The idea of an MVP often garners charged, play-by-play commentary from all ends of the startup spectrum. Rightly so. Despite the necessary, proposed revisions — marketable/sellable/lovable? — it continues to be a foundational concept.
In the following exchange, Sendspark’s co-founder and CEO, Bethany Stachenfeld (@bethany), makes the case for yet another essential revision. Expanding this iterative, systems-thinking manner of building to growth funnels:
— Treating the customer funnel like a product and not just optimizing what’s already working
— The framework, team, and stack that Sendspark deploys to ship a growth bet every single day
— Notes on how they’ve approached the freemium model
— And how they’ve dealt with freemium’s core trade-off
My background is in marketing.
Before starting Sendspark, I was doing B2B marketing for other startups.
And, I feel, what founders need to know is that it’s very different to be running and marketing for an established funnel vs creating a whole new one.
Because the customer funnel is like a product. You cannot optimize it if you haven’t built it. The key idea to ponder as you’re starting out is: what’s the MVP for your funnel? And where do you want it to be in the long run?
Because maybe we know that we should have all these free tools and amazing, free-to-paid programs and offer all these incentives. And perhaps give away a lot.
But we can’t just build everything at once. We have to shape these incentives and the funnels they reside in, like products. Piece by piece. In a way that’s continuously adding value.
What I would recommend is to, first:
Identify your ideal funnel. One that you think will make sense for your business, when your product is fully built, and you have customers, and all’s going well. Figure out what that looks like. So that you can track all the key things you are going to need to get there.
Then distill this ideal funnel and sketch out what is the MVP version for where you’re at right now. And that should be based on features you currently have and shouldn’t require development other than tracking a few key events.
It’s to optimize this MVP that you want to run growth experiments. Starting with the simplest tweaks you can make. Revisiting email automation sequences. Showing retargeted ads based on pages people have visited. And others.
There’s a mistake I’ve seen a lot of companies make. Which is, instead of mapping out the ideal funnel and then the MVP funnel, they just look at what’s already working and optimize for that.
I’ve heard a lot of people give that advice, too.
I think that’s bad advice.
Because if you keep optimizing what’s working, but you haven’t actually thought about all the broader, moving parts of a funnel, you are never going to do this right.
You’ll just say, “oh, I’ve been talking to people and I guess at our stage we need to add more sales folks.” And you invest in human resources and start hiring people to fill in holes. But really, you just need to actually build out your funnel and optimize that from first principles.
For example, our ideal funnel looks something like:
→ Visits website
→ Creates free account
→ Records and Shares a video
→ Gets a win from their video (likely a meeting scheduled)
→ Records more videos
→ Hits a free limit (maybe records 30 videos or wants to use a premium feature)
→ Upgrades to Pro ($15 / user / mo)
→ Wants to collaborate with team members
→ Expands account
However, early on we didn’t have many Pro features or team collaboration features that customers would want to upgrade to use. We just didn’t have time to build them yet!
But still, we started tracking our ideal funnel, knowing that conversion rates would be low. And then we built the features to improve them over time. This way, we were always measuring what really mattered, and improving in the right areas.
As a base for experimentation, we work with the AARRR framework.
Which is really good for a freemium product, as it covers acquisition, activation, retention, revenue, and referrals. And then we’ve found and listed areas, within our product experience, that are important (leading indicators of progress) for each of those steps.
We probably have over 200 ideas classified by what stage they impact. Ways we can move the needle for different parts of the funnel. Drawn, either from someone on the team, a mentor, or a customer who is excited for us. All in a Google doc.
Every week we review our metrics, and ask, where are the biggest gaps? Which are the ones we’re trying to close?
Is it someone going from being a free sign-up to an active user? Or an active user becoming a super sticky user? Or is it someone turning into a paid customer, having been a super sticky user for a while? Or perhaps a whole team trying to expand?
Then we pick something we can ship. Some are super easy additions. Insert another video to an onboarding sequence. Push out a different in-app notification for the new cohort of users.
Some are more complex projects. Say, hosting a huge webinar with some of our customers where they gather to share unique ways in which they’ve been using Sendspark.
From there it’s just trying to find a balance between what’s possible to do in a week alongside everything else we’ve planned for. We’ve tried a bunch of different models. Some were much fancier than this one. But this is what’s working for us really well right now.
We know our funnel. We have a backlog of ideas. We just pick a few at a time.
Sticking to a goal of launching a new experiment (mostly no-code ones to begin with) every day. It might just be a simple blog post but there’s always something growth-informed coming out each day.
There’s definitely a range of work that goes into it.
But we’re doing something every day. We’re getting better every day.
When it comes to the team structure, right now, it’s mostly my co-founder and I, who are running these experiments. We’re looking at the data on a schedule and identifying the best experiments. A lot of them I might do myself, especially those on the marketing automation side.
A portion of these ideas actually go into the development cycle and become product features. But that might take a little bit longer. Shipping every day is great but you have to test things before they make it to production.
Our dev team is amazing. We have an excellent person working in customer success as well. Doing just a tonne to help different customers one-on-one. Creating video tutorials and help responses.
The idea is, we’re all about connecting personally.
So most of these tasks happen manually at first.
If we’ve sent, basically, the same personalized video tutorial to three people. We create a reusable edition, add some dynamic variables using Sendspark, and embed it in automated notes and help articles.
It’s not that we’re rushing to automation. But we automate relatively quickly. It’s more like: Send out a few things personally. And then if they seem to be resonant, weave them into an automated system, so that we can move on to the next thing.
Some of the tools we’ve been using are: HubSpot for a lot of automation on the marketing side. Even our website is built on HubSpot and we use their email sequences. We use Intercom for a bunch of in-app notifications. We use Amplitude to measure data.
We obviously use Sendspark to record the videos and embed them in different platforms. Right into HubSpot emails, Intercom messages, Gmail, or wherever we’re sending things.
And honestly a tonne of Google Docs, too!
This iterative process definitely informs how well we’re able to scale with a freemium model.
Here are a few other ideas that have influenced Sendspark’s free plan.
How does one draw the line between free and paid features? From what we’ve observed, the free parts and the paid ones are all quite different from what they were two years ago and they will be quite different, yet again, two years from now.
The pricing choices, then, must reflect this consistent evolution of the product.
In terms of building competitive advantages, you should try to make sure that customers can access free value in all the features that your competitors offer for free. Plus at least one other thing that’s really cool and different, and makes them choose you over others.
Similar to paid packages, there’s a lot more uniqueness you can serve up in a free plan.
The way we’ve done this so far at Sendspark is that the standard video recording and sharing features are free. The thing that’s uniquely different is that you can embed the videos in different email platforms. For people who want to have videos play in-line, that’s a much-sought option.
Another helpful way to approach the freemium product strategy is that you don’t have to make your core product free. Many companies (SEO software, for example) have free tools that create high-intent virality and drive it straight into the product. As opposed to stripping away parts of the product and putting the onus of validation on select features.
Overall, you don’t want to be the cheaper or the free version of X.
You want to be the best for a very specific group of people who have a specific problem. And then ideally, your market gets bigger over time and you get to solve for more personas/use cases or both. But you really have to start by doing a unique thing really well.
The trade-off with freemium is one of reach vs focus.
On one end, we’ve found that the reach has been worth it. Sendspark is an inherently viral product. The essence of it is that you create videos and share them with other people. So we really want to get a lot of people using the product and sharing it as often as possible.
That being said, we’ve had a lot of additional support to tend to and it can sometimes be harder to focus because you have so many people (who aren’t always your ICP) requesting all sorts of different features and also those who might never become paying customers.
With more users, you get a lot more data. A lot more feedback. People write in and have ideas. And we have such amazing customers too.
The number of times a customer says something like, “I love your product, have you thought about doing X? I really want to see you succeed,” is mind-blowing. I never would have thought of sending that to a company before. But now, we get notes like that all the time.
The only caveat is that you just have to make sure that you’re bringing in the right type of users. So you want to do a little bit of cohort analyses and maybe ask some questions upfront.
Ask, how do they plan to use the software? And if they say something that’s not exactly what you’re building for, you kind of filter out the noise.
When people submit feature requests, instead of just instantly adding them to a log, we dig into how those features are really going to help a customer and then make a value judgment.
For example, some people sign up to use Sendspark for education, mostly to record courses. Sendspark is certainly not the best platform for that.
My co-founder and I started the company because we were both working at another company that was very good for creating videos for education. We left intentionally to create a video platform specifically for conversational video messaging with customers.
If you’re looking to make a course, or if you’re looking to do a high quality edited video, we’re not the platform for you. We are only focused on helping you build customer relationships. .
It’s important to not lead customers on. Instead of trying to make every use case work, we recommend better software for other use cases.
Otherwise, what would happen is, they will attempt to use Sendspark. It wouldn’t work for them. They’d have a bad experience. Some will complain. Often in public. Other people will see those reviews and conclude, ‘oh, it’s a horrible company.” You don’t want that spiral.
And it’s worse with freemium.
Probably the biggest danger of having a free plan.
You want to make sure people know clearly what they’re signing up for, beginning at the sign-up page. Making sure the messaging is very specific.
That’s definitely been a lesson as we’ve built out the free plan. We’re not filtering people out by having them pay, but we can filter people out by using better messaging and being proactive early on when it comes to support requests.
Having too many users sounds like a champagne problem. But it’s not. Really. We need to know these are the people that we’re focussed on and we’ll continue to optimize for them. Other people might sign up. And that’s fine.
But we’re not going to encourage them further into the product.
A nuance on the freemium virality bit, too, is that Sendspark is primarily B2B focussed.
There’s definitely some individuals who sign up and use it, but we aim entirely at businesses. Specifically teams that are constantly communicating with their clients: Sales, customer success, and others.
So whether virality benefits us or not depends on who our clients’ clients are. If our clients are reaching out to other B2B sales and support teams. That, of course, has viral distribution potential. But if they’re reaching out to software developers only, then not so much.
Then another form of viral distribution that helps Sendspark is an internal one. Different people, encompassing entire teams, at the same organizations.
Virality is more than direct referrals. We also benefit from our many users writing public reviews about Sendspark.
We already have over 500 reviews on the chrome store. Around a 100 on G2 and other places. All signaling that people are using Sendspark. They’re talking about it. They’re sharing it. That definitely adds value. And it exists because of the reach a free plan affords us.