10 Years, 10 Pricing Mistakes: What We Learned Building a SaaS Without VC Money

Back in 2015 we launched our first SaaS with no VC and little experience. Most pricing choices came from gut feeling or advice we shouldn’t have taken. Looking back, what feels obvious today wasn’t obvious back then. We learned the hard way.

Tiered pricing.

On paper it looked fair. A team with 15 users paid $39 but adding one more jumped it to $119. Inside a tier customers were happy as cost per user dropped. But crossing a tier made them furious. People churned, demanded discounts or sent angry emails. Instead of rewarding growth the model punished it.

Per-user billing.

Finally revenue scaled with customers. But it attracted many solo users and tiny teams. Because Everhour syncs with project tools, even one freelancer could flood the system with hundreds of projects and tasks. They paid almost nothing but created huge load and were the loudest group: angry tweets, refund requests, endless tickets. Larger teams were calmer and more loyal.

To fix this we set a 5-seat minimum.

You could use fewer but still pay for 5. It filtered out freelancers, raised our average check and aligned us with the right customers. Some complained it was too expensive but those who valued the product stayed. Later we added a free plan for up to 5 users without integrations. This gave small teams a way to stick around and got us listed in “Top Free Tools” blogs. It worked more as marketing than monetization.

Flat-fee pricing in our Trello add-on.

$10/month, unlimited users and projects. Out of 30,000+ active users only ~500 paid. Everyone else left the moment we asked for money. The product was too simple for upsells.

Grandfathering.

If someone signed up on old plan they could keep it forever. No forced upgrades, no surprise hikes. Many are still paying today. We tried nudging them to newer plans with extra features and loyalty discounts, but only a small percentage switched. If their plan worked fine, the barrier to paying more was too strong.

Discounts.

Volume disc for large clients worked well and helped close deals. Urgency promos like “Sign up in 5 days and get 20% off” gave a small conv boost and improved retention. Those customers stayed longer because they didn’t want to lose their deal. But it felt unfair to existing users. Some even tried canceling and resubscribing. We eventually stopped.

Billing mechanics lesson.

First we billed per user and deferred prorates to next cycle. Logical in theory, messy in practice. Customers canceled before renewal and we lost money, especially on annuals. Others didn't recognize totals on bank statements. We tried charging prorates immediately when someone was added. That fixed revenue issues but created too many micro payments and invoices. Customers hated multiple charges when swapping users.

Eventually moved to per-seat billing.

Buy seats in advance, use them flexibly and the invoice stays clear. Still not perfect. Some forget to remove empty seats and complain later. But overall it’s far cleaner. Today it’s basically the SaaS standard.

Lessons.

- Tiered pricing looks fair until you cross tiers. Then it makes customers angry. - Solo users overloaded our system while bringing little revenue. - “Micro” plans added support costs without conversions. - Flat-fee pricing attracted 30k+ users but barely any paid. - No minimum seats meant tiny checks and the wrong audience. - Multiple plans added complexity. Most picked the same one anyway. - Upselling grandfathered users failed. If old plans worked, they wouldn’t pay more. - Discount promos helped short term but felt unfair and were abused. - Per-user prorates turned invoicing into a mess and cost us money. - Free plans drove marketing visibility, not upgrades.

This is near the 4,000 char limit here. More details and screenshots: https://medium.com/@citizenblr/lessons-from-10-years-of-saas-pricing-experiments-4ed45f552171