Failing despite getting 301 users
You know that feeling. When you read a post on IH about someone making more than $10k within minutes of launching their product. You're excited for them, envious even. Slowly but surely those feelings you had turns into despair as you self reflect on how your own products/services are doing.
If you can relate to that feeling above, know that you're not alone. Even as we welcome our three hundred and first user to https://fakejson.com, numbers mean nothing when compared to the actual metric that matters. For our SaaS that number is our recurring monthly subscriptions. For all the excitement that comes from an update about a new user signing up to receiving valuable feedback on how better to improve the product, all we have to show for it is 1 recurring paying user. Uno user. Perhaps we need to pivot, perhaps we need to change our pricing model. Perhaps our approach to a low touch SaaS isn't right.
We know the journey into entrepreneurship and financial freedom isn't going to be easy. We know we have to temper our expectations against the realities of every big hit we hear about, there being 1000 failures we never hear about. This post is more a shout into the void for fellow hackers hoping to make it. Know that you are not alone. For us personally, our failure serves as our motivation towards producing a quality service loved by the masses.
Don't target freelancers and solo devs. They don't pay unless they get enormous value and can't live without. Focus on teams, companies with +5 employees, etc.
Another issue is that there are many free libraries that do this. Even though they are not that powerful, developers like to build their own thing as they like, if it doesn't take too much time.
You need to pivot, my opinion.
Can vouch for this. I'm a contractor and I very very rarely pay for products. There's just too many free options out there.
That may be the crucial piece we were missing, we misjudged our target audience. Thanks for your feedback Crysper.
Looking at the product, I'm not sure it is a good fit for companies that have money to spend. Why would a company with resources need this? At first sight/thought, this smells like hackathons and side projects.
What user personas do you have?
Since you are working on JSON, I'll leave an idea here in case you do want to pivot:
Plenty of bigger companies have data feeds from partners. Reliability is important. At my contract job, I realized I'd find it useful to have a SaaS service to which I could send occasional samples of documents the systems are processing. Json, xml, html, csv, excel, etc. Stick to the easy ones to start, such as json and csv. Your SaaS should alert me when the structure of the document has changed. Have keys appeared or disappeared? Has the structure of a url listed in the document changed? Numbers became strings containing numbers? Lists that always were non-empty started showing up empty, or the other way. You get the idea. You're expecting a steady format and changes are pretty obvious. I don't need to send you every single document instance. One sample document every five minutes to scan might be enough if my system is processing 10 documents per minute. Once I send you 30 or 100 samples, you have a clear picture of what the structure is, if it is indeed stable. Or you can work on the parts that are stable and ignore the unstable parts of the document.
Thanks Heliodor for the feedback. It honestly started as a side project because of prior side projects that required us to tons of fake data and test with them. That's when we thought we could make something out of it. You have a point on companies with resources, why should they spend money on us. The reason isn't very compelling for them.
Right now, we're trying to target individual developers or freelancers working on multiple projects. However, we realize our user personas work on projects on and off and the usage of FakeJSON comes and go with their projects as well. Perhaps that's why our subscription model hasn't been working as we intended.
Interesting thought on a quality assurance data feed checker. How do you go about solving for this right now assuming that there isn't such a solution in the market?
The value metrics of your pricing seem totally arbitrary and hard to get the head around. I wouldn't know which plan to pick even as experienced dev. Why not have one paid plan only and one simple value metric. Reading about value metrics on priceintelligently blog helped me a lot with pricing for SaaS.
I am with Heliodor on the user personas. Targeting Freelancers for something that their customers should buy / pay for seems to me could be the wrong angle.
Am I right that fake data is something you need temporarily until the backend and frontend fit together? I think you are setting yourself up to failure this way, because your product is based on natural churn when you don't need the fake data anymore.
thank you Benjamin for your feedback. We went with multiple tiers because we didn't want our users to overpay for things they wouldn't find useful. Maybe we'll give it a shot at simplifying the pricing plans. I'll definitely take a look at that blog.
You helped clarify something which I couldn't put my finger on. Freelancers would pass the cost back to their clients when it comes to web hosting or purchasing a domain. Now I'm thinking, we've got to tweak our delivery towards allowing freelancers to better pass the cost back to the clients themselves.
Yes you are right regarding the natural churn once obtaining the fake data and having both sides working. However, we also offer rigorous app testing with HTTP responses and naughty strings designed to easily expose defects in one's app.
What kind of software engineer is looking to spend money on projects that aren't making money? I don't know of many! I sure am not. Maybe you can target some sort of persona who has run out of time and needs a shortcut?
So far, I have implemented metrics for the data feeds and I have set up alerts for when each data feed fails in various ways. A failure via changes in data structure would be caught as a generic failure after hours or maybe one day. I'd have to find the cause through the metrics charts and the logs. Having a data structure QA would tighten the reliability quite nicely. It would shorten the alarm gap and pinpoint the problem too. I could sell my boss on this in the blink of an eye. Want to check it for viability before you spend time on it? Go run it by sizable affiliate marketers.
Definitely not me long term (but sadly yes for now)..We'll try to figure out how can we rejig FakeJSON to better cater to our target market. The feedback we get from our users is that it's useful and they're happy with it but perhaps there's just not enough of an incentive to proceed with the paid plans.
Last question if that's ok, do you use other paid SaaS like a data feed checker at your workplace?
I looked at your pricing plan. Do any of your customers hit the ceiling of your free tier or is it too generous? You have 300 users, so you should be able to analyze their usage to have a first take on boundaries for your pricing plans. What should you segment on and what quantities do you put in each tier?
Is there a way I can message you? I don't see one in your profile.
hey Heliodor. We have a couple of users hitting the free tier limit but the majority definitely have quite a lot of leeway left in their limit.
Feel free to message me on twitter @jonleehl
Maybe change of perspective for you on pricing: instead of subscription try credits. Lets say 10 000 credits for 10 USD. 1 credit = 1 request. Minimal order may be 5 USD (5 000 credits).
If they really plan to use your service, they wouldn't mind giving one time payment when needed.
I think that your business is not fit for subsription based model for most of the freelancers. I personally wouldn't want to pay you monthly if I only have specific project. It just isn't worth it. It is too much of a committment. But I would be willing to pay one time (and again if I need to recharge credit). It's like mobile operator monthly plan vs. credit plan. Freedom is everything in a world where everybody wants subscription from you.
"For our SaaS that number is our recurring monthly subscriptions."
Are there many other metrics for SaaS?
Anyway, transforming the user feedback into useful features is tricky, at least to me. Users often, well, lie, for a lack of a better word. They may say they need a featue that they don't. They may say they would pay for a solution and then not pay. This is probably the hardest part, I'm personally struggling a lot with this.
I thought I would learn the lean methods and things would just be as in the books. Boy, was I wrong. Talking to your users and building features that they actually NEED and would PAY for is the real art of the craft of building a SaaS. I'm thinking it might take me another 6 months to a year to get some basic knowledge of how to do that.
Keep your head up my man, you're doing good, I see you don't mean to quit so you're gonna figure it out sooner or later. Good luck.
Did you talk to anyone, live (phone OK) before making the product?
Have you talked with any of your users, including the paying one? I mean actual verbalization, not chat/email.
Hey Sean. I haven't spoken verbally to any of our current users outside of my circle of friends who have tried it. In terms of chat/email, yes I've had some feedback from our users. Not sure where you're going with this?
"Not sure where you're going with this?"
If you do in fact see this as a venture you want to grow, then you need to figure out why your free-tier users aren't converting to paid users. The only way to do that in the early stages is to talk to them.
"Early stage" varies, but for most SaaS products is generally the first 1,000 paid customers or so. After that, it's typical that you've figured out at least one narrow persona to go after and how to land them as a paying customer.
Until then, get talking. Video calls are better than voice-only, voice is better than chat, chat is about equal to email, email/chat is better than nothing.
"Perhaps we need to pivot, perhaps we need to change our pricing model. Perhaps our approach to a low touch SaaS isn't right."
You said it right there in your original post - it's all a guessing game right now. If you want to know where to take things as a next step, you have to talk to your users.
Hey Jonathan,
In the spirit of "no failure, just feedback", I will share a few thoughts from the perspective of a copywriter:
Hey Cynthia, thank you for the feedback. Definitely taking all this in as we redesign the site. On point 2, I'm reaching out to some of our users right now!
Happy to help - good luck, and congrats on getting 300+ users!
I had an idea to build something similar myself. Primarily for integration/e2e testing, so I could stub out various services when running tests as part of build/ci pipeline. Not sure who's your actual user.
hey Tadas. Glad to know that we weren't completely crazy going into this niche. Yeap, the consensus seems to be that the target users aren't exactly well defined which we have to work on.
Surprisingly we both have a coincidences today.
Today I got my 301 free user on my webapp Watermark.ink
After alot of effort,time and money into this project I am in a similar situation as yours.
thanks Prakis. Similar milestones, similar predicament. Will follow along and see how it goes :)
Hey, I'm a little bit lazy but how this is better than https://github.com/typicode/json-server?
Hey Dalius. Json server is great if their existing data set is sufficient for you or if you're comfortable working with APIs to essentially generate your own mock backend server. What we hoped to do with FakeJSON is to abstract away all the hard work needed for setup and easily act as a mock backend through our API.
Double check json server - all you need to do with it is to pretend that you have API (zero work or writing JSON file if you need some seed data).
FakeJSON has a use case, but I think it's a service that I wouldn't use more than 5 minutes on getting up and running. Do I need an account to get started and then make a decision on what plan I want? Ideally, I'd enter my credit card and instantly get an API key to my email that I can use straight away. Usage-based pricing would be simpler too.
And that's our end goal with it as well. How to use it should be a no-brainer, and if it isn't, we've failed in that regards. When someone signs up, they are automatically placed into the free plan bucket. An API token is available straight away and they are free to start using it however they please. We only ask clients to enter a CC number when they decide to upgrade their plans. Thanks for your feedback Hannu
I think the product can have its niche, but I am not the one to ask about how to promote that.
It reminds me of https://mailtrap.io, which has been around for years now, so it must be profitable I think.
Just yesterday I was looking at the Gumroad Ping page, and it lists RequestBin for inspecting: http://requestb.in, but that's a service that's been shut down, and you have to deploy your own open source instance on a web server just to test things out.
It's kind of the exact opposite of what you do, test an API output rather than provide the input.
There are probably similar alternatives, but maybe if you build something equivalent and ask them to be on that page instead of that dead project, it's an easy source of leads. Heck, you could just deploy RequestBin somewhere and host it on a subdomain.
No idea if that page brings traffic, but just a random idea while I sip my first coffee of the day
That sounds like it'd be a candidate for a new side project. Cheers to you as I sip my first coffee of the day as well.
Nice self-realization post, We all been there (still there). One most virtue for all startups is patience. Just give it a while and it will pick up eventually !
Nothing happens in a vacuum. They either determine the issues whether in sales, marketing, product, pricing, etc., and resolve them, or they wither and die.
Thanks Naveen. Tough journey indeed. And the worst part is that we're usually too self-absorbed in our work to even realize how we can better improve it.
(Before reading comments)
why failure? What were your goals for this project? This is not clear in your post, seems to be money, but how much time this service is up? One month? One year?
At first sight your idea and presentation is awesome, from my point of view (as developer), but I think you should consider some points:
This comment was deleted 7 years ago
You have valid points in both. For bigger companies, they would have an in-house team doing validation as well as utilizing more comprehensive and expensive solutions out there.
For a solo dev, the free band would be sufficient for generating the data they need.
We'll try to find that sweet spot between the two audiences.