SaaS CEOs: Measure Customer Engagement – Increase Conversions & Lower Churn

The goal of a SaaS CEO should be to increase the profit they make from each customer (LTV), and lower the costs in sales and marketing that it takes to acquire each customer (CAC). Measuring Customer Engagement is a key tool that will help you achieve that goal, as it will allow you to increase your trial conversion rates, which directly reduces CAC. And it will help you lower your churn rates, which directly increases LTV. (There are also other significant benefits that are described below.)

How customer engagement has changed in the on-line world

In the old world, most of the ways that companies engaged with their customers would involve human interactions either face-to-face, or over the phone. As more and more transactions have moved on-line, businesses are losing the insight they would gain from employees talking directly to customers. That turns out to be a major loss, as understanding how your customers are using your product/service and what they think matters a great deal.

There are some tools that help us judge how customers as a whole are behaving on-line, but so far those tools have not allowed businesses to analyze customer behavior at the individual level. In this post, we will look at two emerging tools, Totango and Apptegic,  that help SaaS companies understand their customer behavior at the individual level.

Why should you care?

You might ask why would you care about how individual customers are behaving? The answer is that this information can be used in three important ways:

  1. Identify customers that need help, or that are ready to purchase, in SaaS free trials. Intervene using either email or human touch to increase conversion rates.
  2. Help identify customers that are about to churn. Intervene using email or human touch to try to turn them around, thereby reducing churn rates.
  3. Help identify customers that are appropriate for an upsell or cross sell

There are also additional benefits:

  • Look for patterns of behavior that are correlated with another attribute. This will help you to identify the key segments in your customer base. For example, your customer behavior may vary according to the following specific attributes:
    • Which lead source they came from
    • Geography
    • Size of company
    • Vertical industry group
    • Job Title
    • etc.
  • The above information can be used to improve the targeting of your messaging and product features
  • Knowledge of how users are navigating through your product can help you improve product design. For example, if you could identify parts of the product that are not getting used, and parts where that are heavily used that can tell you what users care about, and/or where the product navigation needs to be improved. You could use a system like this to A/B test new product features.
  • Two of my portfolio companies have been able to cut support costs by emailing tips and FAQs proactively to customers that are starting to use a particular product area
  • Individual customer behavior histories could be helpful to customer support people when contacted by a specific customer. It would allow them see how that customer had been using the product.

How do you measure Customer Engagement?

Measuring customer engagement requires that you do the following:

  • Identify the key events that you would like to track. (Note that sales will care about a few high level events, while development and support will want a far more detailed set of events.)
  • Assign engagement scores to each action in a way that can be changed as your app evolves
  • Instrument your SaaS application to collect log data for those events, with the specific userID
  • Create a database to store this information, ideally in a form that allows for fast querying, and easy cube-style, cross tabulated analyses.
  • Connect this database with other data sources that contain additional customer attributes that are relevant for understanding segmentation
  • Compute an engagement score for each user that can be used by sales and support.

Handling multiple users from a single company

Several SaaS applications are used by multiple users in a single company. For this situation, it will be useful to track the engagement in the following different ways:

  • For the whole company
  • For each individual user
  • For each type of user (e.g. administrator, sales people, sales managers, etc.)

Computing a Customer Engagement Score

Customer Engagement Score = (w1*n1) + (w2 * n2) + … + (w# + n#)

Where w is the weight given to a random event and n is the number of times the event occurred.

HubSpot’s CHI Score

A long time back the team at HubSpot realized the importance of understanding churn, and being able to look at it by customer segment. They developed a customer engagement system which delivered a CHI score. CHI stands for Customer Happiness Index. HubSpot used this to identify customers that were at risk of churning, and used their support staff to call those and improve their usage of the product.

They were also able to tell which parts of the product were sticky (i.e. less churn if the customer used those features). Another use was to compare the two main customer segments that HubSpot sold to, to understand which were more likely to churn, and therefore less attractive to sell to in the long run unless a fix could be found.

Totango and Apptegic

Totango and Apptegic are two startups that have emerged to help automate the process of measuring customer engagement for you. Both companies are initially focusing on SAAS services. Totango is the more polished application, its UI and integration with other solutions like Salesforce are further along. Apptegic’s user interface is earlier in development but its underlying platform seems to be more powerful.

Let’s start by taking a look at Totango:

Totango – a deeper look

One of the most appealing things about Totango is that it interfaces with Salesforce.com, providing each rep detailed information about what their prospects who are involved in a trial are doing with the software.

Engagement scores are shown in the screen shot below using a color coded bar. This helps guide the salesperson to call or email the customers that not using the product. See the red highlighted box below:

image

 

The Totango Insights column, highlighted in red below, allows the rep to see what the customer has done with the product. E.g. how many other users have they signed up, what functions they have used, etc.

image

 

The Online Status column, highlighted in Red below, shows the rep which customers are actually on-line right at that moment. This helps the rep know when is a good time to call.

image

Sales reps can drill down on any customer to see the full details of what the customer has done, see below:

image

 

And they can add any customer that they are concerned about to a watch list, and Totango will notify them by email whenever anything changes in the status of that customer. That makes is much easier for the rep to stay on top of multiple problem customers without having to keep inspecting the status of each one. See below:

image

 

Apptegic – A deeper look

Apptegic are still in private beta, so my comments here represent an earlier version of the software that was demonstrated to me by the founders. They are making fast progress, so things will evolve quickly.

Apptegic does not yet have the integration with Salesforce.com shown above, and still has some way to go to achieve the same application level polish. Instead what was shown to me was a platform based on a sophisticated Cassandra back-end, with data saved in a form that allows for real time queries. At the time that I saw the demo, Apptegic had layered on top of this platform a query tool, and a dashboard. They had also recently added the following Customer Engagement screen to start to layer on top of the platform simpler tools for salespeople:

image

 

Apptegic are also interested in helping the product designers with information about how the product is being used, which they can then use to improve the product. See the screen below which shows the most popular activities and how people navigate from one area to the next:

image

 

The strength of Apptegic over Totango appears to be their database architecture, which appears to be more fully thought through and future-proofed than Totango’s backend. Apptegic provide features for integrating other data sources, which allows you to import important customer attributes and behaviors that might be stored in CRM, support and billing software. Like BI cube software, they claim that they can take any attribute and use that to look for patterns with their real time query system.

The real time queries could allow SaaS developers to personalize their software in real time. Several of their beta customers have indicated that they plan to use this to help drive in-app marketing based on the customer’s app usage patterns.

Apptegic also claim that they will automatically look for patterns and correlations in the background, and when one is found, bring it to the attention of their users. This might help to show for example that the users that are churning faster all have this particular attributed in common (e.g. size of company, job title of the user that signed up, etc.).

Behavior-driven Communications – HubSpot Enterprise

Another product that is interesting to look at in this context is HubSpot Enterprise (which was formerly known as Performable before its acquisition by HubSpot). This product was designed to allow marketers to send “behavior-driven communications”. To do this, HubSpot not only tracks user events inside of your SaaS application, but also relevant customer events coming from other systems such as ZenDesk where user might have logged a support request, Twitter, Facebook, etc.

Take a look at the screenshot below, which is from their own internal system to see how this can work. This shows is how you can detect a particular event such as a user that has registered for a free trial, but not yet installed their Analytics tracking module, or viewed an Event Report. That indicates that the user is missing out on the key benefits of the system. By sending them an email that contains a link to show how to install the Analytics tracking module, you have a good chance of bringing them back to the system, and solving the issue that might have been stalling their progress. (Salesforce.com is another company that does this today, also telling you how you are doing versus other customers in your category.)

 

image

 

This highlights the importance of being able to easily connect your customer engagement system to marketing automation systems, so that events can be easily used to drive communications as shown above. Currently this feature is missing from both Totango and Apptegic. They need to create an easy feature similar to what is shown above to drive email campaigns in Marketing Automation products like HubSpot, Marketo and Eloqua.

Note: HubSpot Enterprise is not focused on the Customer Engagement marketplace like Totango and Apptegic. Their primary focus is on marketing automation, and allowing a marketer to track events that are relevant to marketing.

What should you be looking for?

To help prospective buyers of this kind of system, I have put together the following diagram as a starting point to highlight some of the functions that you might want to consider:

image

 

Who can benefit?

As shown in the diagram above, there are several departments that stand to benefit from Customer Engagement intelligence:

  • Sales
  • Marketing
  • Support
  • Product management
  • Business intelligence analysts (assuming you have such a role).

We have covered how sales, marketing support and product management can use the product, but not yet talked about the business intelligence that such as system can generate. To cover that, I need to explain why customer segmentation is important.

Why Customer Segmentation matters

Not all of your customers are likely to have the same needs. In an ideal world, you would have the time to get to know each customer and treat them individually according to their needs. However in a world where you are dealing with many customers and needing to keep sales and support costs low, that isn’t practical. The next best thing you can do is segment them into groups that have very similar needs and behaviors. The benefits of this are:

  • Allows you to send far more personalized and appealing communications, with messages that directly resonate.
  • Understand the profitability of specific groups based on how easy it is to acquire them, and how much they spend, and how long they stay with the service
  • Determine the product needs of specific groups that you might have decided to focus on based on their profitability
  • Determine why certain groups are churning, and if possible and appropriate, modify the product to more closely meet their needs
  • etc.

Customers can be segmented according to vertical industry, job title, company size, etc.

When you’re a startup, Focus matters

One of the most important ways a startup can succeed is to focus their efforts on a single customer segment in the early days. This allows them to use their limited resources to build a product that fully fits the needs of that one group, as opposed to partially meeting the needs of many groups, but not being over the bar for any one of them. Messaging can also be very directly targeted to appeal to that one group. To do pick this one group, it helps to know which groups are responding well to the product, and which are most profitable.

Where Customer Engagement intelligence can help

A good customer engagement tool will provide your business analyst with both query tools, and automated pattern recognition tools to help understand customer segmentation. The query tools will allow them to ask questions such as what are shared characteristics for customers that:

  • Churn
  • Purchase
  • Don’t purchase
  • Use specific modules/functions
  • Respond to upsell messages for particular modules
  • etc.

This will help you identify potential new customer segments and persona that might not have been obvious. It should also be easy to create queries to find out how your known customer segments and persona are behaving.  At this stage Apptegic looks to have the upper hand in this area.

Automated Pattern Recognition

Apptegic claim that they will introduce automated pattern recognition in future versions. If they are able to pull this off, it could be a very powerful feature, eliminating the need to do lots of queries to search for patterns.

What comes next?

This particular space is still in its infancy. Neither Apptegic or Totango meets all of the needs that I have discussed above. However both companies are young and nimble, and likely to evolve fast.

Summary and Recommendations

All SaaS companies should have in place a strategy for measuring customer engagement and using the resultant data to improve trial conversions, recognize upsell opportunities, and to predict and proactively prevent churn.

I recommend starting by evaluating both Totango and Apptegic. However given the early stage of the industry, and how things may change, you might find it useful to do your own instrumenting of your app, and maintain your own log files. That way you will have the safety of knowing you have collected the data you need to move to any future system without lock-in.

Be Sociable, Share!

About 

hubspot-cta-2
  • http://www.totango.com Guy Nirpaz

    David,

    Thanks for the great overview.

  • http://www.facebook.com/salman.s.malik Salman Malik

    Great post, as always David.  I also have a hunch that apps like Totango and Apptegic have a home inside the enterprise where IT departments roll out SaaS apps internally and to customers and partners of the enterprise with (usually) almost no engagement tracking.  The rollouts then fail over non-adoption when engagement tracking early on could have highlighted what was and wasn’t working.

  • http://www.forentrepreneurs.com David Skok

    Salman, I think you are right about that. Those apps that represent revenue to the corporation will be the first to see this kind of instrumentation.

  • Nikos

    David,

    Thanks for a great post and overview of Totango and Apptegic.  We at Lingospot have also embraced the notion of a CHI, but to-date we’d have to rely on our observations and all too frequently on our “gut feeling” about each customer.  This makes things much easier.

    Nikos
    Lingospot

  • http://argylesocial.com/ Eric Boggs

    This is great stuff, David.  Thanks for sharing.

  • http://www.facebook.com/skierajs Aaron Severs

    Great article! I’ve been working with Apptegic in their private beta and they have been great to work with. Easy integration so far.

  • http://startupcfo.ca startupcfo

    Great advice David. At Tungle we had e-mail alert triggers to ID users at risk of churning out. We had similar alerts for conversion opportunities. I have seen Totango and think it and the other solutions listed here are important elements in optimizing CLTV. The most important metric to a SaaS business is churn. And engagement (or lack thereof) is clearly a churn driver.

  • http://www.forentrepreneurs.com David Skok

    Mark, thanks for the validation.

  • http://www.teamworkpm.net samkidd

    Another great post. Thanks.

    One issue we often see with our app is that one member of the customers team is really keen, normally the person who signed up, but getting buy in from the whole team can often become a stumbling block. Even though you may feel like you have  good customer interaction, it’s only with one team member and not always the decision maker.  

  • http://www.forentrepreneurs.com David Skok

    That’s a great point. Thanks for pointing this out.

  • http://www.apptegic.com Karl Wirth

    For this reason, I think its important for B2B apps to be able to measure engagement at both the account level and the user level.  We can vary an account level engagement score based on changes in the account’s number of users and their individual levels of engagement.  Sam, do you think this would help address the issue?  Does anyone have other ideas on what else should be done to address it?  For instance, should we factor the users’ roles into the account engagement score?   – Karl, Apptegic

  • http://www.teamworkpm.net samkidd

    I think that might help give a better idea. I know it varies from account to account and it won’t always be 100% but I think combining the amount of user and all their engagement would give a better idea.

  • http://www.totango.com Guy Nirpaz

    Sam, you’re right, one of the biggest challenges of online selling to organizations is understanding the buying and adoption (on-boarding) processes. 

    The solution we found most effective is to present a holistic customer view which presents engagement indicators at the user level rolled up to the account level. 

    Guy,
    Totango

  • http://www.deskaway.com Sahil Parikh

    Great post David! Very helpful. Thank you.

  • http://twitter.com/jlopin jlopin

    Thanks for the great article, David. Here are two ideas on this topic that I’ve been incubating in terms of how we can better leverage CHI at HubSpot to create EV.

    1. The factor most positively correlated with happy HubSpot customers is the breadth and depth of their usage of our software to do inbound marketing. Simply put, customers who use more of our product are happier and get much better results. I would love to carve out real-estate inside our product that automatically displays “CHI-driven targeted information” intended to help customers get more out of our software based on their specific usage patterns. (I even asked Guy to consider building this feature into Totango so I can use it :)

    2. One thing that is true at HubSpot, and I suspect is true at other healthy SaaS companies, is that customers who use our product more get better business results. Since more usage also means higher retention rates, our incentives are perfectly aligned with our customers: we both want to see deeper, more meaningful use of the software. With that in mind, CHI should be a tool for customers as much as it is for folks running the company. I’d like to see CHI drive a customer-facing “to do” list and/or percent complete tracker that guides customers in the right direction to get more value from the software. If we can pull this off then our customers will get lots more traffic and leads (they win) AND they’ll stick around for a long time as happy customers (we win). CHI: something we do with customers, not to them.

    Jonah

  • Yesuifen20

    That turns out to be a major loss, as understanding how your customers
    are using your product/service and what they think matters a great deal.

  • Yesuifen20

    understand their customer behavior at the individual level.

  • http://www.webhostings.in/ hosting server

     I am glad to catch idea from your article. It has information I have been searching for a long time. This looks absolutely perfect.

  • http://www.facebook.com/gil.sadis Gil Sadis

    David, another really good post. I would like to comment on your building blocks diagram. I think it lacks (and we see it from the demand to our product) the connection to the payment provider. After knowing your users’ behavior and segments you need to take actions. And a lot of times these actions are pricing & packaging actions. Finding the right pricing plan for each segment, that will reflect the value your product gives, is a crucial element in the customer engagement process as we see it.

    Gil,
    Licensario