Ask HN: This is my MVP - would you pay $10/mo for my monitoring service?

if i run out of notification credits, do i not find out when my web site goes down?

i hope this feedback helps.

Have you spoken to customers about this already? Who is saying "I already pay for monitoring but crimety I keep getting false negatives because while it checks status codes it doesn't let me fire regular expressions against the content."

Scrap the credit system. It adds mental overhead for customers. People who pay for monitoring systems will generally tend to run fairly reliable setups. If you get one idiot on the service who needs an SMS every hour, you can either educate him, rate-limit him, fire him as a customer, or just subsidize it from your customers who know what they are doing.

Charge more. If I care about uptime, I care a lot more than $10. You'll get pushback from poor college students. So what. They are terrible customers to have and you won't keep them, anyhow, because somebody can afford to offer this cheaper than you can.

Your primary value differentiator (header/response codes against a regex) are more overhead that I have to create when I sign up. For me, just knowing if the system pings and responds on port 80 and 443 is sufficient.

Also especially agree with patio11 about the credit system. I'd only bring that in if certain customers are costing you too much to SMS.

$10/mo isn't a bad deal, but I really think you need to go way above what you're thinking here to dominate this.

Offer to keep a backup/cache of the site at a known point (this is basically a 1-line wget script). Offer some haproxy-bsed auto-failover auto-maintenance page stuff for when a site does eventually go down or get fubar'd.

From my personal perspective, it's good to know when something goes wrong, but it's fantastic to also be able to have some auto-fail or recovery option available (automatic, or user-initiated from an iPhone app) so that if I can't get online to rectify the situation promptly, I at least know users are being hit with 404's and SQL errors, etc.

My initial thought (and maybe my MVP proposal does not reflect this) was to bring some of the features that things like Gomez offers (detail timing, screen caps, transactional walkthroughs, etc) down into the pricing space of the lower end offerings. I've worked for a large internet retailer before where we used Gomez - talking tens of thousands of dollars a year for just a few tests. Ridiculous. But having used it, at least on our site, I can say that it was definitely the fine grained monitoring that saved our asses -- things would NEVER outright 50x, just fail in really subtle/glitchy ways. Maybe it was because our system was for more complicated than Johnny PHP/Rails. IMHO 50x means the network is down, the IT guy unplugged something, you're getting DDOS'd or your system is just inherently unstable. It never meant "Apache died" - because Apache never did die, for years and years. Maybe I'm wrong to try to take "head" features and push them down to "tail" pricing.

Perhaps I could try to create a higher level offering that's still more affordable than Gomez. I don't know what that pricepoint would be though. Watchmouse has a transactional/clickthrough offering for $500/month. They look hard to compete with - their offering is quite robust. As for other offerings, I don't know. They don't tend to publicize a lot of their pricing, seem to have a not-so-self-service sales cycle.

I don't see the point of trying to create a "cheaper than cheap" option and competing with the simple-ping tail.

I guess I just though I saw an opportunity where some people may be getting greedy with their margins, and thought I could fill in that hole with an offering that was equivalent where it mattered, but at a much lower price point. Low enough where people in the tail would maybe even step up into it. Perhaps the world is perfectly happy in a bimodal setup - the "<$5/mo ping and L7 check against one URL" type of monitoring and the ">$x,000 or $x0,000 year total solution" type of monitoring.

I'm a developer, so I know what ping, DNS, TCP, and UDP mean, but for the average Joe with a Wordpress blog, maybe a more compelling sell is:

Check the following: Is my site up? Are my ads showing? Has my site been hacked? Is my site slow? Is Google Analytics showing? Is my email up?

Marketing to this audience might run counter to the "charge more and eliminate needy customers" market, but you might find other ways to generate revenue. For example, you can run analytics on different hosting companies and see which ones are consistently slower than average, and market the fastest ones to your clients (at $100 a piece affiliate revenue).

1. Time the download of resources on each page, similar to what you get from HTTPWatch. Report on which ones are blocking and which ones are giving 404s

2. Provide a way to script the call, with user name and password, and the ability to store cookies and follow the redirect URL, up to N(5?) levels. This makes sure that not only the home page is up but a (test) transaction can also complete as expected.

3. This is more optional, but provide a sample page that users can host on their end or an API they can leverage to restart app or take similar actions, so that if the monitoring finds a problem, the answer is not calling the admin's phone.