Using NPS to improve overall customer sentiment

David Ellin

I've used NPS for years. At first, I wasn't a strong proponent and then I realized I was using the data all wrong.

While I didn't have too many detractors, I was eager to move detractors way up to become promoters. Silly me. That would be like moving mountains.

I finally learned that when you roll pebbles, they gain momentum and become boulders. That's what I started doing with NPS.

The goal had to be moving a detractor to a passive before moving to a promoter so I started focusing on the break points. If I could move a customer from a 6 (highest detractor score) to a 7 (lowest passive score), I impact my NPS. That's just a single point. Additionally, if I moved an 8 (highest passive) to 9 (lowest promoter), I've impacted NPS even more. That's also just a single point. Move a few customers up a single point and you have major shifts in your NPS score.

What's your strategy for moving the NPS needle?

Jay Nathan

This is great @David L Ellin. My strategy is to start by maximizing participation. The people who are not responding to feedback requests are 3-14x more likely to cancel than detractors. So that’s where the real risk (and REAL good feedback) lies. 

whenever I share feedback metrics I always share participation rates and the goals on the dashboard.

Statistics courtesy of @Steve Bernstein 

David Ellin

Participation is so important, @Jay. True that in penetrating down to the non-responders.

Jeremy Donaldson

@Jay Nathan - Participation is better than observation! I agree that if you aren't getting a response that is the first step. Most companies target 20% response as the target for good feedback. If your organization is below 20%, focus on responses first. 

As for scores, the unfortunate reality with NPS is it is subjective to the day that the individual completes the survey. If the poop hits the fan the day before the NPS goes out, you can forget about getting a high score. Likewise, if someone has an incredible day before they take the NPS, the chances of a high score increase. I have customer that rate us a 7, but act in a customer reference capacity every month. 

The second step would be to conduct a deep dive into the feedback provided. If a customer consistently gives the organization an 8, the amount effort spent to get them to a 9 or 10 may be in vane if their philosophy is to never rate someone that high. Likewise, someone who gives a 0 may jump to a 5, 6, or even a 7 with a relatively small amount of effort.

Lastly, consistent follow up and perhaps a resurvey of the customer 30 or 60 days post issue resolution to get a more accurate temperature of improvement in the account. If the effort doesn't lead to better results, then you may want to rethink how you engage with the customer.

Ramya Ragavan

Speaking of an issue where you don't get any responses, I am noticing that several of our client's email server has blocked these surveys. We use Pardot (Salesforce marketing tool) to send out surveys but these are blocked by client's internal IT. 

When I get on a call with a customer, I ask them to check if they received our survey - 

1. If they did, but didn't respond I ask them to take 1 min on the call to fill out the survey and let them know that this information is being reviewed at our leadership level and it is their chance to voice out their opinion. I also mention that we review their responses, and we look at our "call to action" 

2. if they did not receive it, I ask them to touch base with their IT team, and also offer to connect our IT resource to theirs so we can sort it out. 

Would be interested to know how others are tackling this - is it lack of response, or is it not receiving the survey? What tools do others use for surveying - is one better than the other?

Jeremy Donaldson

Interesting take conducting live NPS surveys. I have a tendency to want the feedback to be provided via a source that does not have my input or intervention because the customer may skew results based on my presence. Does the “live” surveys change the results from your experience?

We use SFDC as well, but if you ever have the opportunity to send the survey via a non automated Templar I’ve found less surveys get blocks by IT.

The process I follow is similar to what’s below:

1 month out: Email or call with the customer to example the upcoming survey and get the correct contacts aligned. Also, request IT whitelist the email address.

1-2 week out: Email or call the customer stating the date, time, email address, and email subject line of the survey. I explain the survey may go to SPAM. 

Day before: Send a reminder email

Day after: Send a follow up to non-completed survey responders or a thank you email for completing the survey

1 week after or next call: remind the non responders of the survey

30 days: last notice of survey before we close it

David Ellin

All good points, @Jeremy Donaldson.

David Ellin

@Jeremy Donaldson is spot on. It's important to get the survey email address white listed via IT. Your main customer contact should be able to help you get that done.

Pre-survey reminders ("heads up...we're asking for your feedback...") and non-responder reminders are valuable. I've found that most customers will wait until the last day to respond. My emails suggest that customers complete the survey upon receipt so they can get it off their plate and not worry about it later.

Lastly, remember to include a 'thank you' email once the survey is completed. Remind the customer how important their timely feedback is and how your company is going to use it.

Marijn Verdult

@Jeremy Donaldson @Jay Nathan Love the ideas, what are your practical tips for improving the participation?

For our company, the NPS score is amazing but I keep telling our Leadership that currently it's just a vanity metric and is not a great indicator (or KPI as was suggested)

David Ellin

@Marijn Verdult, improving participation takes effort on any number of fronts. Here are some ways I've been successful in the past:

1. Review pre-survey messaging for a strong call to action and indicate why it's so important for customers to respond.
2. During every QBR/EBR/SBR, remind customers about the importance of responding (if they haven't responded and personally request their participation). If they have responded, show their trending metrics and continue to reinforce the importance of future participation.
3. Mid-survey reminders - we typically keep a survey open for 21-28 days. We send reminders to non-respondents on days 8, 16 and 2 days before closure with an immediate call to action (IE: don't procrastinate...there's no time like the present to have your voice heard)
4. Send an automated 'Thank You' to everyone who has responded. It's important to show appreciation for their participation.
5. Personal outreach - I've called customers to specifically ask why they haven't responded. It's usually, 'I forgot', 'I'll do it later' or something like that. I once had a customer tell me it's against their company's policy to complete surveys.
6. Make sure you get your survey-from email address white listed in customer email systems. That may take gaining access to a system administrator. Start with your champion and go from there.

AND MOST IMPORTANTLY, follow up on feedback. Taking action on feedback is the best way to guarantee future participation. Neglecting feedback is the best way to guarantee diminishing response rates.

What other techniques have others used to increase participation rate @Jay Nathan @Jeremy Donaldson @Jeff Breunsbach ?

Marijn Verdult

Great points @David L Ellin !

Do you target all users individually as well? For non respondent end-users, do you ask the coach/champion/manager to "chase" them?

Jeremy Donaldson

@Marijn Verdult We are averaging right at 20% response rate as a company. One practical change we implemented as changing the subject line of the survey to include the surveyee's name: "Marijn, Tell us how we are doing" or "David, How are we doing?"

The other thing you could consider is putting the 0 - 10 rating scale in the body of the email so all they have to do is hit the number rating they want to select and it auto-populates that score in the survey. 

David Ellin

Great question, @Marijn Verdult. If we survey multiple people at a company where some participate and others do not, we target the individual users. If none of the folks at a customer respond, we'll approach the champion, executive sponsor, etc. to pinpoint a reason and appeal to them to have their folks respond.

Marijn Verdult

@Jeremy Donaldson  do you have challenges with delivery of the emails when putting the rating in the email? 

As soon as we put any formatting into an email our delivery rate plummets due to the firewalls of our customers (some of the largest companies out there) 

Jeremy Donaldson

@David L Ellin You are absolutely correct. It takes a multi-prong approach with most customers to get feedback.

I agree with everything above. The one minor difference is sending personalized thank you notes calling out specific feedback they provided. Why? It lets them know we received it, read it, internalized it, and wanted to call specific attention to it. 

Also, leadership in some form engages with every response received - we have a company wide slack channel dedicated just to NPS responses and we leverage Gainsight CTAs for positive and negative responders.  

Also, we take a multi-contact approach for feedback - we like to survey power users to executives to get a holistic view of the account not just the admin pushing the controls. This does reduce the response % per account, but we get a deeper level a feedback overall. 

David Ellin

I forgot a big one that I'll go back and add to my prior post. This will address your question. Make certain you get your survey-from email address white listed with all of your customers so they don't go to spam or get held up due to formatting.

Jeremy Donaldson

@Marijn Verdult Yes, we have tried it both ways. We are currently back on doing the survey without the rating in the body of the email.

The customer must whitelist the email address if the rating is put in the email body. 

Ramya Ragavan

@Jeremy Donaldson - No, and that's what is interesting. Customers that I talked to over the phone has still given us the score they think we deserve as a company. I clearly mention to them that this is about the company - our products, and services and their experiences with it, and NOT about an individual. Again, this is our very first time reaching out to a broader group via an automated tool. In the past, we have collected NPS during user group sessions, when they had several folks from our team in the same room  

But, I do agree - there is a possibility that results might be skewed - for us it is more important the first time around to collect as many responses as we can, and when we can, while we learn which customers have not even received the survey. 

I liked your idea of tell them, tell them what you told them...reaching out ahead of time, reminding again when it is close to survey time, and following up to make sure they know there is an end date to respond. I am going to take this suggestion, and follow

Jackye Clayton

@Joyce Wells I thought you may find the above conversation interesting!

Manisha Khandelwal

Hi @Jeremy Donaldson in the below statement, what do you mean by 'non automated Templar'?

"a non automated Templar I’ve found less surveys get blocks by IT."

Joyce Wells

Thank you @Jackye Clayton This is a great conversation! I just completed our first NPS survey for Sierra Interactive. I tailored the NPS question just a little toward the industry which is Real Estate technology and I used this new customer inteview software that allowed me to send out the survey to all of our customers and then to resend the email if they had not responded in 5 days with a new subject line urging them to participate. After clients rated us we provided a comment box if they wanted to provide more feedback and I also included a link for customers who were interested to schedule a call with me. I had almost a 20% response rate with a score of 37, of that group 10% left comments that I am currently sorting through and bucketing them into product, support, etc. and another 5% requested to have a call with me in which I was able to document the notes from my call all within this tool. Interesting enough a few of the calls were from customers who loved us and just wanted to tell us that. Another call from a client that was dealing with an issue led to me actually hiring him on my team. He stated that he was moving cross country and leaving his company. He was the trainer for our software in his company and I was looking for someone with his skillset so it was a win for both of us. Had it not been for him requesting a call this opportunity would not have surfaced. I was also able to send out a thank you email to all that responded. Currently I am working on creating follow ups with detractors and nonresponders. Does anyone have any creative ideas for outreach to these 2 groups? As we continue to do customer outreach using this interviewing software I have NPS as a trending question so we can continue to survey customers who did not respond ongoing as we are talking with them whether it is during a QBR or another outreach project we might have set up. I now have the reporting capability to be able to see our results year after year and can easily share out with the leadership team. I am pretty happy with our results seeing this is the first time we have surveyed our customers in this capacity but this is just the start of my work with getting some true customer feedback. So this thread is really timely for me.

David Ellin

@Joyce Wells, congratulations on getting your feet wet! It sounds like you’ve taken a ‘crawl before you walk’ process which is great. Learn as you go and continue to build out your program.

It’s extremely important to follow up with both detractors and non-responders. If you can’t get to the root cause of why a customer is a detractor, you won’t be able to move them to a Passive or Promoter.

For Detractors: Before reaching out, check with your support team, CSM and anyone else to see if there have been lingering support issues, open tickets, or any other known issues that have not been resolved. The last thing you want to do is reach out and not be informed about things they’ve already complained about. If there are open issues, get the status and if/when the issue can be resolved. If it can’t be resolved, be prepared to explain why.

For Non-responders: start with the same process above because if they’ve complained and haven’t seen action, they won’t feel that responding makes any difference. If there are no open issues, let them know that without their feedback, you’re operating in the blind as to their sentiment. Also let them know that using their “voice” gives them input into your product roadmap and future solutions.

If you survey and don’t have a great action process behind it, customers will stop responding. Remember to thank them for their participation whether there’s specific follow up or not.

Here are a few suggestions to increase your response rate:

1. Make sure you communicate the URL/email address from which the survey will originate. Some companies may have to white list the address. If your survey address is [email protected], it may go through. If it’s from sierrasurvey@[surveycompany].com, it may go to spam or get rejected altogether.
2. Send out a pre-survey email explaining the importance of gaining their timely feedback. The email should reinforce when the survey will come out and what address it’s coming from. Politely request their timely response.
3. Send reminder emails during the time the survey window is open reinforcing the need for their feedback.
4. Be careful to not over-survey customers. Survey fatigue is real.
5. Share survey results with customers during their QBRs and reinforce what you’ve learned, what actions you’ve taken, and thei continue need for future responses.