All things “Customer”

What should go into your customer health score?

~ Don MacLennan

There’s a religious debate going on in the community of Customer Success vendors. My turn to weigh in.

On the one hand, some tools vendors are evangelizing the value of predictive analytics driven by product usage data. Other tools vendors are evangelizing the value of rules-based scoring driven by various factors including usage stats, survey scores, support tickets, and more.

While each vendor’s approach can be “right” under certain circumstances, each is generally wrong to characterize a health score as being driven by one approach versus the other.

My take: You need both.

Before I explain why, let’s get to first principles. Customer Success teams consistently say that they need to get away from “firefighting” mode to proactive customer engagement. Think of the health score as an “Early Warning System” that causes a Customer Success person to engage the customer at the earliest possible sign of caution, though direct contact or even a campaign.

Early warnings can come all along the customer journey, including during product trial, implementation and of course, as a precursor to renewal or retention.

The question then becomes, how to construct a score that satisfies key criteria:

  • Runs continuously so that each customer is monitored all the time.
  • Is suited for each stage of the customer lifecycle.
  • Is suited for each type of customer in your customer base.
  • Is trusted so that your team is inclined to react when a low score is generated.

So back to why neither group of vendors is “right.” In order to satisfy these criteria, you need multiple scoring methods. Some examples of different scoring methods for different lifecycle stages include:

  • During a trial, you know little about a user. The only signal you might have is usage. So, a score for trial users needs to be driven by usage stats or a predictive algorithm to spot the users likely to convert or not.
  • During onboarding & implementation, usage might be sparse. Especially if your product is complex and you deliver implementation services. You might, however, have a lot of support tickets as a signal that indicates if the customer is struggling to configure or adopt the product. Or, you’d look at early usage and tickets together.

Some examples that pertain to different product types and Customer Success models:

  • You service a high-value customer tier. The health score should take into account total adoption of your product in each account and all the facets of your high-touch relationship with the account.
  • You service a customer in a lower-value tier with a low touch service model. For example, one Customer Success Manager to 100+ customers. Usage data and maybe support tickets might be all you have to construct a score that spots at-risk customers.

So are predictive analytics a good thing? Of course. They can make for a better health score. However, let’s be clear about when and how they can useful.

First, the purpose of a predictive algorithm is to spot variances. In this case, your “most healthy” and “least healthy” users. If your product is not well instrumented, and you collect just a few usage event types, then an algorithm will be less reliable because it won’t spot variances in usage patterns. Conversely, usage data that contains more than 15 or more different event types will probably tease out interesting differences between users.

Second, if you don’t have a historical set of usage data collected, the algorithm will be less useful. You’re looking at correlations between usage patterns and outcomes such as churn, retention and renewal. One month’s worth of data won’t cut it.

Last, algorithms are most useful when you don’t have other facets of a customer relationship to rely upon. If you’re a Customer Success Manager with 20 assigned accounts, it’s doubtful you’d be surprised by an unhealthy customer nor would usage alone explain the health of that relationship.

In summary, for many Customer Success teams, customer health scores must take into account usage data plus something else. Also, if you want predictive algorithms to drive your score in part or in whole, be sure to instrument your product fully and build up a historical repository. Last, consider a health score driven by usage data alone only when you don’t have any other data to work from.

Happy scoring!

Looking to de-code churn drivers? Look for hot-spots, not averages

~ Don MacLennan ~ 6 Comments

OK, I admit it. I’m recycling parts of a blog post from 2 years ago. 

The Normal Distribution, or “Bell Curve”

Our brains are wired somehow to think of everything in terms of a Normal Distribution, aka the Bell Curve. It’s a trap that can obscure important patterns in data such as churn drivers.

The shape of the curve means that we think of populations of data (such as customers) as being a somewhat homogeneous group if only we could compute the average. For example, how many minutes per day “on average” a user spends in your product. Or, the percentage of customers “on average” who renew their subscription.

The problem is that populations of people and customers almost never behave in a normal distribution. Instead, the more prevalent pattern of behavior is a Power Law, or Pareto Distribution.

The Pareto Distribution, or “80/20 rule”

The Pareto distribution is also known as the 80/20 rule. Except that in online worlds, the ratio can be even closer to “95/5″. So, it’s very likely that the 5, 10 or 15% of your customers that churn each year actually have a lot in common, and differ materially from the “average” customer.

Let’s de-construct this. Start by asking yourself a few questions.

Is churn the same across every acquisition channel?

Probably not. You’ll see variances between channels such as outside sales, inside sales, online channels, and even within the various online campaigns that can beget customers (email marketing, search engine marketing, display advertising, etc.).

Is churn the same across every customer segment you sell into?

Probably not. If you’re B2B and you sell into various industries, then you probably have some industries with better retention than others. You will also see differences between small, medium and large customers according to their company size (their annual revenue and/or employee count). If you’re B2C and you sell into various demographics, then you probably have some demographic segments with better retention than others (think age, household members, etc.)

Is churn the same across every region you sell into?

Probably not. Different countries have different retention characteristics. Sometimes, even across different states or provinces.

Is churn the same across product usage patterns?

Probably not. There are probably usage patterns such as feature use and frequency that correspond to higher and lower churning customer segments. Also, think about usage across a population of users inside an account. Some accounts will have a couple of very engaged users and the rest no so much. Others may have widespread, evenly distributed usage. You wouldn’t see these differences if you were working with “averages”. Yet, these different segments will probably have different retention rates.

De-coding churn is about the search for “hot-spots”

The game of de-coding churn is about finding variances from the average, instead of focusing on the average itself. There’s probably a high-risk segment in your customer base that can be described by a combination of multiple factors including acquisition channel, region, customer demographics and usage pattern. Think of this example: “churn is higher in the U.S. through the reseller channel by 25%, than the regional average” If you knew that, what would you do? You’d probably have a conversation with the person who recruits and trains re-sellers in the U.S. about doing a better job of it. You just found a “hot-spot”.

So why is the search for hot-spots so hard? It requires all the data about your customers to be in one place and fully dimensionalized (think “data cube”).

For most companies, getting data into this type of structure is done using Pivot Tables, which is time-consuming, error-prone and hard to maintain. This is why Big Data and analytics systems are becoming commonplace.

Happy hunting!

Mining for gold: 3 sources of product usage data

~ Don MacLennan

Update: the vendor landscape described here has evolved considerably since writing this. But the basic concepts still apply.

In a prior blog, I talked about ways to drive revenue growth with product usage data. Including reducing churn, spotting up-sell opportunities, increasing trial conversions etc.

In a follow-up blog, I discussed the four dimensions of product usage that you need to analyze:

  • Frequency: how often a user interacts with your product
  • Features: which features are used
  • Volumetrics: how often a feature is used, or a data record is created
  • Configuration settings: ways a user configures the app to suit their needs

By now you’re probably interested to unlock the value of product usage data for your company. Which begs the question: where do you get this data in the first place?

There are three sources of usage data. Generally speaking, usage data comes from:

  1. Clickstream data
  2. Log events
  3. Database queries

We’ll review each in succession.

Clickstream data

Clickstream data is generated by an end-user’s interaction with your product interface. For example, logging in to a web browser. Or performing an action in your mobile app.

This type of usage data can help you understand frequency of use (usually but not always) and coarse-grained feature usage, depending on how you “tag” usage events from the browser.

Several products are good sources of clickstream data for your browser-based app:

  • KISSmetrics
  • Mixpanel
  • Segment.io, a popular Javascript plug-in that feeds lots of other clickstream tools

An aside: why not Google Analytics for product usage? Unlike these other tools, Google’s terms of service prohibit you from collecting user-specific data of any sort. Thus, it becomes very difficult to understand which user or even company is accessing your application and how. Stick to the products above.

Getting usage data for Mobile apps can come from several of the packages above, plus some that are purpose-built for mobile:

  • Flurry. Note that Flurry monetizes by driving ad placements, so it’s not for everyone
  • Tapstream

Pros of clickstream data

  • Easy to deploy
  • Little to no engineering team involvement
  • Good for basic engagement metrics

Cons of clickstream data

  • Not detailed enough to reveal important features and user segments
  • Brittle to maintain

Log events

Depending on how well your engineers have instrumented your server-side code, they may be generating usage data as log events. For example, your Web or App Servers might be generating Apache logs that contain details about the user’s actions, especially feature usage.

Pros of log data

  • Log events can be very specific and accurate in depicting feature usage, compared to page-level clickstream data from a browser

Cons of log data

  • Log events need to be parsed, which can be challenging if you’re doing it for yourself in a database or Excel file
  • Log events can contain types of events that aren’t meaningful to you, because they describe system behaviors not user behaviors (think error logs)
  • Your engineers need to be involved to do a good job of log instrumentation

Database queries

On the server side of your application is some sort of database (such as MySQL, Hadoop, MongoDB, etc.). Each user action may have a corresponding “transaction” or record in the database that forms a picture of usage. For example, if a user started a new “Project” in your Project Management app, then that record was created in the database at a specific time by a specific user.

These records can be queried from your database to produce events or counts of events (such as daily summaries).

Other functions of your application may behave as “set it and forget it” where your app is automating processes without requiring a user action each time. In this case, usage events are generated even though the user hasn’t logged in lately. Database queries may be the only means to collect this type of usage event.

Pros of database queries

  • The most comprehensive picture of usage

Cons of database queries

  • Your Engineering or Operations team has to retrieve the data for you

Conclusion

There’s no “silver bullet” to getting usage data.

In many companies, you have easy access to one type of usage data and not the others. And, no single source of usage data depicts your application’s usage in a comprehensive way.

Think of it as a journey. It’s best to get started with the data on hand. As you learn to make sense of it, and drive business results, you’re armed with the justification to get other types of usage data. Sometimes this means further instrumenting your product. In other cases, it’s simply about getting another team to help you access their data.

But the journey is well worth the effort, because usage data is the foundation of understanding the health of your customer.

“Net Churn” is a bullshit metric

~ Don MacLennan ~ 4 Comments

TLDR: Certain subscription-based vendors have conceived of a metric called “Net Churn”. The concept seems to have originated in publicly-traded SaaS companies. It’s probably in response to Wall Street analysts’ enduring attempts to decode the underlying measures of financial health in their renewal revenue streams.

Here’s how Net Churn is calculated:

  • You take the renewal revenue pool for a given month; let’s say it’s $1,000,000 of potential revenue
  • You take the actual renewals PLUS actual up-sells/cross-sells that you get from those customers at the same time, PLUS actual cancellations value PLUS down-sells; let’s say these sum to $950,000
  • Divide actual over potential and voila! Net churn rate

By using this math, “Net Churn” can be a number as high as 95% even over 100%.

This is financial chicanery.

An organization should measure each metric separately. Each is a unique and important financial lever. Combining measures into Net Churn blurs how well (or not) you’re doing at each.

For example, in a B2B business with annual subscriptions, consider these metrics:

  • Logo churn – this is the count or rate of customers in the renewal pool who actually renew. The calculated rate is always 100% or less
  • Revenue churn – this is the count or rate of renewal revenue in the pool that was actually renewed. The calculated rate is always 100% or less
  • Cancellations – this is the count or rate of cancellations in the period between renewals
  • Base renewals – this is the count or rate of revenue from customers who renew at the same financial commitment as the prior period
  • Up-sell/cross sell – this is the additional revenue captured beyond the renewal potential
  • Downsell – this is the renewal revenue from customers who lower their financial commitment for the same product for the renewal period versus their prior level (downsell is also referred to as partial churn in some organizations)

Ideally, you measure and optimize on all levers:

  • Logo churn – useful to see churn patterns across your whole customer base
  • Revenue churn – useful to see churn especially from high-value customers
  • Cancellations – a particularly concerning outcome worth understanding in depth
  • Base renewals – useful to see the customers who are steady-state. These customers may end up as future at-risk customers because they aren’t growing
  • Up-sell – useful to see the performance of premium versions of your product offerings (if you have them). Your marketers spend a lot of time tuning your offer hierarchy and this is a great way to validate their efforts
  • Cross-sell – useful to see the performance of complementary products in your portfolio (if you have them). Many vendors diversify their product portfolio, at great expense. The payback is often measured by how well your customer base consumes the new offerings
  • Downsell – useful to see customers who lessen their commitment, which may present clues to overselling license capacity in a prior period

Leaving aside the games that are played with Wall Street analysts, focusing on these metrics individually will quickly highlight several drivers of financial performance. Each metric reflects a unique opportunity to drive revenue growth.

Six best practices for expectations management in Customer Success

~ Don MacLennan ~ 2 Comments

TLDR: The role of Customer Success might be one of the most connected and inter-dependent jobs in your company. In addition to working with your customers’ people, you have lots of internal stakeholders to work with too. Including executives, sales teams, marketing teams, finance teams, product teams and support teams.

With so many people involved and so much communication flying around, the art of expectations management becomes an essential tool for survival and success in the role.  I’d like to share our six best practices from many years of managing many relationships.

1. Consistency & timeliness

  • Regular communications beat sporadic ones. A clue to a problem in your current situation: somebody requests an update that you didn’t otherwise plan for
  • Consistent updates beat variable ones. Suggestion: find out what people want to be kept informed about, and make this the basis of the regular flow of communication

2. The power of “no”

Some of us hate to say no.  Yet it’s powerful when used appropriately:

  • “No” is required to build trust and credibility.  Saying “yes” to everything creates skepticism that you’ll deliver everything, and rightly so
  • If you like to under-promise and over-deliver, “no” is part of how you do it
  • Be careful to say “no” for a reason. Try using “no, because…” to justify your position, and negotiate if needed

3. Timely follow through

The longer somebody waits on us, the more likely they are to believe something’s gone wrong:

  • Any commitment should be met in days, ideally 1-3. Suggestion: if a commitment takes longer, de-construct the deliverable and commit to interim steps that last 1-3 days (e.g an update call to share progress, or an interim deliverable)
  • “Constant improvement beats delayed perfection”

4. Bad news must travel the fastest

Take a cue from the professionals who do crisis communications:

  • Be pro-active and control the message. Bad news + delay = elevated emotions by other parties

5. Active listening

Active listening is a way to assure the other party that they are being heard and understood:

  • Play the message back before responding. Suggestion: a person says something. Instead of responding, first play back what you heard: “what I heard you say was….”. And, use the word “because”. An example: “What I heard you say is that item #2 is most important, because you committed this to a customer”. Once you have confirmation, then respond.

6. Document the plan

Artifacts have a powerful way of aligning expectations.  Feedback is often easier to get by having someone react to something that’s written or visual (versus verbal only):

  • Write down and share information so that expectations and deadlines can be easily referenced later
  • Prior documentation can be excellent way to review progress and get credit for commitments you delivered on

The metrics-driven SaaS business

~ Don MacLennan ~ 1 Comment

I founded Bluenose in part because there are major changes happening across technology-based industries. In this post, I’ll review those changes and what it means for technology vendors. In particular, I’ll discuss the importance of adopting new ways to manage businesses, even as the old ways still hold true.

There’s a business model revolution going on.

Our friends at Zuora coined it the “Subscription Economy®”. It’s fundamentally changing the patterns of consumption:

  • in music content, we’ve gone from buying a CD for $18 to buying a song or subscribing to a monthly streaming service
  • in movie content, we’ve gone from owning the DVD for $49 (ouch!) to on-demand viewing or a monthly streaming service
  • in desktop software, we’ve gone from buying a boxed CD at the computer store to a monthly subscription. And lots of freemium mixed in.
  • in business software, we’ve gone from perpetual licenses with up-front payment to monthly or annual SaaS subscriptions
  • in mobile phones, we’ve gone from long-term contracts to shorter ones, and in many cases pay-as-you-go
  • my favorite example: GE Aircraft Engines used to sell you an engine for a few million dollars. Now, you purchase “flight hours” and your billing is metered accordingly

It seems like everywhere the internet touches our lives, the business model has been transformed in turn.

A new dynamic between customer and supplier

When suppliers got up-front payments for their products, their attitudes toward customers often sucked. You have a software problem? Call tech support, where you’ll be rushed off the phone as soon as they can. After you were on hold for too long.

After all, at that point you as a customer are a cost. Repeat after me: a cost, as in to be minimized.

In recurring revenue models where you consume and pay over time, it’s a new game. The supplier must keep loving you, or you and your revenue stream go away. Study the organizational design of a SaaS businesses, and you will encounter new departments called “customer success”, “customer advocacy”, “customer experience management” or “customer retention”.

What’s going on?

Follow the money. The new shape of revenue is causing a new way of thinking about the customer.

By the way, we as customers are liking this quite a lot. We’re holding the leverage now, and are enjoying the newfound attention as a result. For that reason, I think this model is here to stay. Marc Andreessen said “software is eating the world”. Maybe it’s more like SaaS is eating the world.

Time for new metrics

As suppliers, lots of the old ways to measure success are still there. We still care about revenue, profits, cash flow, access to cheap capital, competition, etc.

But this new business model forces us to master some new metrics in turn, like Lifetime Value (LTV). Or Customer Acquisition Cost (CAC) to LTV ratio. Or Churn Rate. Or Renewal Rate.

The new metrics that I find most interesting are ones rooted in the measure of a customer relationship over time. And, the linkage between relationship health and the associated financial outcomes: renewals, churn, up-sells, cross-sells, etc.

“Subscription Economy”® is the registered trademark of Zuora, Inc.

6 ways to increase revenue with product usage data

~ Don MacLennan ~ 1 Comment

TLDR: Analyzing product usage data can lead to several revenue generating opportunities. While most SaaS companies have some form of usage measurement, few have achieved a complete view of usage.

Let’s review all of the revenue levers at stake.

One: increase revenue by increasing trial conversions

Most trials convert because the user achieves rapid time-to-value, such as minutes or hours from the time you provision their trial account.

The conversion rate drops off significantly when provisioned users don’t actually engage with your product until days into the trial. In fact, most provisioned users who don’t access your application in the first few days probably don’t convert ever. So, usage data gives you an “early warning” about provisioned users who aren’t active in the first hours; you can react by stimulating their access with emails, phone calls and such.

As you master usage analytics, you’ll also discover the features that are most correlated to conversion. You can then engage trial users with a purpose: nurture them to experience the “best” features for conversion.

To summarize, the blueprint for successful conversion is about getting users to achieve rapid time-to-value, and to specifically experience the features that lead to conversion.

Two: increase revenue by spotting up-sell opportunities

Highly engaged users tend to adopt your application faster than others. In fact, the velocity of adoption is a critical measure for your business. These users are open to expansion revenue propositions and should receive attention when exhibiting “high velocity” behavior.

Even in the “average” user segment, progress toward some license limit should be the trigger to capture expansion revenue. Some examples:

  • users who have used 80% of their licensed seats
  • users who have consumed 80% of their capacity (as expressed in data volume or some other metering method)

The other important consideration is to present the up-sell opportunity at the earliest moment of need. Too often, vendors use an artificial event such as a renewal date to propose a higher-priced offering. But what if the customer had a need months ago? Selling at the moment of need is the best way to maximize this revenue stream. Pay attention to velocity and license limits as a result.

Three: reduce churn by spotting declining usage

Just as increasing adoption is a very strong positive signal, decreasing adoption is a strong negative signal. One could even argue that if usage is not growing, trouble lies ahead.

When a customer is at-risk as evidenced in usage data, the time to intervene is now. Whatever is causing customer dissatisfaction needs immediate attention. This “moment of opportunity” shouldn’t be confused with the renewals cycle. Customers can and do “churn” before their contract expires, by ceasing their usage. You’re highly unlikely to receive a renewal if you address their dissatisfaction at renewal time instead of the time they began to disengage.

Four: increase revenue by finding and mobilizing evangelists

Evangelists are a far more effective means to acquire new customers than any other means thanks to the power of recommendation and social proof. The challenge is that evangelists are often hard to spot. You have to find, enroll and activate them instead of waiting for them to naturally behave this way.

Usage data is a goldmine here. The heaviest adopters, and users of your most differentiated features, are hidden in your usage data. The programs you have to activate evangelists are super-charged when you’re identifying all of the potential recruits in usage data.

Five: increase revenue by improving usability

If we believe that adoption begets customer retention, then barriers to adoption must be systematically found and fixed. Usage data enables you to spot areas in your application that have usability issues.

Some examples of how to find usability issues in usage data:

  • No actions beyond login. Is the user having trouble getting started? Does the product have a good “first-use” mode?
  • Patterns of navigation that suggest users are getting “stuck”: a user is on the cusp of performing a task, then goes to a product help page, then goes back to the task and never successfully completes it
  • Users abandoning a guided process partway through, such a wizard. Is the process designed in a way that they can’t, or don’t want, to complete it?

Six: increase revenue by increasing prices

When you understand the correlation between usage and retention, you understand the sources of value creation in your product. Did you price your product on this basis? In other words, are you leaving money on the table by not linking your price to the features that generate value? Most organizations set prices for many other reasons than their unique selling points.

Where to go from here: getting and using product usage data

Now we have a framework for the value drivers of usage data. And they can add up to major revenue lift.

The 4 dimensions of product usage

~ Don MacLennan ~ 1 Comment

There are six ways to increase revenue with product usage data:

  • increase revenue by increasing trial conversions
  • increase revenue by spotting up-sell opportunities
  • reduce churn by spotting declining usage
  • increase revenue by finding and mobilizing evangelists
  • increase revenue by improving usability
  • increase revenue by increasing prices

Collecting and understanding each of the four dimensions of product usage enables us to maximize the revenue increases that result from this analysis.

Product usage dimension one: frequency

This is the easy one.  How often does a user engage your application?  Think “logins per week”.

Declining frequency, or no use at all, is a strong signal for churn risk.

And that’s about it.  Don’t be tempted to infer too much from this one metric, such as “active use = satisfied user”.  You’re covering the basics of measuring any use versus no use.

Dimension two: feature usage

Now we’re getting to the juicy stuff.  Imagine all of the features and functions in your application.

  • At a global level, which features are used the most?  Which are not used at all?
  • And for a specific user, which features do they use?
  • And for a group of users inside a business customer of yours, what’s their feature usage pattern as a group?  Is the product widely adopted, or only by one person?

To get started, consider the granularity of what you’re trying to measure.  Don’t go too deep at first by measuring every button action in your user interface.  Otherwise you get a list of dozens or even hundreds of unique actions.  Which is too much data and introduces “noise” into your analyses.  Even powerful statistical analysis will yield inconclusive findings if too many unique actions are thrown into the mix.

Instead, start with a list of 3-6 high-level features.  Where “feature” is a fairly high level concept like “uploaded a photo” or “shared with someone else” or “created a new project”.  You can always refine your approach to get to 10-20 features in your list.  You might even go beyond that once you master how to interpret and act on the data.

Remember: start with a few features to measure and grow only when it suits your needs.

Dimension three: configuration

Many products, especially business-to-business products, are highly configurable.  Even in a consumer app, users can configure many settings like notifications.

The act of configuring an app beyond its out-of-the-box settings is a form of user investment in your app.  And with that investment can come loyalty.

Just like feature usage, the goal is to get an understanding of the patterns of configuration settings that correspond to loyal users.

Looking at patterns on configuration settings also helps you spot new segments of users, and areas of the user experience you might want to optimize.

Dimension four: volumetrics

Volumetrics means the degree to which your app is used.  For example, you’re the vendor of an online project management app.  How many projects, over a period of time, does a user run in your app?  How many projects in total does a given customer run?

Another example: you’re the vendor of an online backup/sync/share app.  How many files does a user upload, and of what size?

Most apps have 1-3 primary data objects worth measuring.  The greater the volumetric usage, the greater the dependency of the user on your app.

Like measurement of features and configuration settings, volumetric measures can also lead to insights about user segmentation and user experience issues.

Summary

We’ve reviewed the four dimensions of product measurement: frequency, features, configurations and volumetrics.  We’ve talked about the revenue insights that can be gained, such as trial conversions, churn prevention and up-sells.

Customer analytics | 8 data sources you can use now

~ Don MacLennan ~ 2 Comments

Customer analytics for SaaS companies is about reducing churn and increasing lifetime value of your existing customers.  While every company looks at acquisition metrics, SaaS vendors must do more because the customer lifecycle begins at the point of web conversion.

A customer that enrolls in a trial or subscription is beginning their relationship with your company. However, your SaaS offering will need to be constantly delivering the value your customer is seeking or you will fail at customer retention.

How do you know if you’re meeting customer expectations?

There is a recent body of research called behavioral economics.  In a nutshell, it states that what a customer does, versus what they say, is a truer measure of their satisfaction and intent. This means that a customer who says they are happy in a survey might not actually be happy.  Even worse is the customer who never tells you they’re unhappy, and they simply churn.

Customer analytics unlocks the power of this research by enabling SaaS vendors to understand their customers’ actions, focusing on product usage combined with everything else you can know.

How to get started?  Make an inventory of all of the data you have about your customers, and ways it can be used.

  • Traffic and volume metrics: Attributing customers to the channel that they came through will enable you to find patterns of valuable or less valuable customers.
  • Product usage: Logins, modules used, processes run, usage frequency.  This tells if your customers are using your product and what parts of it they are engaging with.
  • Sales CRM: (when you have a sales-assisted acquisition model)
  • Support CRM: Issues your customers needed help with. Use it to spot your customer’s frustration with your business.
  • Purchase & product plans: What your customer purchased and how much they paid.  Identifies customer tiers, potential up sells, upcoming renewals and payment status.
  • NPS ® (Net Promoter Score ℠): Surveys taken on a subset of your customers to gauge satisfaction.  Allows your customers to tell you qualitatively vs. quantitatively if they are happy with your service.
  • User comments: In emails, on support forums, or even Tweets. A second data source of customer sentiment that can be tied with NPS and Support CRM.
  • Customer intelligence: Key changes within your customer’s organization that could signal potential changes in the stability of your relationship.

Most SaaS companies review metrics from each of these sources independently.  Customer Analytics happens when we analyze our customers’ data across multiple touch points.   This reveals valuable insights you would have no way identifying if the data remained apart.

Look across all of your customer data, and consider what it means across the lifecycle.  The result? Better products and increased revenue.

Don’t get caught using averages (part 2)

~ Don MacLennan ~ 2 Comments
Pareto/Power Law distributions: the needle in the haystack

I wrote previously about the prevalence of Pareto/Power Law distributions in product users’ behavior here.  Wow, that’s a lot of alliteration.

But the discussion stopped at only one dimension of data.  For example, a single dimension like Free versus Paid users of a Freemium product such as online backup.

The story gets really interesting when you consider multiple dimensions (aka variables) of data at once, each with its own Pareto characteristics.  The outcome can lead you to a some very interesting places.

In the first scenario, a small set of users in Dimension One (let’s say, Paid product users) also represents a small set of users in Dimension Two (let’s say, country of user origin).  This can mean that a tiny percentage (sometimes less than 1 percent!) of an entire user base represents almost all of the revenue or commercial value.

When this happens, it’s incredibly important to know who these users are; you’ll need to hang onto them for dear life to protect your revenue stream.  For example, you might cater to the specific needs of users from their country of origin.  Do you think users in China have different product needs than in France?  Probably.

In this scenario, you’ll also need to consider a revenue diversification strategy to protect your risks of relying on such a small segment.

Another scenario is that users in Dimension One (again, Paid users) don’t belong to the majority (or, “head”) of the distribution within Dimension Two (again, country of origin).  In which case, the implication is that country doesn’t matter in targeting your best (e.g. paying) users.

You can go astray in this scenario by looking at country of origin in isolation.  Maybe you have a huge pool of users from Germany.  The temptation would be to conclude “Germany is my most important market”.  Unless you knew that paid users didn’t cluster around a single country and that Germany was comprised of lots of free users.

What to conclude?

One: make sure you know if your most valuable user segment is much smaller than a Normal distribution would imply.  Most people think that their most important user segment is something like 10-20% of their base.  If 1% of your users drive the business, know who they are, find more like them, and don’t lose them.

Two: don’t let any one dimension of data drive your definition of user segments and internal decision-making.  If you hear sound bites inside your company like “German users are our most important”, that’s being too imprecise.  It generally takes 2-4 dimensions/variables to be precise about a user segment and to know how to best treat them (“Paid users with broadband PC connections in Germany are most important”).

Three: if you truly have 1% of your users driving the business, consider diversification strategies.  You’re carrying a lot of risk, but you also have 99% of your users from which another valuable segment can be found and served.

Last: as I argued in the prior post, it’s easy to dismiss the Pareto effect as only applying to obvious examples like Freemium for online consumers.  I’ve found the same patterns in other businesses.  In which case the gap between reality and perception is even wider!  Spend some time hunting down these patterns inside your company.  I promise you will be rewarded with new insights.