Leave a comment
Get the GH Bookmarklet

Ask GH

I'm trying to get a sense of LTV for a customer, and they only offer annual plans, starting less than a year ago, so I have no idea what churn rate will be. Any advice on ways around waiting for the year mark? I was thinking of looking at engagement levels as one proxy, but maybe someone else has a better idea.

  • SE

    Sean Ellis

    over 5 years ago #

    My first reaction was that you can't do it. But I agree that engagement levels are probably a good proxy. I would define a period of time of inactivity that indicates the account is "dead". Maybe something like 45 days. Over time you can adjust your model. For example, you might find that 10% of accounts with no activity during 45 days start using again.

    At Qualaroo we went from monthly and annual plans to only annual plans. And this was great because we didn't have to worry about churn for a long time. But we recently add quarterly plans to the mix. We did it for a variety of reasons, but one of the benefits I believe is that quarterly plans keep us honest with ourselves. A great sales/marketing process can't mask a crappy product. Quarterly plans are enough to really give people a chance to use the product, but we get much faster feedback about the use cases and user types that continue to renew. In the short term it might not be as good for our business, but in the longer term I think it helps us build a healthier business.

    So you might want to consider quarterly plans if you want a faster churn feedback loop.

    • RG

      Ruben Gamez

      over 5 years ago #

      I really like the suggestion of looking at engagement/activity (or whatever you want to call it) in this way. Yeah, I'm looking at this during the trial period to help me improve activation, but I've not done as good of a job focusing on that after they've become a paid customer.

      Since I do monthly billing, I've relied on customers canceling to help me measure and improve churn -- because the feedback loop is relatively fast.

      But is it really?

      I was just listening to the most recent episode of the "Jobs to be Done" podcast with Des Traynor (from intercom.io), where they talk about churn and engagement a bit: https://itunes.apple.com/us/podcast/jobs-to-be-done-radio/id499859427?mt=2

      At one point they had a great comment that often a customer really churns way before they actually get around to canceling. I know that's the case for myself with services that I've used, and there have been customers I've talked to where that's also true. Not always, but enough to realize that relying on them canceling (as a proxy for engagement -- which is the reverse) probably isn't ideal. I need to do a better job of tracking/measuring that directly -- both of these things.

      • LM

        Lincoln Murphy

        over 5 years ago #

        Cancellations are a lagging indicator that you have a churn problem. I'd say one of the laggiest.

        Churn is rarely only an end-of-life event. It usually starts earlier in the lifecycle... and I've seen the seeds of churn planted in the sales cycle where promises are made that can't be kept later on.

        I had a few posts where I touched on this:
        http://sixteenventures.com/improve-saas-churn-rate-engagement
        http://sixteenventures.com/saas-churn-threats
        http://sixteenventures.com/growth-hacking-retention

      • CN

        Chris Neumann

        over 5 years ago #

        I've talked to other SaaS companies where they had certain customers that were not using it but still paying, and eventually they'd flag them so they never got contacted lest they remember to cancel. :)

        For b2b stuff, I always wished for a way to not bill people around Thanksgiving or Christmas since often that's when people that aren't using it suddenly "remember" to cancel.

        All that said, I'd much rather have someone using my SaaS, since they'll tell their friends. Otherwise, if someone asks, they'll respond "oh yeah, I was using that for a while, but stopped. I gotta go cancel...." Not really what I want.

        Lincoln Murphy actually suggested auto-canceling people that aren't using the app and giving them their money back. That seemed interesting, and I wonder if he actually got anyone to do it, since the goodwill might outweigh the "free" money from the non-users.

        • LM

          Lincoln Murphy

          over 5 years ago #

          I think what Lincoln said was if you suddenly grow a conscious and admit that you know people have been paying you for 6+ months without using your app, to maybe do the right thing and cancel their account and refund their money.

          But the real message was not to let that happen in the first place. I referred to customers that pay but don't use as "zombie" customers in this post:

          http://sixteenventures.com/saas-customer-success-zombie-customers

          ...and yes, we've done it a few times and it creates much goodwill.

          The folks at Gainsight - purveyors of a Customer Success Management app - had a post recently where they also used the term "zombie customer" but instead referred to customers that are paying you AND are using your app, but that will end up churning out.

          I like their definition better since it really seems more like a zombie... but I like the meme I used better:

          http://customersuccess.gainsight.com/no-silver-bullet-solution-customer-churn/

          That post riffed off of this original one:
          http://customersuccess.gainsight.com/3-myths-on-the-indicators-of-churn/

          Ultimately, what those posts are illustrating is that even just measuring use - or engagement - isn't a silver bullet, because some customers will churn anyway.

          Instead, Gainsight advocates - and so do I for what it's worth - a more holistic approach to Customer Success that aggregates multiple inputs - many generated by actually talking with customers - and generates a Customer Success Index, Customer Health Index, Customer Happiness Score, etc.

        • SE

          Sean Ellis

          over 5 years ago #

          Auto cancel would be extreme, but it could be really effective to say "I noticed you haven't been using the service and wanted to remind you to cancel before May 10th (your renewal date). Of course I'd rather help you get value from the service, but if you aren't using it, you shouldn't keep paying for it. Please let me know if you have some time this week when I can talk about how we can start meeting your needs." That type of approach would certainly get my attention.

          • LM

            Lincoln Murphy

            over 5 years ago #

            It's amazing to me how often people running SaaS companies say to me that they're afraid to contact a customer (as has been mentioned in these comments numerous times) because they don't want to remind people to cancel.

            If they weren't afraid to contact the customer initially - to, you know... get them to use your app - then you wouldn't have to be afraid to contact them later.

            But a lot of the time, if you were to send an email like the one Sean suggested, that would be the first time the customer has heard from you in a long time... so yeah, they'd be like "oh crap, I should cancel that."

            But what if you - instead - reached out with something that attempts to get them into the app? What if you reached out individually to see if you can help them get started?

            Will some be "reminded to cancel?" sure... but probably - just speculating here - they didn't originally sign-up so they can cancel 7 months later and take food out of your baby's mouth (evil customers).

            They probably wanted to use your app. You made it complicated to get started (or, I'll be nice, they got busy/distracted/etc.) and just never got on-board.

            But they did want to use your product... why don't you see if you can help them actually do that instead of just telling them they could cancel?

            I'd say it's not an either / or thing... those > 6 months of paying w/ no use... send them the "you can cancel" email or just do the auto-cancel/refund thing.

            Those < 6 months, try to re-engage and get them using the product... that's what they wanted initially, right?

          • CN

            Chris Neumann

            over 5 years ago #

            Since you're the CEO of a SaaS business, I look forward to your feedback on how well that works. :)

            • SE

              Sean Ellis

              over 5 years ago #

              Not sure I'm bold enough to try that on all, but we might try it on a few...

        • RG

          Ruben Gamez

          over 5 years ago #

          I actually remember hearing someone (can't remember exactly who) say they automatically cancelled after a year of no activity, and got a couple of complaints about it because they still meant to go back and use it (and knew they were putting it off).

    • LM

      Lincoln Murphy

      over 5 years ago #

      I agree with Sean... as far as using real numbers to estimate a customer's lifetime, you can't.

      And I'd say your year one churn numbers will (probably) be less-than-helpful in estimating second year churn because - hopefully - you improved things as you went along.

      As I said somewhere else in the comments, Engagement is great but you may want to take it to the next level with an overall Customer Success Index, Customer Health Score, etc. because that'll give you a better picture of who's a churn threat and who's not.

      Then - and really only then - can you have some idea of what churn looks like in an effort to estimate customer lifetimes and then the value of those customers (LTV).

      This is a great post from David Skok that brings together Customer Success experts from Gainsight, Totango, and others to dive deep on reducing churn that might give you more ideas:

      http://www.forentrepreneurs.com/customer-success/

      I've also written a few things about engagement and Customer Success:

      http://sixteenventures.com/improve-saas-churn-rate-engagement
      http://sixteenventures.com/saas-customer-success-eliminate-dead-ends

  • LM

    Lincoln Murphy

    over 5 years ago #

    I've posted a lot of answers throughout the comments that deals with engagement and the fact that it may not tell whole story.

    But to answer your question a different way, you could estimate would be to do some research on churn rates other companies have and use that as a proxy.

    You may have to employ some creative thinking to get these numbers for private companies. You may be able to get this level of granularity from annual surveys by Softletter or OpexEngine, though I doubt it.

    But we're all growth hackers, right? I bet we could come up with several ways of obtaining this information in a very ethical, very above-board way.

    Oh, let me be absolutely clear... the only way this has any chance of being the least bit accurate is if you get data on other companies in the SAME product category and the SAME market position/price point.

    Don't go look for generic SaaS churn rates that come from aggregating every type of SaaS (Enterprise, self-service, cheap, expensive, etc.)... SaaS is a business model, not a market or category of product; you want to make sure you're doing an apples to apples comparison.

    Then you have some idea where others stand and you can do what ifs around being better than them or worse.

    One note.... I've often seen lower renewal rates with annual plans because of the lack of engagement (and low customer health score) so whether you use engagement as a measurement tool... make sure you use it for the other thing it's good at: keeping customers around.

    If you let a year go by without any contact with the customer, if they do renew (or are auto-renewed) you may be issuing a lot of refunds or dealing with chargebacks.

    Oh, and a lot can change in a year... make sure you're using a payment processor that updates credit cards and that you're doing pre-dunning to get people to update cards... on annual plans especially, bad CCs can wreak havoc on your retention rate.

    But overall... just don't hide from your customers.

  • JG

    Jim Gray

    over 5 years ago #

    Churn is an authoritative measure of perceived value, but it gives you an integrated signal which samples from a long window of user sentiment. Potentially very long, in the case of annual billing.

    It's also a signal in the sense that "hitting the ground" is a signal that your airplane's flight controller was off true. It's hard to correct from in most cases.

    So what to do? Back up a level and use behavioral signals to model user perceived value with higher time resolution. These models can be either very complex, or very simple. But you should probably start simple and follow up with one-on-one contacts as a sanity check.

    You can probably think of a few good measures of whether a customer is getting value from your product right off. Some of these will be correct, some will be less accurate. Test them against your month-to-month subscriptions to find out which are actually important to your users.

    Here's an example of clusters to look for:

    * Consistently active
    * Consistent but less active
    * Random sprees of high activity
    * Random use, low activity
    * Activity only after (event) occurs
    * Not active

    When users move from one cluster to another, it's probably a good idea to follow up with them in some way - either to reinforce changes that are likely to increase CLTV, or to try to recover from service problems or loss of perceived value.

    You can automate that a bit with triggered events and such, just don't be spammy and impersonal about it, and be sensitive to the implicit sentiment which may be associated with it. Also be sensitive to the fact that often your service isn't the cause at all - maybe they got sick, they're short of work, they changed employers - lots of things that are NOT your service can affect user behavior.

    Annual billing is great - but it also carries in increased responsibility to make sure that users keep getting value from you. Watch more signals than just churn, because churn is a signal that you've already let them down.

  • TW

    Tommy Walker

    over 5 years ago #

    I'm actually in the process of writing an article on this now, should be published Wednesday...

    But, what you need is a predictive calculation to help you understand your churn behavior.

    This post on shopify gives a very useful calculation at the end that can help you.

    http://www.shopify.com/technology/4018382-defining-churn-rate-no-really-this-actually-requires-an-entire-blog-post#axzz2usZ5VbVX

    Before applying the calculation, make sure you're only doing it with your different customer segments. (pricing plans are a good start)

    Hope that helps.

  • BH

    ben hoffman

    over 5 years ago #

    I recommend reading these two blogs. There is a TON of info on LTV and CAC…. I can't remember exactly which post discusses your question but I do believe they address this exact question.

    David Skok: http://www.forentrepreneurs.com
    Chris Janz: http://christophjanz.blogspot.com

Join over 70,000 growth pros from companies like Uber, Pinterest & Twitter

Get Weekly Top Posts
High five! You’re in.
SHARE
11
11