This week we hosted the very first english version of SaaSCast, a live video podcast (webcast) where we talk about SaaS topics: growth, customer success, admin, everything you can think of.
In the past, we had the pleasure to host french speaking episodes with great companies such as Salesmachine, Segment or Dougs.
For this new version, Maxime Berthelot, Product Manager Marketing at Buffer joined us for a 40 minutes live session.
We talked about how Buffer handles growth differently from most saas and how they use their customer feedback to enrich their growth experiments reservoir.
I think it’s interesting because we have a very different of vision of growth compared to a lot of companies.
The growth team at buffer owns the flow of customers that comes in and out. We also work closely with the marketing team but our core function is to take care of the customers once they signup and during their lifecycle until they churn.
Our mission really is to show the value of Buffer to our customers. Our job is not to find some hacky way to drive traffic to the website.
Buffer is a really customer oriented company. We try to get a deep understanding of how our users are using our apps.
To do so, we have a team a customer researchers who basically spend a lot of time with our customers.
We mix this qualitative data that we get with quantitative data from our 3M signed up users thanks to our data science team.
Those data points enable us to detect patterns and create solutions to better adjust our growth experiments.
We don’t ship features, we modify the actual features and the flow to bring customers a better experience of buffer.
Thanks to the amount of traffic we’re getting we only need a few days or a few weeks only to validate an idea or experiments.
I think they should. Usually we don’t get that much traffic on small portions of the funnel. We have more at the top of funnel and less at the bottom.
So when we try to iterate on features at the bottom of the funnel we don’t get as much traffic.
But it’s still interesting to dig in that data anyway. Specially if you have a low volume, then you should focus on gathering some qualitative data.
We’re using Hotjar to get some quantitative and qualitative data. It’s really cheap, but the point is you should talk to your customers and jumping on a call with them is even cheaper.
We’re really inspired by Brian Balfour.
First step is to understand every stage of your funnel and see where you’re getting the main drop-offs.
Once you find your target issue, you pick up the numbers and you spend 40 to 60 days trying to increase that number.
For every experiment I’m going to launch, I will try to increase that number.
Every week you’re going to come up with experiment ideas to improve the income (e.g the steps in a trial conversion funnel) to eventually increase the outcome (the trial conversion rate).
You basically you want to improve all the those small numbers to increase your final outcome. This is exactly what we do at Buffer, we use the qualitative/quantitative data we mentioned earlier to brainstorm and come up with experiments ideas.
Then we design the flow with designers and engineers to understand the cost of shipping that experiment. We aim for the best ratio cost/impact.
At the beginning we were trying to optimize the little things like a CTA but eventually we realized we needed to have a higher vision of the global flow.
We needed to that to better the understand the customer experience and make sure we were bringing the best experience possible.
We went from shipping 5-10 experiments a week to thinking about what would be the most important things to test, even if that’s one big experiment.
Also you’re going to ship a lot of experiments eventually, and a lot will fail. And sometimes it will be tough. But every time you will learn something new. Either you succeed or fail.
It’s totally by hand, the foundation of the customer research team is the customer development book called Lean Customer Development by Cindy Alvarez, Director of User Experience for Yammer.
Once you have reached a dozen of users and actually talked to them you may start noticing some patterns. Then we try validating that pattern with data and dig deeper.
One of the biggest success was when we shipped our tracker bar during trial. People had to fill in their profile and do certain actions to fill up the tracker bar.
We came up with a very rough first version. Even with that first version we managed to get 20x more profiles connected and team members added to the account.
By changing that we knew we would have an impact on our trial conversion rate.
The conversion window was about 14 to 30 days since the trial was ending after 14 days but just by looking at the behavioural data you could notice the change in the first few days.
The thing is we are really ok with having free users. We’re cool with the fact they might use the free plan forever. If they’re happy with the product then we’re happy.
We’re not pushing everyone to upgrade because some of those guys just don’t necessarily need our premium plan. If they’re happy with the product they will talk about it anyway.
It’s a community really as well as a reservoir of leads for future opportunities.
Usually people start using Buffer for themselves and then they upgrade when they start using it for their own business.
Ship early and regularly a few experiments to build a certain growth mindset. It will help you build the process and frameworks you need. Then you have to really focus on the small incomes in order to increase the outcome (e.g the trial conversion rate.)
Hope you guys enjoyed, next episode of SaaScast will be hosted at the end of October so stay tuned! We will talk about predictive analytics and sales with Madkudu.