Show HN: Use landing pages to test your product | Hacker News Reader

Show HN: Use landing pages to test your product

quicklytest.it

43 pointsemoray1337 comments

Hey. Congrats on shipping.. that's more than I have been capable of this year.

Lots of constructive things I want to point out. Don't think I'm bashing on the idea... (But you did post it as a Show HN so like most hackers I tried and find out how things work).

- I successfully created an account without paying. (Contact me through my email if you want details. You can find it in my profile).

- You write all of your clients details to disk. I can view them all here. http://learnmoreabout.it/ I can also overwrite other peoples tests. (Not sure if you are aware of this). (Could also be used as a very easy denial of service vector killing available inodes)

- Being able to access stuff like this is wrong. http://learnmoreabout.it/usr/share/base-passwd/group.master

- You have an active RAM disk which is readable through HTTP. I've been able to download and mount and read from it.

- You've got an SQL injectable page. (Took a while to find one, but it's there)

- You are doing validation that should be done on the server, via the client. (ie. Checking that the name doesn't contain invalid character.). So this ends up with people writing to directories they shouldn't. See the file directory 'thisShouldNotBeHere' within your web accessible /usr directory? http://learnmoreabout.it/usr/share/mysql-common/

I've uncovered way more information than I ever should have been able to (it was almost as fun as the Strip CTF). If you want any advice, I'm always infront of an email client... (and PHP is my main language)

You are definitely at a very early MVP stage, but already I'm reminded of this blog post from last week: http://blog.ryankearney.com/2012/10/never-give-your-informat...

And, in case anyone says I should have followed responsible disclosure, I prescribe to the same ideals as Silhouette in this comment. http://news.ycombinator.com/item?id=4619657

It was fun replicating the things you found. Knowing something is possible makes it a lot easier to do.

You did pretty good in 18 minutes.

By the way, MVP refers to features. Failing basic security practices like this has nothing to do with whether something is an MVP or not, it's just sloppy/irresponsible coding.

I think I may have spotted an XSS vulnerability also, but I wasn't sure.

I'd be grateful if you could drop me a quick email to discuss? questions@quicklytest.it. Thanks again.

Hi,

First off let me appolgize about the sloppy coding. Unfortunately I'm not a programmer, I outsourced the programming.

The information you posted is invaluable and I'll drop you an email shortly to follow up.

It's dishonest to tell visitors the queue is due to high demand (what message do they see if demand is low?) and fail to disclose that the service does not exist and may not even be feasible at the offered price point. This kind of chaff cast into the marketplace makes it harder for customers and legitimate service providers to find each other. Market research is fine (with consent) but it's not a license to deceive the public about what you're doing.

I'm curious as to what others think of this.

I've been using the same technique to test a market and have been struggling to figure out how shady this is. One on hand, it does feel like we're deceiving people, on the other hand, what we're doing is actually benficial to consumers. If they signed up, that means they want what it is that you're trying to build and they're letting you know it.

Why not just state the product isn't ready yet, but you'll contact them as soon as it is?

You'll lose people that are mildly interested and clicked "buy" out of impulse.

I've heard it put before as the difference between telling a friend / stranger about a business idea and saying "would you buy it" which will often result in a non-committal "yeah" to, "great I have a load of them in the car, £5 each, how many would you like?"

Only the second is a real test of if the idea is valuable and viable. This way you end with with a 'hot' mailing list. You also can't worry about losing these early potential clients, if your client pool is really that small then this is not the way to test. There is a great section on this in the book 4 Hour Work Week.

I absolutely hate this trend of creating landing pages for bullshit and nonexistent products. Turning the public into sheep and testing ideas with no plan to execute on them is unethical, yet I hear it all the time from garbage theories that generally have the word "lean" in them. It's crap.

Is it better for someone to spend months coding something only to discover that nobody wants it? What do you really lose if you find out that a service you like isn't ready yet?

Great point Stavros, I believe this is the biggest source of failure for startups.

Hi Prodigal,

Although chrisacky flagged up some major issues in security above, there is more than one server/ domain running the system which they didn't discover. Once there is around 50 users on a server / domain I've created a new one.So thankfully chrisacky only managed to discover a small amount of users active on the service currently.

For easy landing pages you own, I wrote a thing:

https://github.com/skorokithakis/landing-page

It has email list subscription functionality, just drop two templates in, and you're done. It uses AppEngine, so it's pretty much free to host, and easily extensible.

That's pretty cool, will definitely give that a try

Looks cool, but if you ask me a monthly subscription model isn't optimal for this kind of thing because the churn[1] rate will be so high.

Why do I expect a high churn rate? Well, there are two options. Either the splash pages created with quicklytest.it get some traffic or they don't. If they don't get traffic the user will get demotivated and stop. Cancel subscription. And if the user does drum up attention for his idea? Then he's going to replace the quicklytest.it page with his own thing. Again, cancel subscription.

I like the idea, but it will be very tough to make money from it.

[1] Percentage of subscriptions canceled each month

I think thats a good point. Simply put: for this product to be beneficial to its users they shouldn't have to use it for a long time. The goal is to find out quickly whether or not people want to buy your test product. It's not something you want to pay for long term. Unless maybe it's an agency or someone that launches multiple products regularly. But I think the people that fit into this category are pretty limited.

It might not hurt to look into a higher one-time price point or I suppose you could base expected revenue with the assumption that most people may cancel after a month or 2.

Hi JD, Bryan,

I have thought this too, however the barrier of a higher price point was too much to stop people making that initial purchase.

I have done exactly what you have suggested and predicted revenue from a 2 month turnover.

This is no means meant to be 'startup' in the massive growth sense.

I like to build products that are useful for people. If I can earn some beer money out of it at the end of them month then great.

Nice. But I was a bit disappointed that when I clicked on the "get started now" link it didn't take me to a "coming soon, enter your email address to be first in the queue" form...

Disappointed that it didn't do that?

Yes, it would be beautifully ironic...

I seem to remember a service just like this on HackerNews about 6 months ago. I search hnsearch.com but couldn't find it, that said there seems to be a ton of advice on increasing landing page conversion (page 2, 3, etc..)

http://www.hnsearch.com/search#request/submissions&q=lan...

Launchrock?

Somewhat off topic but out of curiosity, for what genuine technical reasons would one show "There is a queue" message on the signup page? One would be very large no. of signups immediately following the launch. Anything else?

I don't subscribe to the idea of landing pages for non existent products but I feel that if lot of new products start showing such a message when there aren't many technical reasons to do so, no one is going to believe about the queue thing anyway. OTOH, people might start using it as a filter instead!

I personally did it for quicklytest.it as I wanted to great each of my new signups and ask them for feedback straight away.

Interesting. So it's a cheaper and lighter weight version of Unbounce. It definitely could be of use for super early stage startups.

Hi eskim,

It's not just startups that I'm targeting, but individuals who might have a couple of cool ideas that want to make that first step and see if they're as cool as what they thought.

So is your value add that you are actively steering traffic to the test sites (Google Consumer Surveys, Amazon Mechanical Turk, etc.) or simply creating basic landing pages and relying on the customer to generate his/her own traffic?

Currently the customer drive their own traffic (via the guides on the website). Mechanical turk is something I've thought about, but it's not really an accurate representation of a wide market, so I'm not sure the results would have much value.

Interesting concept. I use http://www.unbounce.com . Landing page platform with a/b test features and integration with lots of third party systems - mailchimp, kiss metrics, aweber, lots of templates etc

(1) How can simplistic "landing pages" validate a product idea?

(2) And if you actually do have a good idea, ins't using this more like "quicklyspill.it"?

Given what chrisacky observed, this is not even close to an "MVP" since it's not even close to being viable.

Hi Bill,

thank you for the feedback.

1) I believe that a landingpage can validate and 'idea' as you can use it to gage interest, via the visitor clicking through for information, time spent on the page, number of times the visitor returns etc.

2) I think it's fairly well documented that ideas on their own don't mean anything, they have little if no value. It's the execution that counts and if quicklytest.it is the first step in execution then great! It gets you out of the door.

or just hijack graphics from dribbble.com and use them on a private smore.com page and send it to customers. takes 15 minutes and is free.

Additionally - I've added a chat box to the website (olark) if anyone wants to catch up.

Reply on news.ycombinator.com