Digital Ocean’s journey from TechStars reject to cloud-hosting darling

http://techcrunch.com/2014/03/23/digital-oceans-journey-from-techstars-reject-to-cloud-hosting-darling/

by bjenik • 12 years ago

94 56 12 years ago

techcrunch.com

comice• 12 years ago

Article doesn't disclose that Crunchfund has a stake in DO.

politician• 12 years ago

"rejection" -- I had to reread that headline 3 times before I figured out what it was trying to say.

neotek• 12 years ago

"Reject" is fine, it means "a person or thing dismissed as inadequate or unacceptable."

larrys• 12 years ago

Totally misleading linkbait title.

Sure they were "rejected" by NYC (for valid reasons - apparently treated very well) but then went into the Boulder program after applying. And no doubt the reference given by Tisch must have certainly helped I'm guessing.

Nux4• 12 years ago

Every day, working with customers' products or witnessing stories such as DO's I realise that success is 90% marketing and 10% actual technical stuff.

You don't have to have the greatest features or too many of them or even all of them working as advertised. You don't need to take security very seriously (hello DO, WhatsApp!) or use buzz-stuff such as OpenStack (hello Rackspace, HPCloud!) or any existing "cloud standards" in order to succeed.

You just need to market it and promote it right.

I'm not even sarcastic or anything. As a techie I always frickin overlook it and this is the single most important element that can help you make serious money/business.

There's a whole bunch of people out there who don't know shit or don't really care about how great your product is technically (or not ;> ). All you need is it satisfies some basic "need" and a shiny packaging, it's enough to go out there and get some milk flowing in.

Kudos to the DO folk!

solve• 12 years ago

Alternatively: all the ways that it could fail don't matter if there's at least one way it could succeed.

giulianob• 12 years ago

I think at the end it came down to price. No one else was offering SSDs at the time for that cheap. Sure you get what you pay for but for a lot of people I guess it's good enough.

alien3d• 12 years ago

Yes.because it cheap and i want to test ssd. I want to test how include or require in php can be speed up.

jonesetc• 12 years ago

or use buzz-stuff such as OpenStack (hello Rackspace, HPCloud!)

Maybe I misread the tone of this post or I'm fuzzy on the history, but it sounds like you're saying Rackspace is trying to be hip by jumping on the hype-train. My understanding is that Rackspace co-created [1] Openstack and let it out into the wild.

[1] http://en.wikipedia.org/wiki/OpenStack#History

Nux4• 12 years ago

What I meant is Rackspace (also RedHat, HP, Mirantis etc) would have you believe that if you're not running Openstack then your "cloud" is shit and has no future. DO clearly demonstrated otherwise.

They're inflating the hype because they generate $$$ this way - from their public cloud, private deployments, training, certifications etc etc. Nothing wrong with that I guess, it's business.

mbesto• 12 years ago

Totally agree.

People who are going to argue against you are going to say "But what about products that are so great that they go viral, people share them and your customer acquisition goes through the roof!". This is marketing. Making frictionless experiences that satisfy needs very quickly is a great way to market your product. Note - Many of these experiences are frictionless because the security sucks. Security is accessibility's worst nightmare.

WhatsApp did this by making it easy for my parents living in the US to download the app, add my UK phone number and voila, I can IM them for free.

DO did this by making the interface on a VPS provider "not suck" and the pricing in line with all of the junk shared providers out there.

What I've noticed about "Bubble 2.0" is that making great, frictionless products is the new marketing. It is still marketing, and I feel it's best that people recognize it then pretend it simply doesn't exist.

straws• 12 years ago

If this kind of "marketing" is truly the thing you overlook when making things, are you making valuable tools that are hard or frustrating to use? What are the parts of your products that have nothing to do with the friction in using them?

stormbrew• 12 years ago

I think you're kind of wrong here. DO's biggest thing for me is that it's incredibly easy to use while providing all the baseline features I actually need for casual hosting. I guess you could count this as 'marketing', but I think there's a strong technical element to knowing your market and how to make what they want to do as easy as possible. DNSimple occupies a very similar niche.

In fact, part of the appeal may be that they don't use openstack, because openstack implementations to date have all been quite limited and complicated so far.

hyp0• 12 years ago

Yes. Two parts of business: engineering and marketing.

Two parts of marketing: promotion and meeting needs.

The key "marketing" of DO was:

  Existing providers offered pricing plans that were too complicated and platforms
  that weren’t always meeting the needs of their users.

It's not enough to make a "better" product, technically. You have to make a product that better meets user needs (relative to existing alternatives).

redwall_hp• 12 years ago

For me, it's all the pricing-reputability ratio. They've always seemed like a stand-up company, while offering a comparable product to Linode and similar...at a fraction of the price.

I jumped over to them around the time NewsBlur wrote that post about scaling to meet the new demand when Google Reader shut down. I had never heard of DO before that, checked them out, and made a the switch after a bit of research.

donniezazen• 12 years ago

That seems so now a days. Folks like easy things. Plain and simple.

jedrek• 12 years ago

I use DO as a private little always-on server, I don't actually host anything worthwhile with them.

I signed up mostly because of the $5/month price and the $10 credit. That's half of what I pay for my main hosting account (over at WebFaction) and year, is about what a nice meal out costs. It's nice, quick enough for all my needs and took minutes to set up.

Nux4• 12 years ago

Yeah, I guess the line between marketing and technical is not very clear all the time.

Re Openstack, indeed it has failed the "VPS" market which is still going strong.

tim333• 12 years ago

In my case DO succeeded more through the product than the marketing. I signed up after reading several 'who's the best hosting service' discussions and the consensus was DO was the best for small developer stuff, Linode and AWS being sometimes better for larger stuff. I've never seen an ad for DO. It's all been user recommendation.

pbreit2• 12 years ago

I think he way you are explaining it may give the wrong impression that the technical aspects are easy or unimportant which couldn't be further from the truth. In fact the technical aspects (and a compelling offering) are difficult and critical but merely a baseline to get in the game. Then, all being equal, the winners are those that figure out how to gain customers (marketing, for lack of a better term).

frakkingcylons• 12 years ago

I know that TechCrunch isn't interested in discussing any actually useful details of what made DO successful...but for me it came down to them doubling the RAM per instance and using SSDs in 01/2013, creating a collection of excellent sysadmin articles, and cheap/free backups and images.

rythie• 12 years ago

I used slicehost, though when it became rackspace cloud it became way more complicated. The main draw for me was the $5/month pricing combined with a $50 voucher they gave for black friday and they use SSDs. Even with all those benefits there are still switching costs due to having to re-setup on the new host, though I imagine most of their growth is people starting new projects.

computador2• 12 years ago

What the HTML is that?!

computador2• 12 years ago

...was their catchy phrase on their earlier video commercial on Youtube.

https://www.youtube.com/watch?v=vHZLCahai4Q

I'm sure this contributed in a way to their current success

coreymgilmore2• 12 years ago

DO really is a tremendous product. The simplicity, pricing, and speed of setting up an instance is awesome. For setting up a quick dev environment or a simple host for X or Y, it is definitely easier than through AWS.

I use DO pretty often and I hope they have some cool new offerings in the future. Maybe really simple load balancing?

ycombasks5• 12 years ago

I wish they could address these issues: http://www.redbottledesign.com/blog/five-reasons-why-you-sho...

I think "There's no way to increase storage space without major down-time" is the most important one. One hour of downtime is no good for production.

btgeekboy• 12 years ago

For what it's worth, #4 is a moot point: DO has said that they're not currently billing for bandwidth overages.[1]

[1] https://www.digitalocean.com/community/questions/where-can-i... (Kamal Nasser February 13, 2014)

rschmitty• 12 years ago

For #2 vote for this https://digitalocean.uservoice.com/forums/136585-digitalocea...

and

https://digitalocean.uservoice.com/forums/136585-digitalocea...

I don't think the time behind the scenes will ever be zero so I'd rather 100% be guaranteed not to lose my IP address. Spin up a new VM and assign prod-new the IP of prod-old similar to how you might do it with AWS Elastic IP

coreymgilmore2• 12 years ago

I do agree with this. Adding storage to an existing droplet would be nice. Even if it required adding a separate "disk", this would be useful. But obviously, adding, for example, 20GB disk after disk after disk could lead to a lot of annoyance.

mobiplayer4• 12 years ago

You don't care much about downtime if you're using only one droplet for your app in production, do you?

joshmn3• 12 years ago

DigitalOcean is not a cloud or anything that resembles what a "cloud" is.

If you consider "cloud" hosting backup-enabled virtual private servers with multiple operating-system choices and multiple locations, well, hate to break it to you, but that's has been around since 2005.

Everyone is hopping on DO because it's marketed as a "cloud" and it's part of a tech accelerator (and maybe because it has SSDs and is so damn cheap). Other than that, there's really nothing special about it.

Source: me; hosting-industry expert.

ohashi2• 12 years ago

Everyone is hopping on DO because it's marketed as a "cloud" and it's part of a tech accelerator (and maybe because it has SSDs and is so damn cheap). Other than that, there's really nothing special about it.

The marketing is good. The quality of the product is good. It's incredibly easy to get started with. The tech accelerator I don't think plays any factor. I searched my database and there are 24 mentions of TechStars out of 3213 positive mentions. That's 0.7%, I can't imagine it has much impact right now on their growth. To give some other comps cheap is 115, affordable is 36, ssd is 133, fast is 174, amazing 212, easy is 134.

Source: http://reviewsignal.com/webhosting/company/101/digitalocean

joshmn3• 12 years ago

The marketing is false though.

That would mean I can go ahead and market a Galaxy S6 as an Apple device and watch all the fanboys and teenage girls hop on. But of course I need pretty colors.

zaidf• 12 years ago

The biggest reason DO took off has little to do with the reasons you cite. It's got to do with simplicity and hourly billing. AWS had the latter without former. Your typical VPS host has a bit of both but wont tell you launch 20 instances for a few hours for pennies.

joshmn3• 12 years ago

Your typical VPS host has a bit of both but wont tell you to launch 20 instances for a few hours for pennies.

DigitalOcean limits you to 5 off the bat, but that's just being picky.

RackSpace, Atlantic, Gigenet, VPS.net, Oxeo, HPCloud, Upcloud, Dotblock, StormOnDemand (LiquidWeb) to name a few. And that's off the top of my head.

philtar• 12 years ago

I think the SSDs and the price are why people are jumping on. Not the marketing. Their marketing is horrid.

"You've been developing like a BEAST and your app is ready to go.."

priyankt• 12 years ago

My opinion. I am a DO user and I don't think you can use it for production application. Casual hosting is fine.

For production servers, deterministic backups are very important. DO does not provide that. I faced a lot of issues when transitioning my server from non-backup to backup mode. To do that, you need to take a snapshot(server shutdown is required for this)of server. Then, you need to destroy it to get the same ip. There is a chance that your ip may go to someone else & any services that are dependent on ip may need changes. I was lucky to loose the ip for a while & get it back after some trial & error.

Also, you don't have any control on the backup. It is taken by the system and is supposed to be anywhere between 2-3 days.

xur17• 12 years ago

Are those the only backups you take? I have all of my code stored in git, and my database (and assorted files) get backed up each night to aws. I figure a snapshot is my first line of defense, and aws is my backup in case this doesn't work.

jaequery• 12 years ago

i dont know why but i'm so tired of these posts

soci• 12 years ago

I wonder at what scale a business like this is sustainable with this low pricing schemas. Whatsoever, will it ever be?

teekert• 12 years ago

I really loved DO, that is until they emailed me that my server had to be rebooted and I found it with all the config files being replaced by a bunch of @@@'s (I had to use the browser VNC to log in).

I asked them what had happened and if it can be fixed, they tell me to make a new droplet. I ask what will prevent this from happening again... no response...

Bye bye DO.

Digital Ocean’s journey from TechStars reject to cloud-hosting darling

Related Stories