Leave a comment
Get the GH Bookmarklet

Ask GH

I can see 30 day trials getting you a higher conversion rate (greater value), but 14 days to payment rather than 30 is good.

Which should you use? I can see a standard answer being "split test it and find out," but for many companies that assumes a certain amount of traffic. For small operations just starting out, what factors can help make an informed decision in the absence of data (i.e. not instead of, but until they get enough traffic to test)?

  • SE

    Sean Ellis

    about 7 years ago #

    A really important factor in this decision between a 14 day and a 30 day trial should be whether you have a touch or no touch sales model. If the majority of your sales are not assisted by a sale team, then 30-days generally works pretty well. But if your sales team manages a pipeline, then 14-days can lead to a lot more sales per sales person. A short trial period means that each sales person has a lot less prospects that they need to manage at any one time in the pipeline.

    If you are purely no-touch ecommerce then a split test can be the way to go (assuming you don't have the constraint mentioned in your question details).

  • JG

    Jim Gray

    about 7 years ago #

    Free trial should be long enough to accomplish activation, and long enough to persuade pre-trial customers that they will adequately be able to evaluate it (given complexity of product, complexity of corresponding business process, number of stakeholders...).

    Which means that length isn't really something that makes sense to mess with outside of an analysis of your onboarding process. What sort of product is it? What concerns do customers have? What behaviors signal activation or lead to the customer extracting value?

    You can A/B test the effect of "the number" on initial conversions, but the number that become activated & the number that end up at a given pricing tier, and the long-term effect on churn & retention - this is more complicated and more dependent on what you DO during the trial period, how the service handles new users.

    So honestly, I'd A/B test to see if there's a meaningful impact on conversion...then I'd buckle down hard on studying and improving "new player experience" to see what has the biggest impact on driving activation + subscription + retention. Which is way harder than running an A/B test.

    See numerous posts by Lincoln Murphy around the topic:


    • LM

      Lincoln Murphy

      about 7 years ago #

      Jim nailed it... and thanks for the links, man.

      The easy way to think about this is that a Free Trial length has two sides... customer-facing and inward-facing.

      Customer: Long-enough for them to think they'll have time adequately evaluate. It's a marketing gimmick.

      Vendor: As short as possible.

      Don't confuse the two... don't buy into the external-facing trial length as the time you have to stick to to convert 'em. Most SaaS vendors do this. I think it's safe to still say most on this.

      It's all about what you do during the trial to get 'em on-boarded, engaged, invested... and then convert 'em... the "trial length" just helps get 'em in.

      I came up with a way to think about this called CCAs - Common Conversion Activities - that, when applied to a Free Trial, allows you to more effectively engage and make offers to convert. More on that here: http://sixteenventures.com/free-trial-metrics

      In my experience, 30-day trials usually convert on day 31 because the vendor doesn't ask for the sale until (if they ever) the trial expires.

      As an extreme example of what can be done when you decouple external Free Trial lengths from your internal sales process, I have one SaaS company I work with that took their 42-day average conversion time down to 3 days. On a 30 day trial.

      Oh, and their customers convert on average - in those 3 days -with a 33% higher Average Subscription Value (ASV).

      I'm just kinda a little proud of that.

      • LM

        Lincoln Murphy

        about 7 years ago #

        Oh... I didn't address how product complexity, level of touch involved during the sales process, etc. affects the decision on Free Trial length.

        It doesn't.

        Well, it doesn't from your perspective. It only does from the perspective of your customer. If you are replacing or competing with big, legacy products that took years to implement and you have a 14 day trial (or even a 30-day trial)... the customers you seek may avoid you because it's clear that your product is just a toy.

        It may not be... it might be awesome, but you don't get how to position your offering in the market. Maybe don't even lead with a Free Trial... let alone a self-service one.

        But where we usually get into trouble is when we talk about Free Trials and situations where "our product is too complex" or "our sales process is so high-touch and takes so long."

        Those are far too often just excuses to let you off the hook.

        It allows you to not focus on simplifying your Free Trial so a potential customer can experience an emotional "Wow!" quickly... it allows you to avoid the reality that your sales process is so long because you're making it that way.

        I'm sorry, but even in the scenario I mentioned at the start of this comment, a complex product is not a selling point. It's not something to be proud of. Even in that scenario, while I may not lead with self-service and ease of use to get their attention, that may very well be exactly what I use to get them hooked once they're in.

        Sell 'em what they want... give 'em what they need, right?

        So, what if you thought about your customer, figured out when in their buying cycle they're ready to start the trial (not always right at the beginning), and what a successful trial might look like for them?

        What if you put book ends on the trial (the "length") that gave them enough time to evaluate the product, and put the trial in front of them when they're ready to evaluate the product?

        And then, what if you made it so that while they think they're evaluating the product... you're secretly getting them to actually use it... to become invested in it so they stop evaluating other products and becoming a paying customer is the most logical next step?

        And what if you did that within a complex sales process, so that a Free Trial was considered a well-orchestrated part of a high-touch sales process and not the either/or or "let the customer kick the tires and play with it and I'll check back in 30 days" way people tend to think of a Free Trial?

        If you did that, you might be like one client that took their Free Trial - in a high-touch market - from 83 days (on a 30-day Free Trial) to 21 days (still on a 30-day trial).

        How'd they do it? Well... there was a lot to it (part of it was completely changing everything they were doing), but one of the key elements was understanding the buying cycle of the customer, understanding the different personas, the catalyst for change, and engaging the right people at the right time to drive the sale.

        This means knowing your Ideal Customer better than they know themselves, understanding all the players involved in the sale, understanding how to use the Free Trial to get low-level "users" to drive engagement with higher-level decision makers during the trial process, and connecting usage behavior to the sales folks so they can push the process along and reach out when the CCAs have been met to close the deal.

        So even in a high-touch sales process, you can apply these same principles... but the Free Trial may not be what you lead with, it may come into play in the middle of the sales process.

        And I'd also say there's not really no-touch and high-touch sales processes... or there shouldn't be... it's much more nuanced than that, and may be for the same product at different price/complexity points. And how you apply a Free Trial in each one depends... but we need to stop the one-size-fits all approach to Free Trials to make it work.

        Free Trials are such an interesting thing... so many companies - SaaS or otherwise - use them... so few understand how they really work and even fewer have really unlocked their potential.

        • GG

          Gab Goldenberg

          about 7 years ago #

          Amazing stuff Lincoln. Do those blog posts linked above include the case studies you mention in your comments here?

  • BH

    ben hoffman

    about 7 years ago #

    Before giving my opinion i would advise installing optimizely and simply running an a/b test for 30 to 45 days. It's super simple and will pay off in the long run.

    Choose a winner and then wait and additional few months to see how many became paying customers and how many did not churn.

    You might also want to check engagement levels.

    I know you don't want a "test it" answer but this is really one of those situations where you genuinely have to test and analyze.

    With that said, if what you can accomplish in 30 days can also be accomplished in 14, then go with the latter. With 30 days, usually customers take their time and don't feel the rush to test drive the product. With 14 days there is more of an urgency to use the product.

  • PH

    Paul headly

    about 7 years ago #

    We have tested this and it didnt really make a difference. I think that it depends on the complexity of the product and the market it addresses B2C or B2B? . If your product is an easy to use B2C solution and it is understood at the on boarding stage then a short trial time is adequate to test effectively.

    If you product deals with a more complex task them they will need more time. If it is above a certain pricing threshold they will probably need a sign off which means that the trial should be around 30 day minimum

  • TD

    Tiffany Dasilva

    about 7 years ago #

    If you could make it work with 14 days though there are huge benefits. Finding out how your advertising is working, how tests are performing quickly is a huge asset. (Having to wait 30 days minimum to see if a test lands you customers is tough). So there are benefits to making this work but again for all the reasons listed it has to be the right call for your company.

  • JS

    Javier Sanz

    about 7 years ago #

    Ha. I'm drafting a post about this. From my point of view, and with the little research that I have done so far, give your potential users the enough amount of days to see some results with your tool/produce/service. No split tests, no copy of other similar ammount of days from other products within the same industry, etc - that are some of the ideas here suggested.

    Identify how many days takes your product to output what your potential user wants, and play with it.

  • SM

    Stuart McKeown

    about 7 years ago #

    We recently moved from a 30 day to a 14 day trial model & the results are astounding (for us anyway).

    Typically when someone runs a campaign with us, it will almost always last ~30 days. So giving people a trial for the full duration really doesn't make sense.

    Instead if we give people trials for 14 days (half duration) it means they have to make at least 2 decisions:

    - Do I want to shorten the length of this campaign to 14 days? (generally the answer is no)
    - Am I willing to pay to continue (generally yes)

    The nature of our product obviously worked in our favour, but just thought I'd share our experiences :)

  • LM

    Lincoln Murphy

    about 7 years ago #

    Anecdotally, I've noticed that companies that switch from a 30-day (or longer) free trial because "it's not working" (whatever that means to them) to a 14 or 15 day trial sometimes find that their conversion RATE goes up. Sometimes.... not always.

    When we see that happen, often it's because they've bought into the fact that their "sales cycle" is somehow tied to the public-facing Free Trial length.

    As I've said in other comments, that's generally a bad idea... but in this case, that's a good thing. Why?

    Because having 50% of the time they used to to try to convert a prospect into a customer lights a fire under them and they work harder creating a system that on-boards, engages, and converts customers in a way the 30-day trial didn't. They might even ask for the sale!

    So what was the reason for a higher conversion rate? The "14 day Free Trial" or the fact that the vendor worked harder?

    And just because their rate went up, how were the actual numbers? Did the shorter trial length increase, decrease, or maintain the number of folks signing-up for the trial in the first place?

    Often - not always and this is just my experience - I see shortened trial lengths having a negative impact on number of sign-ups because they go so low, the customer doesn't think they'll have time to fully evaluate the product. So a higher conversion rate with lower overall numbers isn't ideal, right?

    Before you arbitrarily decide to "A/B test" this stuff, those are some things to think through.

    But there are some other awesome comments here that show the differences between every product that should further illustrate how it's best to understand why this stuff works (or doesn't) and then "test" from that baseline.

  • JM

    Jason Miguel

    about 7 years ago #

    From a customers perspective. 14 days is not enough time for me to get to the "WOW" moment for 90% of software i signup for.

    Even on a lot of 30 day trails i'm emailing the company asking for a trail extension.

    At the end of the day you have to A/B test it.

  • BR

    bastien rabaute

    about 7 years ago #

    I agree with Javier. Maybe a limited period is not the best for every product. I think it really depends of how your tool will interact with the business calendar of your customers.

    For example, if your product is used to send administrative documents usually at the beginning of each month, it's maybe better to let your prospects being able to use it at least 2 "beginning" of month. If they try it at the 10th of the month, they will have to be able to test it again at the beginning of the next month.

    In such a case, maybe it's better to limit the number of documents he can send but not the period when he can do it.

    The cycle you send them emails or call them have to be in the same rythm too.

  • PO

    Phil O'Connell

    over 4 years ago #

    We recently switched from a 30 day trial to a 14 day trial at Tettra. As with every experiment we run, we came up with the idea and then dug into the data to justify it.

    A critical part of our analysis has to do with PQL (Product Qualified Lead) triggers- thresholds of activity during product usage that indicate a user is an ideal lead for our sales efforts. We run a wiki platform, so for us our PQLs are how many teammates the creator of the account invites and how many pages of content they add to their wiki.

    When we dug into our usage data we realized these thresholds were being hit before the 10th day of the trial in almost all cases, meaning someone who qualifies themselves to be ~80% likely to subscribe is already getting themselves to that point before a 14 day trial would end.

    We made the change last week and are already seeing results. No regrets at all.