In conversations with any top executive of companies, you are almost always going to hear the mention of NPS or CSAT as a key metric used. Today, almost 67% of the Fortune 1000 firms say they measure it and use it as a strategic input in all planning.
NPS campaigns/ surveys are launched for a variety of audiences-
· End users or customers- to know how our product or service is doing
· Internal customers or cross-functional teams- to gauge performance of our business unit
· Employees- to know leadership is doing, how they feel at the organization
NPS, in its current form, came to be about 16 years ago by Fred Reichheld in 2003, in an HBR article titled “One Number You Need to Grow”. Some of the biggest reasons for its widespread adoption- and the reasons why it still makes sense, and might even be important are:
· Is easy to measure
· Is quantified- and hence reportable, trackable
· Is believed to be a representation of customer sentiment
· Is definitely a confidence booster- high NPS scores do tell you that top customer retention will not likely be your biggest problem compared to competitors.
· Helps you plan for new customer acquisitions- if the score is low- you know you can’t rely on referrals or Word of Mouth (WoM) to acquire new customers. This can help you strategize and plan better for marketing- on both acquisition and retention side.
· You can predict churn on the negative side- you know a low NPS is not getting you positive WoM- but negative WoM is most certainly happening- and this will have a downstream impact.
Rolling out an NPS campaign or a survey- and its subsequent measurement and reporting has become a “must-do” at most companies. Sadly, more often than not, not much thought goes into what the true aim of the process is and what the results really convey.
People either love NPS or hate it. Haters have a very good reason- the numerical NPS score really doesn’t tell me what to do with it. The current NPS result tells us how many promoters/ detractors my product or service has. A lot of question remain unanswered with the focus only on the MAGIC NPS number:
· I have a lot of promoters- so what? What are they doing- to my product or service? What do I do with them?
· Similarly, what are the detractors doing? And what do I do with them?
· What does a score mean?
· When do I say “This is good enough”?
· OK, so this score seems bad! I need to move it up. To what score?
· How do I move the NPS up?
Several experts in the field of surveys, campaigns and executive strategy, do believe that the NPS methodology is obsolete and the value it provides is minimal. There are other, more effective, means of measuring customer or employee satisfaction- which also provide clear, actionable feedback.
In the recent years, there has been an increasing thought about how NPS/ CSAT measurement is deeply flawed. If that is so, why is this campaign/ survey still so popular?
NPS is NOT entirely useless. After all, there is a very good reason so many of the Fortune 1000 companies use this score. The problem is the way many companies handle its design, execution, measurement and, most importantly, reporting. In my opinion, the root cause of the faulty handling and the biggest problem is the tyranny of the leaders who misuse NPS/ CSAT score as a convenient data point to lash the whip of terror on unsuspecting colleagues.
This article will focus on how to make the best use of the NPS system which you have in your organization. And in case you are planning to starting measuring NPS, how to best do so.
Assumption: Readers of this article are aware of what typically a Net Promoter Score (NPS) campaign or a Customer Satisfaction (CSAT) survey is in its basic form is.
Why are you measuring NPS?
Be very clear and write it down- what is the reason you want to measure NPS.
· Are you doing it because it has been done for years and is a mandate for you to report as a number in the annual report?
· Are you doing it because it is used as a tool in the calculation of anyone’s incentives/ hikes/ promotion?
· Is the company using the results to drive crucial marketing decisions?
· Are the inputs used in product or service design plans?
Keeping the campaign/ survey recipients truly “random”
In organizations with huge customer/ employee base, it will very often not be possible to send the survey to every single one of them. Hence comes in data sampling/ segmentation.
With a plethora of data science and simple common sense tools at our disposal, we should be honest in making the best effort to keep the sample really random and still sizeable to be a true representation of the audience.
All the important demographics should be factored in. For customers, it will be important to also bake in tenure, their lifetime value (LTV), potential value, geography apart from the regular demographic factors. What is important is to ensure true and sizeable sample from each segment to ensure the results are not skewed.
In my experience, some companies which run more than once- a-year NPS campaigns, often use the same “skeleton” of sampling and just pull out fresh contacts each time. While it is true that the base might not have dramatically changed over a quarter, it is still worthwhile to re-examine the sampling logic to ensure coverage and representation. What if a new product was launched- did we factor it in? Did we expand to a new geography? Did we close business some place?
Reaching out to the audience right
While right sampling ensures you reach out the right people, what is also important is reaching out to them at the right time in the right way.
Take out time to research what frequency suits your needs best? If yours is a fast changing market, may be quarterly campaign works best. In case you are a relatively mature business (say, a bank)- may be twice a year is good enough. Annual may still be effective- but then a lot of time passes without hearing from the audience and you are possibly leaving a potential opportunity/ issue untapped or unresolved for a substantially long time.
Also, work out what channel or combination of channels will work best. With employee satisfaction surveys, this is fairly simple since we know all our employees will have easy access to what system at what time. With customers, there should be sincere effort here. Are all your customers tech-savvy and mobile? If so, may be mobile UI based campaign would serve best. Are most of my customers middle aged semi-urban small business merchants? In this case may be a combination of telephonic and door-to-door visits would be the best idea. Of course, time taken to complete the exercise, data consolidation and budget would play deciding factors here. Any of these three going haywire would render the whole survey moot.
Once these are in place, it is important to launch the campaign at the right time. In a sales driven business, sending out the survey towards end of the quarter may be a bad idea. People are already under immense pressure of meeting their targets and the survey would only look like waste of time to them. In case of employee surveys, sending them out just after the appraisals will really not work. They are either buoyant or unhappy- and the survey responses are going to have a heavy “recency effect”.
Ensuring complete and true data collection
In case you had a multi-channel execution, it is important to consolidate all data, clean it up and ensure data completeness, accuracy and integrity. There could have been instances that one person responded twice using different channels. Data cleansing, hence, becomes difficult — but important. Given that there are excellent statistical tools and techniques available, in case the data capture was done well, data cleansing can be a relatively painless process.
It is important to remember- that as with all data driven processes, even NPS is fully prone to the “garbage in- garbage out” syndrome. Thus, even if it takes time, data preparation should be given high priority and honest, intentional effort should be made on the same.
Multi-faceted response analysis
Once the results are calculated, the one thing you know for sure is your Net Promoter Score on a scale of -100 to 100. Great, keep this aside.
Deep dive into the qualitative feedback- with all the expertise we have in sentiment analysis and pattern recognition, the most worthwhile exercise would be to implement these on the NPS survey feedback. You could choose to bring together all feedback from last several years together, or even choose to analyze year on year to study patterns over time.
Look for right things to address- Once the analysis of all qualitative feedback is complete, try and pull out areas which seem either most recurrent or most different from what you expected.
Ultimately, we are all trying to put the “Customer First”. So look for areas which talk about things that you can do or do differently to enhance the experience of the customer.
In my experience, segmented response analysis has proven to be very insightful. Along with the scores, the one thing which is also done easily is segmenting your respondents into “detractors”, “passives” and “promoters”. The way to go about segmented response analysis is to segregate the feedback from the detractors, passives and promoters into three different sets- and analyze them separately. This is very effective in finding out what are the attributes that are driving a certain user behavior or are the cause behind a certain feedback in a given bucket. Hence, more actionable planning can be made at targeted audience segments- resulting in better efficacy.
In your analysis of the feedback from the promoters, you should look for what they are expecting more. Often, your promoters will give a lot of ideas in the feedback- these could very well go onto the drawing board for your planning sessions.
The feedback from detractors will give you immediate actionables- since it is mostly around shortcomings in your current service or product. They also talk about what other non-core expectations they have from you- which may be your competitors are providing. These should be focus areas you look for in your analysis of the detractors.
Passive customers are the crucial “swing” bloc and their feedback should be taken very seriously. In case recurrent themes appear, the feedback should even be able to shape the direction in which your service or product is heading. The way your evolve the service/ product will decide whether this bloc moves to being promoters or detractors.
Validation of the analysis
By now the one thing we have been consistent at- is keeping the customer first and customer experience at the center. So, it only naturally follows that we validate our findings or analysis with them.
There could be various techniques to do this: Focus groups, personal interviews, etc. depending on where you are in the feedback analysis and action implementation plan. Irrespective of which stage you are at, it is imperative to go back to your customers and speak with them.
Draw out the top 5 points from all the analysis and discuss them with the customers/ employees (whoever was the survey audience)
Go deeper into the feedback and gather more information. Is the problem genuine? Is it recurrent? Is an idea you got really worthwhile.
Validation of the analysis will also add a lot of confidence in action planning along with supplemental information which would be valuable in implementation rollout.
Analyze the implementation as well- not in detail- but to the extent that it allows you to firm up or reject a recommendation. Design thinking methods have proven to be the best bet in this. The real-win-worth it approach designed by MIT will make it objective and simple.
· Is the problem/ idea real?
· Is the problem worth solving? OR Is the idea worth implementing?
· Can we really solve the problem/ implement the idea?
· What is the cost of not solving a problem/ not implementing an idea?
· What will it cost to solve the problem/ implement the idea?
· What will be the cost of not changing?
Staying honest in reporting
Stop looking at the NPS score- this has always been the biggest distraction. The promoter/ detractor numbers really don’t tell us much and we get stuck up on this number for no good reason.
Forget what the senior management would “like” to see on the report- report on the facts, on the real findings and on the true, uncolored recommendations.
The key is to remember that- the goal of NPS campaigns is NOT to get high, awesome scores. The goal is to “hear” our customers. And this is the fundamental premise where most companies today fail. We “tweak” the design, execution and measurement long enough to have the numbers tell us what we want them to tell us- because our focus is on getting a high NPS- instead of unearthing what the Voice of Customer (VoC) is telling us.
Please remember that NPS score is a trailing indicator. This means that you are measuring how well you have done on something AFTER the event(s) has happened. By the time you measured it, there has been a problem- that has gone unaddressed for weeks, or even months.
Also, another important myth to debunk is that NPS is really not about loyalty. It is more about HOPE. We are asking customers if they will recommend us to someone- this is not loyalty. Loyalty would be more like- have our customers been recommending us to others.
In summary, NPS may or may not be the best tool of choice out there to measure the satisfaction of your employees, customers or stakeholders. But as long as you are using it as an instrument of measurement, it would be best to design it use it right- to reap maximum benefits from the same.
I hope you found this article useful. Please share your thoughts in the comments section- I am sure I have so much more to learn! J