, so it may be useful to know it. London School of Economics scientists discovered a statistically significant correlation between revenue growth and companies’ NPS scoring. According to the research, an NPS increase by 7 points correlates with 1% growth in revenue. It may be worth trying.

Now let’s move to the 2nd point.

University students’ case study. How to apply NPS + KCD questions?

I frequently have classes with university students. If you want to excel as a teacher, you need to evaluate your work and try harder. I’ve asked my students to rate me on an NPS scale and asked them three feedback questions: “What to Keep, What to Change and What to Kill in our classes?” Simple as that, I immediately got four relevant data points to check the hypotheses:

• NPS: How my “customers” (students) evaluate me? What is their distribution? What are the differences between them?
• Keep: What works well in my classes, and how does that associates with numbers? E.g., whether “10” on NPS for “Promoters” is more frequent for students who liked my approach to classes or not?
• Change: What should be changed immediately? What are the quick-wins? E.g., whether dissatisfied students were also “Detractors”?
• Delete: Which element is not needed? Maybe I’m doing too much somewhere, and I can reduce to make it leaner? E.g., whether Passive students find something not engaging at all?

To conclude, I have one quantitative metric, which I separate into three buckets and three questions for three separate groups: Promoters, Passives, and Detractors.

Why is that important? I had classes with two separate groups of students. If I had calculated the quick survey results with average, I would score 9.62 for first and 9.00 for the second one. It’s not a statistically significant difference to start thinking about it, but what if we take the NPS scale? In the first group, my NPS was “+93,” and in second, it was only “+70”. It was still terrific, but the difference was already visible.

I started to think, also looking at qualitative Keep/Change/Delete answers. Some students in the 2nd group mentioned I should be “more to the point,” “keep the tempo”, “be more focused.”

It got my mind, so I began to A/B differentiate the groups and found out that the variable, which had an impact on opinions, was “time of classes.” One class was earlier than the other.

Simple as that, I realized that I was more tired and less focused and productive as time went by. I was too engaged in the first group, using about 70% of my energy, and didn’t have it for the second one. It was game-changing for me and the way I teach classes right now. Nowadays, I try to be calmer, slowed down, and more balanced, so I don’t exhaust too fast and save energy for later. Overall, teaching is also like long-distance running.

I have done this analysis after a few pieces of training I’ve conducted; one can use it as a feedback question on retrospective team sessions; your customer success manager can ask it directly to the client. I keep on asking these questions to my clients, students, and even co-workers. It’s mighty, considering how simple it is.

This type of quant/qual analysis can be improved, changed, and tailored for your needs. The only thing you need to remember is the pattern: foundation (quantitative) question + informative deep-dive (qualitative). It gives you excellent 80/20 coverage, which you can easily leverage in the future. It is much more insightful than just a standard Net Promoter Score, which many companies heavily use.

Aiming for a daily „100”

As you see, there is a considerable difference between the average opinions and those calculated with NPS. Standard calculations wouldn’t be right to see the hidden patterns. Applying quantitative metrics + qualitative insights is a potential way-to-go for every business but also a performance-focused person.

If you aim for the “100” in your NPS scale, you may achieve loyal customers, satisfied employees, and an engaged audience. If there are any deviations from the top score, you can always deep-dive and see the reasons.

It’s an easy hack for those who are not experts in survey-typed questions or don’t have time to focus on that while growing their business. You can even put a threshold: “if it’s lower than X, I’m going to investigate.” I love rules-of-thumb, and in this case, it’s the way to do it!

I hope this helps you. Please shoot me an e-mail at maciejwilczynski24@gmail.com or comment if you like to discuss or ask for anything.

Maciek

Questions to think:

1. Do I have a similar quantitative + qualitative approach?
   If not, why I don’t do it?
   If yes, which metrics and questions I’m using?
   How to apply NPS + KCD questions in my daily operations?
   Are there any different “turnaround analyses” I can do to achieve the goals?

‍