De complete handleiding voor het bouwen van een kader voor klantentimatie [deel 2]

Dit is deel 2 van een reeks van 3 delen. Kom volgende week terug voor deel 3!

Vorige week bespraken we de eerste essentiële stap van het Customer Empathy Framework, met de gebruiker, deze week gaan we de volgende 4 stappen behandelen. Dus laten we doorgaan!

Stap 2: Begrijp uw vroege gebruikers

Leer zoveel mogelijk van early adopters.

Early adopters zijn uw eerste potentiële toekomstige pleitbezorgers, dus zorg voor hen, praat met ze en leer zoveel mogelijk van hen.

But be careful not to spend all your time talking to these users. This is tempting, self reassuring and ego satisfying. These people like your product, and it’s natural to enjoy talking to them all the time. But they are already committed and since you might have 90% of your other users coming in and then leaving your product on Day 1, you definitely want to talk to them, too.

But if some people are using your product consistently, you can learn a lot from them that could help you nail activation of new users. In Hacking Growth, Sean Ellis and Morgan Brown go through the following questions that you should ask yourself about your most engaged users:

1. 1What features do they use most actively?
2. 2What features do they use in the first week (7 days)?
3. 3What’s the average retention (return use) of an engaged user?
4. 4Which platform do they use? Desktop? Mobile?
5. 5How often do they use the app?
6. 6What time of day do they use the app and on which days?

Then you can look into further details:

1. 1What sources were they acquired from?
2. 2Was it an ad, a promotional email to the chain’s customer base, or some other place?
3. 3What is their demographic background, including age, income, and more?
4. 4Where do they live?
5. 5How many of them are paid vs free users?
6. 6Which other (competitive) product do they use?

What you are after here are patterns in these engaged users. Are they all using a particular feature? Do most of them come from a specific channel? What are the common traits of these engaged users?

How do you turn more users into engaged users?

If possible, try to get some help from a data scientist who can make the data speak. By clustering your most engaged users’ actions or attributes into groups, you can find what the most engaged users tend to do more than others and when. Are they sharing the document with a colleague in the first week? If so, maybe you want to make that part of the onboarding to see if such an action helps your users activate. It’s probably worth running as an experiment.

That’s for the quantitative side of things. Now, for the qualitative, talk to them. This doesn’t mean just surveying them. I really mean talking.

Meet with them on Skype or at a conference, invite them on Slack to chat with the team or invite them to your office to meet the team. There is a big difference between user interview notes you are sharing through Slack or email, and what the team hears as live feedback from real users. This is because the social and human aspects of the conversations are important. While you might forget your notes, you are less likely to forget that live conversation.

Now that you have a few active users you are talking to, it’s time to start a community. You might be thinking, “Well, that’s a small group" but that’s fine. Through word of mouth these users can bring more people and there is nothing better than word of mouth.

Step 3: Seed a community

Start seeding a community of advocates.

Third, you want to seed a community. On Spark Post, the last product I led, we very early on treated social media as a key qualitative channel. Within hours we were able to get the pulse on an idea, survey our community to validate some hypotheses, and test a new beta we were working on.

We leveraged social as a two way street.

With all products I have built at Adobe, I have always treated social media as a two-way street. Social media is not TV, so give a voice to your followers and implicate them. This is another great source of qualitative feedback from real world users for your team. Notice that we went from early adopters to social media. We are now scaling things a bit.

Investing time in building a community is critical for multiple reasons. First, as mentioned earlier, it becomes a new qualitative channel for your team, Second, it drives word of mouth for acquisition.

Community drives word of mouth which drives sustainable long term growth.

It is hard to beat word of mouth in terms of conversion. If a friend tells you to install an app she loves, you will install it, period. So leverage your most passionate users to seed your community.

To achieve this at scale in Spark, we started selecting hand-picked users from social media. We found people who loved our product on social media and we invited them to our Facebook group. Makes sense, right?

The thing we discovered quickly is that someone sending you kudos on Day 1 does not mean they actually really use your product actively. So our group was not really active because while we had selected enthusiastic users of the product, they were not really our fans. So we changed our strategy. We elected people to the group based on their behavior in the product.

When we detected a very active user (based on key actions we considered indicative of high engagement), we elevated them to the Spark Insider VIP status:

If a user is highly engaged, we give them a chance to become VIPs.

Notice that we have a dedicated logo here for this community to indicate that it is something special. You just unlocked access to the VIP level and who doesn’t want to be part of such a special group?

Congrats, you made it to VIP.

As expected, the conversion here was around 46%. Right after signing up for it, our Spark Insiders got the following email:

The Welcome email for the Spark Insiders VIP status.

The Facebook group we invited them to is a secret group. Here, we keep providing a VIP experience. Since they are our most active users, they are precious and we want to provide a “white glove" red carpet experience.

Quickly, we realized that selecting users based on their behavior was the right thing to do. We started seeing awesome organic conversations and were able to elect some of the most active members to moderators status in order to reward them and allow us to scale things even more:

One of our Spark Insiders Ambassadors, Tami engaging with the community.

Instead of surveying or paying a research group to come up with a way to describe our product to inform our landing pages and go-to market strategy, we asked our most active users how they would describe the product to their friends. The answers we got were some of the best tag-lines we heard to describe Spark:

Our Spark Insiders helping us with marketing copy.

Designers from the team would also weigh in and ask questions to do qualitative research autonomously to inform future experiments:

Our designers, asking for feedback.

If a new beta was available we wanted them to test, it was simple to get them involved:

An easy way to get beta builds tested by our Spark Insiders.

One question we had all along was: Would this VIP status drive any engagement? So we kept a holdout group of engaged users that we never elected to VIP to see if being part of the group drove engagement. What’s your guess?

We saw a net increase of engagement once people made it to the group.

Why? Our hypothesis is that being part of a community increases your confidence, inspiration, and, as a result, your engagement.

Technically, we decided to leverage Facebook Groups as a platform so that we had nothing to build and this was a natural place to be, given that:

1. 1Spark is a product that allows you to create marketing collateral, so Facebook just made sense, as it is a platform where our users would post their content.
2. 2All the members of our team were on Facebook, so every day, in our newsfeeds, we would see the content and conversations from our community and questions they have. This allowed anyone on the team from engineers to PMs to respond to comments and loop members of the team for any issues or questions that needed addressing.
3. 3With the recent Facebook newsfeed changes announced, this turned out to be the right call. Today, Facebook Groups allow us to always reach our community and keep our Facebook presence and engagement high.

One last benefit was that, through Facebook Groups Insights, we could learn more about our most engaged users. Through that, we noticed that we had quite a lot of Spanish and Portuguese members, which was insightful to driving prioritization for localization of the product:

Facebook Group Insights provides useful details about your community.

One last thing we did was to create a Lookalike audience with these emails in order to optimize our Facebook campaigns and reduce our CAC (Cost of Acquisition). As a result, the Spark Insiders was useful not only as a channel for qualitative purposes, but also in optimizing our acquisition costs.

So now, you’ve got your community, but remember what we mentioned earlier… don’t forget about another key group: your churned users. The vast majority of your users might be coming in every day and not sticking around and you definitely want to find out why.

Step 4: Understand your churned users

Understand what’s not working for some users.

Fourth, you want to understand your churned users. If they tried your product and left, you should do whatever it takes to talk to them. Your most engaged users are already sticking around and so there is no urgency, but these people are leaving today and probably forever. So take action now.

Initially, you might need to do this manually, but soon you’ll want to automate this process. In Spark, just like for the Spark Insiders VIP status, we made it behavioral. We detected lapsed users who had become dormant and asked them a simple question: What could we have done better?

Automate the process.

Of course, you’ll get some bounces or people who want to be left alone, but you’ll also get many people who will spend a long time explaining in detail what did not work for them. For us, every single answer was posted on a dedicated Slack channel automatically and then the team could read the comments at any time.

Again, that stream of feedback coming in every day and in front of everyone is what sparks conversations and makes everyone aware of weaknesses in the product and opportunities ahead:

Lapsed user feedback coming in Slack, visible to the entire team.

As always, you should reach out to these users and let them know you read their feedback. Truth is, we have not done that enough with Spark, and that’s something we should start doing more. They might not use your product again immediately, but you will definitely gain their respect and trust by telling them you read their comments.

Once you feel like your product is now better for them, reach out again and encourage them to give it a second try. Now, remember this, if you are asking these users to try it again, you’d better nail it this time and give them a good reason to stick around. In order to nail this, you want to make sure you know:

1. 1Where these people left off?
2. 2How long have they been dormant for?

Based on that, you can segment these users and come up with different reactivation campaigns. If a user stopped using your product after Day 1 and never activated, the way to win her back will be different than if she used it for 2 months actively, then stopped. For these types of reactivation campaigns, the longer you wait, the less likely people are to give you another chance.

That’s why it’s best to automate this and experiment and reach out as soon as you detect signs of churn. Ideally, if you have the chance to have a data science team, you want to leverage predictive modeling that can detect signs of a future churned user and types of incentives for each segment that will be most likely to succeed.

Step 5: Collect Net Promoter Score

Use NPS as a pulse check on how people feel about your product.

Fifth, start collecting your Net Promoter Score (NPS). I like to think of NPS as an ongoing pulse check on the product. The idea is simple. Every quarter (that’s the cadence we use), ask your users the following question:

How likely are you to refer {your product} to a friend?

Users provide a likelihood score from 0 to 10, and, based on that score, you can start identifying segments of detractors, passives and promoters:

Based on the score submitted, users can be detractors, passives or promoters.

When users submit their score, they can also provide a comment. This can be quite useful as users generally provide additional context on why they submitted that score. Now, there are lots of debates online about the pros and cons of NPS. The issue lies in thinking that NPS is a silver bullet and can be the one metric to rule them all. It isn’t.

It is just another variable in the customer empathy framework, another channel to keep an eye on to detect symptoms that might be worth looking into before it’s too late. For NPS, as with our information from churned users, we have automated the process so that the feedback goes inside a dedicated Slack channel.

Thanks to the thread feature, we can tag any PM from the team when we need to follow up or draw attention on a specific comment:

We loop members of the team in on comments worth following up on.

The ultimate goal with NPS is to move detractors to promoters. NPS should lead to an action. So what we do in Spark is we follow up with people by reaching out personally in order to learn more about what frustrates a particular user, so we can make the product better and hopefully turn this user into a promoter in the future. Here is the type of email we‘d send:

Example email we would send to detractors to understand what we can do better.

A lot of detractors will be happily surprised that you are following up with them, and while it may not turn them into users right away, from a trust and brand perspective, you are winning points. Now, honestly, we have not been looking enough into this yet. We read, respond, and track our average score, but we have not been tracking how often we are turning detractors into promoters in a quantifiable way. This is an area of improvement for us.