The Ultimate SaaS Guide to Setting Up the Perfect NPS Campaign


Guest Post by Chad Keck, Co-Founder and CEO, Promoter.io

As I’ve said before, if you’re not seeing at least a 10x return on the cost of measuring NPS, you’re doing it the wrong way.

Doing it the right way means that you’re proactively reducing churn, improving your product based on qualitative and actionable feedback, identifying new opportunities and growing your revenue from organic recommendations and referrals each month. All combined, the cost of measuring NPS is a drop in the bucket compared to the results.

As Jason Lemkin has stated, up to 80% of a SaaS company’s revenue is driven by referrals and word of mouth (this isn’t far off for most other industries either).

If you’re instituting NPS incorrectly, solely focusing on the number, or simply not measuring it at all, you could be missing out on the biggest revenue opportunity you have today.

That’s why we say, when done correctly, NPS is a measurable profit-center, not a cost center.

But, what is the right way to measure and implement NPS? How do you know if you’re doing it right?

Worry not young NPS Padawan … it’s never too late to learn the ways of the force.

The guide below is meant to walk you through each step of the process, to ensure that you’re set up for success.

Who should you survey?

Most SaaS companies have both free (i.e. trial) and paid customers. Both groups are incredibly important to measure, but not necessarily together. You should treat them separately.

Why separate the customers (hint: trials are not customers, yet)?

  1. If you’re using your Net Promoter Score as a KPI, your trial customers are likely to drive that number down. For all intents and purposes, this isn’t a real reflection of customer sentiment, but rather a skewed result. There is little basis for true feedback during most trial periods since the user has had little time with the brand or product.
  2. While the score isn’t as important for trial customers, the insights are. Trial customers are great at telling you where you need to improve to earn their business or what is not clear to them from a value perspective. When combined with a trend analysis, your product roadmap starts to become crystal clear.
  3. Paid customers require the most immediate attention, which is best managed when they’re isolated from trial customers.

The best way to set up this process is to create two separate campaigns which run in parallel with one another. This will allow you to create separate survey triggers for each customer segment (which we’ll cover below), as well as unique scheduling cadences.

Alternatively, if you’d prefer to have everyone on the same campaign, you could use attribute filtering to segment your customers after your results are in.

As trial users convert, they can be moved to the “customers” bucket/list for future surveys to keep your data aligned.

Assign attributes to your customers

Screen Shot 2016-08-31 at 3.02.52 PM

To put it in simple terms, attributes are the additional data points that help you define each customer. For example, job title, location, plan-type, avg. revenue, number of users, etc.

I can’t stress enough how valuable it is to be able to filter your results down by customer variables and dig into differences in sentiment by various customer segments.

Let me give you an example. Let’s say that you offer 4 different plans that range in price from $25 to $600 per month. Your overall NPS score is 30 (which, for the record, is a pretty good score), but you’re curious how that breaks down.

Using the filters you put in place, you discover that the 50% of your customers that fall into your 3 top plans have a combined NPS score of 50, while the other 50% of your $25 customers have a combined score of 5.

Do you think that is more insightful than your overall score? Can you see how filtering your results by attributes can help you narrow the focus and understand where to dig in further?

Adding attributes to contact records can be as simple as adding additional columns to your data import or passing those attributes in systematically through our API or another available integration (from your CRM, ticketing system, analytics tool, etc).

If you’re triggering surveys based on events, you’ll want to be sure to send that additional data as part of your integration.

Here’s a step-by-step document that you can follow to help you out.

How should I be sending surveys?

This is one of the most important strategic steps to the NPS process and has several components to consider.

Sending Method

If you’re looking to measure the sentiment of your customers overall experience (true Net Promoter Score), there is no better method than email. It’s personal, non-intrusive, and will provide you with the most holistic insights from your customers.

As a SaaS company, you may be tempted to deliver your surveys in-app. I mean, it seems like a logical place to reach your customers.

If you’re simply after just a score, in-app surveys may serve your needs. If you’re only looking for feedback relevant to what your customer is doing in the moment when they saw your survey instead of their overall brand experience, in-app will work.

Simply put, in-app surveys largely drive transactional feedback while disrupting the workflow of your customers. While you may get a slightly higher level of scoring responses, you’ll get less open-ended insights (which is where the bulk of the value comes from) and it’s hard to determine the accuracy of the scores since some users want to simply dismiss the pop-up.

Check out, Why In-App NPS is a Bad Idea, for further information.

Triggering Surveys

Screen Shot 2016-08-31 at 3.03.03 PM

While in-app NPS delivers transactional feedback, email can still be used to trigger surveys based on transactional events. In fact, that’s what you should be doing.

Essentially, there are two approaches to triggering your surveys: 1. Upload a database of your customers and send or 2. Integrate your CRM (or other tools) to trigger surveys based on an event, such as a conversion.

  1. Uploading a database is a fine approach to begin with, especially if you have a lower number of customers or you have a lower volume of new transactions.
  2. Triggering a survey based on an event will help build a consistent cadence, not to mention that it’s designed to fit within your current workflow. These can be easily set up with either our API or tools such as Segment and Zapier.

For best results, you’ll want to be sure to set a survey delay for new customers so there is time to form a basis for constructive feedback (this is usually anywhere from 15-45 days for a new customer depending on the product).

Sending Approach

Whether you’re triggering your surveys individually or uploading your contacts all at once, you want to get into the habit of a daily cadence.

Sending all of your surveys out at the same time is a bad idea for many reasons.

The first of which is feedback management. Depending on the size of your customer base, it’s possible that you could be inundated with feedback, making it nearly impossible to respond within a timely manner. The second reason is that sentiment changes rapidly, you want to ensure that you have a constant stream of feedback to stay on top of the pulse of your customers.

If you’re triggering surveys based on an event, you’ll naturally build a frequency into your sends. If you’re uploading a database, be sure to drip your surveys out over an appropriate amount of time, effectively removing any gaps between reoccurrences. Promoter has a built-in option for facilitating drip surveys to make this extremely simple to implement.

Engagement Frequency

As I mentioned above, customer sentiment changes rapidly. Minimizing churn is essential if you’re to succeed as a SaaS product. The only way to stay ahead of unhappy customers leaving is to know when they become unhappy (and why).

The best way to accomplish this is to survey/engage your customers on a recurring quarterly basis. Statistically, anywhere from 30-50% of your detractors will churn within 90 days, but by surveying them once every three months, you’ll have the opportunity to stay one step ahead and step in before a customer has made the decision to churn or worse, spread their negative sentiment to others.


To get an additional boost in responses, be sure to send a reminder survey if they didn’t respond to your first attempt. Over the millions of surveys we’ve sent, reminders average a 5-15% increase in survey completions!


Following the approaches I mentioned above will ensure that you’re maximizing the results of your customer engagement process and growth efforts using NPS. Never miss an opportunity to learn, grow your business and build stronger relationships with your customers.

Chad Keck is the Co-Founder and CEO of Promoter.io, the most comprehensive customer intelligence & engagement platform built to drive growth and customer retention using NPS (Net Promoter). Sign up today for free and be engaging with your customers in less than 5 minutes.

Published on August 31, 2016


  1. As cofounder of the NPS platform Wootric, I 100% agree with the power of NPS beyond the score — and the absolute importance of qualitative feedback. We at Wootric live this for our customers and ourselves.

    However, I fundamentally disagree with your steadfast focus on email as the Net Promoter high road. SaaS customers live in a multi-channel world! We need to learn where our customer wants to engage and be intelligent about the right channel and the right time to survey the right customer. We find that customers will answer the question they are asked, and are absolutely able and willing to provide higher level feedback when prompted in a web or mobile app.

    The high response rate that in-app NPS surveys deliver really does matter, especially for SaaS businesses focused on reducing churn. I’d rather get a gut impression from more of my users that I can follow up on than radio silence from a passive or unhappy user that ignores my email survey.

    While email is the ‘traditional’ way and right for certain types of users, NPS is evolving along with our customers and our businesses.

    1. @jessica_pfeifer:disqus I had a feeling I’d hear from you 🙂 Always happy to engage in constructive conversation on this topic.

      Let me reiterate what I’ve said to you before when you’ve hijacked our other pieces of content and injected your opinion and product promotion.

      Speaking candidly, your approach to NPS is fundamentally flawed.

      You simply can’t re-engineer a methodology that wasn’t broken to begin with, change the point of engagement (which completely changes the results) and call it an evolution. You’re misleading companies to believe that what you’re doing is driving the NPS of the future, which is just simply wrong.

      In your new “ebook”, you state that one of the benefits of in-app NPS is that, “they won’t have to try to reconstruct what they were doing in your app three days ago”.

      THIS is the whole point we are making — NPS isn’t an attempt to measure sentiment about what customer was doing in an app “three days ago” … that is a different type of feedback … that is not NPS. Is it useful? Perhaps, but not in the same way. We want the customer to have the time to reflect on all of their experiences as a whole when telling a brand how they feel and why they would/wouldn’t recommend them to others.

      NPS is a tool designed to measure the OVERALL brand relationship with a customer, not their specific product/workflow feedback which could be good/bad/neutral at any point but may have ZERO reflection on their overall sentiment. This is very often the case.

      Again, transactional product feedback can be useful, but it’s not predictive like relationship focused NPS feedback.

      From our own testing along with countless customers who’ve switched to Promoter.io from various in-app solutions, I’m honestly not sure that it even matters what question you ask via pop-up since you are only effectively capturing that “gut impression” you mentioned.

      This is further proven by wildly different scores for those same customers once they’ve run their first email/relationship focused NPS effort.

      We don’t maintain a steadfast focus on email because it’s the “traditional” approach. We do so because it’s the correct approach. This isn’t just some “high road” position, it’s backed by data. Most importantly, it’s a customer-first approach.

      It would be easy to build out an in-app widget, but it would be a disservice to any company looking to us for the most optimal approach to measuring NPS.

      If you want to continue to offer a tool for getting a gut impression from more users by interrupting your customers for transactional product feedback, by all means have at it. In the meantime, we’ll continue to build the future of customer intelligence underpinned by relationship NPS data utilizing the proven process the creator of NPS, Fred Reichheld (who is also an investor in Promoter.io), has created.

      I’ll leave you with a quote from a customer who recently switched to Promoter.io and dropped their in-app efforts:

      “Our NPS is not predictive to churn at all and never really tells us anything. A lot of false positives and negatives and no trends. And hardly anyone fills it out.”

      This isn’t about the number of clicks on a numerical score, it’s about the quality and consistency of the verbatim feedback as it relates to the customer relationship.

      1. Hi Chad, your bias is showing :).

        In-app input is much much much better at scale for software experiences. The reason? Representative answers and broad customer interaction. This isn’t just true for NPS. It’s true for marketing campaigns, CS research, product experiments, etc.

        Email delivery and open rates are plummeting. This isn’t something you can change.

        By continuing to advocate marriage to a channel where customers aren’t listening and responding, you’re advising your customers to purposely be deaf to the data (and assume that they won’t hear from the majority of their customers). I appreciate that you are in a tough spot, but it’s a real shame that you’re pushing an agenda out of line with the results that are required: representative input from your customers (all of them, ideally). Email has its uses, but when it comes to defending it as a channel that is “best” it’s out of line and I think your characterization of Wootric or anything else that is in-app as “fundamentally flawed” really de-legitimizes your own arguments for some of the strengths of the channel.

        As an observer of your arguments online in other venues, it’s a real shame that you are so defensive when Jessica comments and accuse her and her team of “hijacking” a conversation. A focus on customers and their voices would be something much more beneficial to both of your companies and the market as a whole.

        1. Hi Robi,

          Thank you for jumping into the discussion!

          Please don’t misinterpret what I’m saying, in-app feedback has its place. In fact, we use it ourselves for lead-gen, product announcements, support, etc. When it comes to quick-hit messaging and a customer needing support in a quick and efficient manner, in-app messaging is quite convenient.

          We don’t contend that in-app isn’t convenient, nor do we contend that in-app surveys don’t drive results. It’s the type of results that it drives that disqualifies it from being an effective channel to measure NPS specifically.

          Quite frankly your claim that email is a dying channel is simply incorrect.

          Anecdotally I can tell you that our customers average a response rate of between 30 – 40% (that’s response rate, not open rate). More importantly, nearly 70% of those that respond are leaving verbatim feedback. And most importantly, the feedback they’re leaving is based on their OVERALL brand experience, not just what they’re doing in the app at the time of the survey.

          Capturing verbatim feedback is the critical factor here and based on all the importing of NPS data we do from other platforms (primarily in-app solutions), there is very little captured to begin with.

          As I stated previously, transactional product feedback driven by in-app surveys can be useful, but it’s not predictive like relationship-based NPS feedback because it’s only focused on what the customers is doing “in the moment”. This is rarely indicative of their entire relationship with a brand.

          If you want more than our own internal data, there are plenty of other studies that support email as a customer’s preferred channel of communication. Marketing Sherpa found that over 70% of customers prefer email over any other form (https://www.marketingsherpa.com/article/case-study/customer-communication-by-channel). Direct Marketing Association came up with similar results, finding that 64% preferred email to only 24% that prefer in-app (http://raconteur.net/infographics/the-value-of-customer-loyalty).

          So, when you say that email delivery and open rates are plummeting and customers aren’t listening or responding, I have no idea what you’re talking about. It simply doesn’t align with reality.

          Regardless of whether you agree with email as an effective delivery channel or not, the point is the results driven by one over the other. In-app is transactional, email is relational. NPS is a relationship-based methodology. All of the predictive, trending and intelligent benefits associated to the program are thrown out the window with transactional feedback.

          Again, it may be useful information, but it is not NPS.

          P.S. I am not defensive about Jessica’s comments here and I’m always happy to participate in the discussion. As you mentioned, the focus on customers and how to provide better insight and actionable intelligence to brands is beneficial to us both, I couldn’t agree more.

          My only contention is when my content is chased down and responded to with the same response repeatedly without any data to support it. Jessica and her team are welcome to share their narrative and how they approach customer engagement, I just don’t intend to try and pick that apart on their blog or every place they choose to post content. It’s bad form in my opinion, that’s all.

  2. I also completely disagree. We had email surveys for about a year before moving to in-app surveys for our end-users, which we’ve been doing for the last 2 years. We’ve experienced improved response rates and are able to move much faster with the feedback we get from customers in-app. We’ve also heard from customers that they prefer the in-app surveys because they are already in our product using the product when they get the survey as opposed to a random email when they are not even thinking about our product.

Leave a Reply

Your email address will not be published. Required fields are marked *

Share This