We launched the public version of Headjack about 2 years ago, and have changed our pricing multiple times during that period.
Finding the right price point for a SaaS product is hard, but crucial. In the beginning, all you have is a hunch, pricing pages of competitors, and an endless amount of highly interesting, but often contradicting blog posts (we liked this one). In the end, it is mainly trial and error.
We can’t be certain we got it right this time, so in this post I will go over the reasoning behind each price change, and share what we learned from it, so you can draw your own conclusions.
Headjack is basically a content management system for virtual reality video. These videos are at least 4K in resolution and therefore have huge file sizes. When we then estimated the amount of views one video would receive, we soon realized that we would be faced with enormous bandwidth costs, and the only way to make this work was to make bandwidth the key factor in our pricing strategy.
We knew of another company who was faced with these large video files, Zencoder, and decided to base our pricing on their model.
Zencoder had a watermarked FREE plan, as well as 3 paid plans. As you upgraded your Zencoder plan, you received 10x the amount of data (minutes of output video in their case) while the price of your plan “only” went up about 7x. The price per additional minute also went down with each upgrade.
We tried to follow suit, but first we had to check what price Amazon would charge us for hosting our user’s files, as well as the download bandwidth they would generate. It turned out it would cost us about $0.11 per GB.
One great thing about Amazon is that if you use enough data, the price per GB drops dramatically, as low as $0.05 per GB! We expected our platform to easily consume dozens of terabytes of data per month, and so we believed we could earn a good amount on the data, while not charging users more than they would pay if they hosted their files on Amazon themselves.
This resulted in the following pricing model (we couldn’t find any screenshots of our earliest pricing pages anymore unfortunately, and yes I did use the Wayback Machine).
|100GB bandwidth||1,000GB bandwidth||10,000GB bandwidth||100,000GB bandwidth|
|10GB storage||100GB storage||1,000GB storage||10,000GB storage|
|$0.11 per extra GB||$0.10 per extra GB||$0.09 per extra GB|
|1 app||3 apps||10 apps||Unlimited apps|
The $99/mo plan was popular, but none of the other ones were purchased, and many users told us that the price jump between the plans was way too big. Also, we greatly overestimated the amount of download bandwidth that users would generate. VR video is simply not yet at the same level YouTube is at when it comes to views, and so the bandwidth and storage packages turned out to be way too big. It was clear we needed to make some changes.
We (naively) thought that showing the ridiculous amount of $7999/mo on our pricing page would result in big companies taking us more seriously, even though we only just launched. Now, to make the price jump between plans a bit more realistic, we removed the overpriced Enterprise plan, and instead added a Studio plan in between the Pro and Agency plans. Our pricing now looked as follows.
Our new sales manager also mentioned that LinkedIn gave users the option to try their Premium plan for free for one month, and we figured this could work for us as well, since the Free plan contained watermarked videos and so was not really usable.
Therefore, we created a “Try Pro for Free” coupon that anyone could use to upgrade from Free to Pro for one month without being charged, giving them the possibility to fully try out Headjack without a watermark on their videos.
This method did really boost conversion from Free to Pro, but also resulted in an enormous amount of churn, which created very muddled sales figures. However, our Monthly Recurring Revenue (MRR) was steadily increasing, and so we didn’t worry about it too much.
At this point we only had monthly plans, and especially for some larger clients this was not ideal. Large companies tend not to like monthly charges, but would rather just pay a larger sum once, since they work with budgets that have to be allocated within a certain period.
For this reason we added the option to pay annually as well, for which users would receive a 20% discount on the total price. For a SaaS company, annual contracts are very valuable, since it means that you have a good amount of cash flow early on, and can also be certain those users won’t churn for at least a year.
But it did not end with adding annual plans. We also started worrying that at the current price point, our business plan might not actually be sustainable. Let me explain what made us think that:
The Pro plan costs $99 per month and includes 1,000GB of bandwidth. If a user uses the full 1,000GB, this will cost us 1,000 * $0.11 = $110. In other words, we would make a $11 loss, and this does not even take into account all the other costs that are involved with developing and maintaining a complex platform like Headjack!
And since we wanted to offer a discount for annual plans, the potential loss per month would become unrealistically high, and so it seemed like the only way to ensure our company’s long term survival would be to increase the price or to decrease the amount of bandwidth we offered. Increasing the price did not seem like such a crazy idea, since one of our main competitors was asking $299/mo, or $199/mo when paid annually.
Therefore we decided to change our Pro plan to $249/mo, or $199/mo when paid annually. This had the added benefit that the gap between Pro and Studio became smaller, which we thought would make it more likely for people to upgrade eventually. Headjack also aims to be a high-end solution, and a low price did not seem to be congruent with that image.
Users that already had a $99/mo Pro plan would continue to pay that low price for another year, after which we planned to upgrade them to the new price. In the end we decided that we would not upgrade them and keep these users at their original price. We also offered existing users 6 months of Studio for the price of the new Pro plan, which amounted to a $1200 total discount. Instead of increasing the price of the Studio and Agency plans, we lowered the bandwidth for those particular plans, since no user was coming close to reaching those limits.
We saw a large dip in automatic conversions after the price change. We hypothesized that if users could only test the product with a watermark in it, then this might not be convincing enough for them to upgrade to a $249/mo plan. But without the watermark, some users would never have to leave the Free plan at all, right?
When researching online, we came across a story which was pitched as:
The accompanying graph was also telling:
Mealime, a meal planning app, decided to open up their Free plan, which resulted in an initial drop in MRR, but a way faster growth rate. This convinced us that we had to had to remove our watermark from the Free plan (read more about that move here).
Without the watermark to distort the videos, users on the Free plan could now actually use their Headjack apps in the field. We believed this would allow users to show off and sell Headjack to their clients, after which they would eventually switch to a paid plan. We also gave users access to our Cinema remote control solution from the Free plan already, so they could test it out and make a more informed decision about whether or not to upgrade.
To further assist conversion, we added PayPal as a payment method in addition to credit cards, since a lot of European clients don’t have a credit card.
After we removed the watermark and promoted this through our various channels, we saw a short boost in signups, but automatic conversion was still not really picking up. When researching how to improve our sales, we came across a highly informative video on SaaS funnels by venture capitalist David Skok:
We started reading more articles from David on his blog For Entepreneurs (especially this one, this one, and this one), and became convinced that we had a good product, we had product/market fit, and that all that was left to do now to win the startup game was to fix the bottlenecks in our SaaS funnel and build an “automatic sales machine”.
We started by making it easier for users to find resources and to learn how to use the platform, as is evident from the changelog of May 2018:
- Added improved onboarding tour
- Added FAQ behind (?) button
- Added thank you page after signing up with links to resources
- Added account setup progress on Dashboard
- Added links to resources on Dashboard
- Added links to tutorials on App Signing page
Then we redesigned our entire frontpage from a customer perspective. This took way longer than expected, but the result is a clean, informative website that can now be measured and optimized. We also simplified our data packages to “small”, “medium”, “large”, since users don’t really want to deal with gigabytes.
We were so caught up fixing our SaaS funnel and building the new frontend, that we forgot to look at the actual figures. We worked on several big custom app development projects during 2018, which included annual Headjack plans, and so we thought we were doing quite well.
However, when the dust had settled after launching the new website and after finishing our largest project to date, we finally noticed that automatic conversions had still not picked up the way we had envisioned. The new website helped, but was not a magic bullet. Had the price increase been a bad decision after all?
We sat down with our sales manager, who speaks to our users on a daily basis, and he said that the vast majority of companies he speaks to are either freelancers or small VR startups that were unable to afford $249 per month. When we soon after read Lean Analytics, which said you should price according to what your customers can afford to pay you, we knew that we had to lower our prices again. The Pro plan is back again at $99/mo, and costs only $79/mo if paid annually (existing $249 users will of course be migrated down to $99/mo).
For monthly payments this keeps it just under the psychological barrier of $100, and for annual payments it is just under $1000. This did increase the price gap between Pro and Studio, and so we made Studio more enticing by giving that plan unlimited apps, even more bandwidth, and included one free app store submission. We also removed the Agency plan to further simplify our pricing, and instead highlighted the custom work we can do for clients who require advanced features.
But what about the data costs being higher than the price of the plan? This is still true, but on average users don’t use up all their data, and so there is still a margin there, although we should keep a close eye on this in the future as the VR market grows and the number of views increase.
And is the gap between $99/mo for Pro and $449/mo for Studio not way too big? It felt that way to us at first, until we looked at the pricing models of larger, established companies like Unity, Hubspot, and Contentful.
Unity’s Pro plan costs 5x more than their Plus plan.
Hubspot jumps from $50 to $800 dollar; a 16 fold increase!
However, Contentful takes the cake! Their Micro plan costs just $39, but their next plan is a whopping $879, or 22 times higher!
We have clearly been wrong on multiple occasions about our pricing, and this has cost us dearly. However, you have to go from A to B before you can see C. You only learn by doing and by giving it an honest try, the only thing that matters is that you learn from your mistakes and keep trying new things to see what sticks.
We can’t say for sure that our latest price change will finally be the right one, but it does feel like our pricing is now better aligned with our actual users than ever before.
On top of that, we tried everything we could think of to make the higher price point work, so now that we lowered it, things can only be amplified!