Monthly Archives: Aug 2012

Customer Relationships on the Web are Broken (and why CRM might be the answer)

The relationship between web and mobile apps and their users should be fundamentally different than traditional customer relationships in two ways:

  • Companies can see how customers are using their product at all times
  • They are getting paid continually over time – either via a subscription, ongoing purchases or ad revenue

In theory, this should make for an extremely satisfying customer relationship – companies are strongly incentivized to make users happy (because financial gain is directly related to how long customers stick around) AND companies have an abundance of data to help make customers happy.

The Reality of how Customers are treated on the Web

However, in reality, this relationship doesn’t play out as well as it could. Instead, we’ve defaulted to a world where, despite knowing exactly what each individual is doing, we treat nearly everyone the same. This problem has a few direct outcomes:

  1. Lots of people get lost between sign-up and becoming real “users”.  Whether they get stuck or distracted, too often there is no re-engagement by companies to bring people back afterwards.
  2. A huge number of features go undiscovered and unused. Because feature marketing is rarely tailored based on user needs or past behavior, it’s both less effective and plays to the lowest common denominator.
  3. Non-product user experience is poor. The majority of communications I get from web apps I use aren’t particularly helpful or communications are just generally non-existent.

Overall, businesses and customers have weaker relationships on the web. While we don’t want all of our relationships to be personal, may of the richest relationships we have with businesses are. The way customers are treated in Apple’s retail stores is a great example.

Why is Everyone treated the Same? 

I’d propose two challenges to building strong customer relationships on the web via personalization (in addition to the lack of face to face interaction):

  1. Most web companies don’t actually have an easy way to see what a given person has done. Most of the tools we have to collect data on our users (Google Analytics, In-house systems, Mixpanel, KISSMetrics, etc) are focused on high level experience changes (funnel optimization, site redesign, etc) rather than on an individual’s actions. Furthermore, these systems aren’t linked to the other sources of customer info (email services, help desks, etc). Want to understand who’s in day 7 of a trial, hasn’t used feature Y, and also has been ignoring the emails you send? You’ll be cobbling together different systems and combining everything in Excel.
  2. Taking personalized action requires too much complexity and work. Even if we could easily understand a complete customer profile, we need to translate that knowledge into action. To do this, we need to be able to figure out want we want to do to each customer at each stage in their lifecycle, based on what they’ve done, and then automatically reach out.

Why CRM for the Web might be the Answer

If you are building a web app and really nail a way to build strong and happy customer relationships through personalization, you have a unique advantage over other companies. Whether you build personalization into your product or into your other customer interactions (emails, support,  etc), the challenge is providing a personal user experience without drowning in money and time investment.

While there are plenty of ways to go about solving this problem for your business, this challenge is why we’ve built Klaviyo – to help you build stronger and happier customer relationships that are good for everyone involved. We think most current CRM systems do a great job of managing the sales-cycle – but generally don’t focus on helping you truly manage relationships with existing customers. If your product is online, the idea of a real “customer relationship management” system that doesn’t capture usage data just doesn’t make sense.

CRM for the web must be different – less data entry, more integration with other data sources, and a tight link with tools that can take action.

Leave your thoughts below and give Klaviyo a shot to see how we can help your business.

Read More
A Blueprint for New User Onboarding on the Web

Over the last 6 months, we’ve spoken to close to 100 software companies about how their new user onboarding process works. These companies covered the entire spectrum – from massive enterprise software firms to consumer internet startups. The goal of this post is to lay out what we learned about the process of onboarding (as compared to the product side), with a particular focus on the emails that we consistently saw companies sending that drove real impact.

Our hope is that you can use the list of emails below as a thought-starter (and then tinker with days / triggers / etc based on your application). Here goes:

Day 1

The Welcome Email

  • Trigger: Someone Signs Up
  • What: This is probably obvious – but make sure people know what to do next, and try to establish a relationship with them.

Day 3

The Come Back Email

  • Trigger: The New Sign-up hasn’t returned
  • What: Not everyone hits the ground running immediately. The goal of this email is to pull back in customers who signed up but didn’t actually get started.  For us, this email is short and personal, but also contains a simple statement of how we think our solution can add value. Many of the companies we talked to sent 2-3 of these emails over the first couple of weeks if users still hadn’t come back.

The Offer to Help Email

  • Trigger: The New Sign-up has returned, but hasn’t done a key action that’s part of getting started.
  • What: For most companies we talked to, there are a few things that users need to do before they can really get value out of your app. This might be adding data, uploading a profile picture, connecting with others, bidding on a job, etc. The goal is catch these issues before the user gives up entirely.

Days 3-60

The Next Step Emails

  • Trigger: Users are active but there are still features they haven’t used, setup they haven’t completed, etc
  • What: The goal is to give users a next step to take with your software that is based directly on something they haven’t done, while outlining the reasons they want to do it. By making this pathway (and the value) both clear and personal, your chance of getting them to do it is much higher – and ultimately it helps both of you out.

Day 45

The Feedback Email

  • Trigger: User signed up but never converted to an active or paying user
  • What: The stated goal of this email is to see if you can learn anything from users who didn’t see value in your service by asking them for feedback and giving them something in return.  The unstated and less obvious goal is to re-engage users and get them using your product again. Many of the companies we talked to found that users who might not have been ready initially responded to this follow-up email by re-engaging and ultimately becoming loyal users.


Based on these conversations, two high level lessons that carried through all of our conversations about onboarding process:

  • Always give users a clear next step. Letting users know what they can do next to get more value from your software gives them a clear and easy path forward.
  • The perfect can be the enemy of the good. Many of the companies we talked to had put a lot of effort into talking about how to do this, but hadn’t gotten around to sending a single email.

Looking to get started with improving your onboarding process? Try out one of Klaviyo’s onboarding recipes to set up automated emails and to measure the impact.

Read More
Using Klaviyo for User Tracking and Customer Development

In the process of building Klaviyo, we’ve been leveraging Klaviyo itself extensively in our customer development process – specifically to understand how our customers are getting value from our product so that we can rapidly iterate on both the product itself and on our marketing and customer service.  While our software makes this easier, the underlying process is relevant to any company trying to learn as much as possible from its users.

At its core, our customer development process is rooted in one important fact:

What people actually do is far more important than anything they say.

Understand who is getting Value

Our starting point is to understand the customers who are regularly using the core features of our product that we believe directly drive value for our customers. These customers can provide the anecdotes that help us focus our development / sales / marketing energies on the right problems and companies – in short, they tell us what we got right.

For us, this actually involves looking at a few different groups in our system each week:

  • Users who login very frequently
  • Users who actively create groups of customers
  • Users who have sent email to a group of their customers through our system

Analyze Customers who aren’t and develop Hypotheses

Equally, if not more, important to successful users is understanding customers who aren’t getting value. For us, we limit this group to be customers who have sent data (i.e. they are at least somewhat actively engaged in using our software), aren’t on our team, and haven’t been active in the last 7 days.

This list makes a great starting point for understanding what isn’t working – for instance:

  • Our value proposition isn’t high or clear enough
  • Our product isn’t valuable enough to this type of company
  • Our product doesn’t solve the problem
  • Our process (training, onboarding, etc) failed them
  • Etc – and likely part of all of the above.

Once we have this list, we use it for targeted outreach to better understand our users and the problems they are trying to solve.  It’s worth revisiting the point we started with: no matter what people may tell you, having them use your product to generate value is what’s most important. By diving deeper on a few specific use cases (and having conversations about their business and problems), we develop ideas about ways to iterate on the product – and judge our results by whether they add value to our clients.

Analyze Key Features by User and in Aggregate

In addition to focusing on individual users, we also focus feature by feature to understand overall usage and who is specifically using or not using a given feature. Doing this once a week keeps us in touch with what people are doing, especially as we’re thinking through future development priorities.

Lessons We’ve Learned

Distilling the above, we’ve learned several valuable lessons about why to focus our attention on individual behaviors and segmentations compared to just using aggregated stats:

  • Evaluating your performance by what people do gives you the most realistic picture of how you are doing
  • Understanding what defines the most successful users gives you powerful clues to where you might want to go in the future
  • Identifying hypotheses about why certain people aren’t using your product gives you tangible actions to test

Listen to the user – both what they say AND what they do.

Interested in seeing how Klaviyo can help your customer development?  Get started with a trial (and if you’re a startup and are one of the first ten people to email me at, I’ll give you a discount code for half off on the monthly fee of the basic plan).

Read More
Your Free Trial is a Conversation with Me – Don’t Ignore Me.

Say I sign-up for the 14 day free trial of your web app, get halfway through installation, get frustrated, and then never return.  Will I hear from you? What will you say to me? Do you have a plan for converting me from a free trial user to a full-on paying customer?

This should be a high point – you’ve convinced me to try out what you’ve built, and all you need to do is turn me loose on your excellent product and it’ll immediately start improving my life and I’ll convert to a paying customer right away. Awesome for both of us right?

The reality is that it isn’t that simple. I get distracted. I can’t figure things out. I set something up wrong.  I don’t get the value of your product.

Whatever the reason, the question is how you treat me during this process – I mean, we just met after all, and if I like you, I’ll give you money and introduce you to my friends.  If you ignore me, I’ll think you are a jerk.

Tangibly, here’s how I’d love to be treated during a free trial:

  • Welcome me to your product in a way that’s personal yet useful.
  • If I get stuck completing the basic setup, send me resources or offers to help.
  • If I get up and running but I’m not using some of your best features, email me to show me the way – but don’t email me to tell me to use things I’m already using.
  • If I have a bad experience (poor analysis results, I lose every game, I hit lots of bugs, etc), then explain to me why this won’t always be the case in an email that encourages me to come back.
  • Don’t make all of your emails one-sided.  If every email you send me is titled “Your trial is ending in x days” and the content tells me nothing other than how to pay, I’m not particularly likely to start using your product.

My interactions with you tell you a lot about me and how I currently perceive your product, and it doesn’t make any sense that you’d ignore them. Listen to me, make me happy, and I’ll gladly give you my money when you are adding value to my day.

Read More