Takeaways on building a SaaS (bootstrap, 230k users, x2.5 YoY) part.1

[annies]

Good day, everybody

I'm reposting since we discussed with the moderators the format in which it would comply with the rules.

I want to share my experience and what my team has learned while product development: Saas, bootstrap, no VC, 4 years, 230k users, x2.5 YoY, team 5→21, Productivity & Design tool (B2B + B2C2B).

I hope these will be useful for makers and teams that are launching now:

1/ Freemium is a must

My opinion about the freemium model is fully shaped by the values of the Product-led Growth strategy.

Being able to use your service for free (even with free plan limits) lowers the threshold for entering the product (no need to pay, no credit card needed, no monitoring the trial period) and creates a wave of organic mentions, recommendations, and shares.

Freemium for us is not only a "function", but a tool for user adoption and acquisition. It allowed us to build a growth model based on organic and WoM drivers.

Please consider freemium as an opportunity for users to test drive your product without any pressure. Users appreciate it.

References: Notion, Slack, Miro, Figma, Dropbox, etc. (explore their freemium models).

How this worked for us: Our paid users before purchase: 95% free plan, 5% 14-day trial.

2/ Onboarding is your infinite point of growth

Onboarding success lies in understanding what tasks your users have and how to deliver value through your product as simply+faster as possible.

Think not about what you want to say, but what your users need to get (what are their pains, challenges).

How this worked for us: How we improve onboarding: demo project, onboarding emails, project examples, explanation videos + tips inside the app, product tour, student and edu plans, external content.

3/ Paid traffic at launch leads to the wrong place

Based on my experience, a strong preoccupation with paid traffic leads to negative long-term effects. If you spend all the resources (time, money, passion) on this "quick audience dopamine," other channels suffer. Paid sources don't have the "cumulative" effect that search organics, WoM, content marketing, etc. do.

By launching paid growth channels right away, you lose the opportunity to find out what your early users love and value your product for, which is critical in the early stages of development.

I always recommend leaving paid channels for dessert when conversions, value proposition, and other attributes of a healthy product are optimized and won't account for 100% of inbound traffic.

Paid traffic is the "fastest" option, but also the least useful in the long run. When the money in the "Facebook Ad Manager" runs out, the music stops.

How this worked for us: Growing to 230k users with a $0 marketing budget.

4/ Design will be even more important in 2022

First of all, I mean product design.

Watching interesting startups, I notice a low UI&UX execution level. It's a big stopper for user adoption. No-code trend leads us to unprecedented dev speed, but also to template solutions.

We believe that product design is an incredible advantage in an era of competition for user attention.

How this worked for us: One of our early investments was into "clean & simple interface": easy to understand, pleasant to work with, familiar working patterns. A lot of feedback from the users was followed.

5/ Launch priority is retention, not revenue

At a product launch, the most important thing is early feedback and user experience, not $30 you can earn. What's more valuable to you?

Don't be greedy.

You have to try hard to get your first early adopters, please don't build paywalls before the product.

How this worked for us: We had no sales at the start, but we talked to the first 50 users and it influenced the further movement. We compiled all the feedback into a document, marked it up (product value, feature requests), and adjusted the backlog and strategy.

———

I think there are ~20 points more, hope I can handle them, too → you can find the sequel in Part 2

Good luck with your Saas products!

 

[613jono]

@annies that sounds impressive, may I know what kind of users are your SaaS targeting?

I'm trying to figure out if there's different pricing strategy for SMB-targeted product or Enterprise-targeted product

 

[annies]

@613jono To tell you the truth, we are still completely bad at Enterprise sales. The product is not ready for it yet, we don't have much expertise in it yet.

I think to make a breakthrough in this direction requires a whole set of measures: new experience, team, strategy, marketing, etc.

 

[danixp]

@annies If the service is already very valuable, but not ready for enterprise - I'd suggest to find a big customer and work on their requirements.So contract will include all the missing features/services and you will obliged to implement all this in agreed time, after that - customer obliged to buy.This will work same as MVP - customer will drive your feedback loops and pay for it at the end.This is also good for customer - they will have exactly what they need, not a tradeoff as with some already existing solutions.

 

[annies]

@danixp Agree, this is the right direction.

But we have already seen that the two Enterprise requests can have completely different configurations. Each large company adjusts its vendors to unique parameters (security, compliance, user access, etc.).

Of course, over time, the cumulative effect will work for us, with this I fully agree.

But at the start it takes a decent effort (it seems to us so).

 

[sacral]

@annies Glad you are open to this. It is interesting how certain b2b saas businesses (and importantly, teams) seem to "fit" with different models. For example I was closely involved with a saas that launched with freemium and after iterating for 2 years was close to shutting down. We brought in an enterprise sales veteran and immediately starting earning customer contracts. As the business grew we continued to experiment with freemium, but eliminating it entirely and going all in on human-driven sales (of course, the major source is still organic leads to the site, but they all get processed by the sales team), was the thing that corresponded with massive revenue growth.

TLDR, it seems like freemium is best for some, enterprise is best for others, and hybrid works for some as well. It seems totally unclear to me how to predict which one is best for an early stage company.

Important note: success in enterprise requires an approach where: sellers have no shame in quoting prices that feel ridiculous when compared to comparable customers from self service (e.g. 10x+ higher in some cases), and sellers that are comfortable quoting customers based on a sense of their max budget as opposed to things like seats/storage/features/etc.

 

[annies]

@sacral Thank you so much, those are interesting thoughts!

It seems to me that the world is not black and white, so many models/approaches can exist and work perfectly. The journey of any company is through a series of transformations, where a company goes from one model to another, tests a third, makes mistakes and starts over and over to constantly optimize its sales funnel.

What works for one company doesn't work for another. But the opposite is also true: you can take someone else's practices and modify them to suit you.

Enterprise is not so much a new financial sales model, but rather an indicator of a company's maturity, when the product moves from SMB clients to large corporations.

 

[angi12]

@annies This is why freemium doesn’t really work well in the enterprise / SMB market. I have worked for 2 larger SAAS companies that tried this and it failed, horribly.

The issue with larger products is it’s very hard to show value in a free trial. First, to get the value they need to know how to configure the tool for their needs. We built out a wizard that asked a few questions that configured based on their answers. That was actually pretty easy and worked well.

The big problem is people wanted to see their data. These were CRM type products, so they wanted to look at their members and play with the tool to see real value.

We added mock data, so at least it wasn’t blank, but conversions were non existent. It still required a sales person. These are $5K - $20K a year products with implementation costs. People aren’t going to invest that without a sales person and multiple demos and conversations.

I think having a free trial might be a great way to collect leads, which then can be worked by BDRs, and turn into sales but I’ve yet to see anyone really pull off an automated growth funnel. But if someone knows of one, please let me know because I’m very interested in it!

We are trying a version of where we have much smaller interactive videos that illustrate one job the customer is trying to get done. I’ll let everyone know how this one works out!

 

[annies]

@angi12 Thanks for this great example. Yes, complex products with expensive (and long) implementations are a separate class of products working on their own models and sales funnels.

In a case like this, I would recommend not making a freemium model for the whole CRM, but rather creating separate side services that can bring micro value to the target audience. These users are the point of contact and approach for selling a larger product.

In that case, you can take full advantage of the freemium Saas service, including distribution and self-service.

 

[annies]

@613jono Thanks!

Our target audience: teams & standalone pros involved in the development of websites and digital applications. Our tools help with UX&UI projects.

There is a huge difference in pricing models between standalone, SMB and Enterprise: cost, selection factors, requirements, plan bundling, applications, and so on. I would say that they are practically different planets.

 

[danixp]

@annies Hi, thanks for sharing.

How was your conversion to paying users look like at the beginning?

What was a conversion percentage?

How long did it take usually for user to decide to upgrade?

What are your limitations on free vs paid?

Did you try monetizing free audience(e.g. ads)?

What was your steps to improve this conversion?