Here is how I see this, as an IT manager and an ex-software developer:
a) Amazon has its own document database (perhaps partially based on Mongo’s older code, perhaps not), which it believes scales as well or better than MongoDB.
b) It wants customers for this new offering. However, they recognize that asking people already using the world’s most popular document DB (Mongo) to cold-switch Amazon’s offering may be a tough proposition.
c) Therefore they make their documentDB compatable with MongoDB at the API level, to make it easy for current MongoDB users to switch.
d) They have to go with a slightly older API specification because of recent licensing changes Mongo made, but they are betting that this isn’t a big deal for most people, and they are probably right.
e) Going forward, the two offerings will continue to diverge, as Mongo will continue to develop its software, and Amazon will develop documentDB in their own direction.
f) It’s anybody’s bet who will end up with a better product in the end, but Amazon has everything to win and Mongo everything to lose here, because:
Mongo has the customers to lose here, and Amazon to gain.
Amazon will almost certainly underprice Mongo’s Atlas offering to make it even more enticing for customers to switch.
Amazon’s engineers are pretty good, and it’s a solid bet they will have a damn good document database offering in the long run, and it will certainly play very nicely with AWS as a host platform. This will be attractive at least to some customers, because all things being equal, from an operations perspective, life is easier when all the things you depend on run in AWS (simpler for billing, vendor management, resiliency planning, security management, training, etc.)
Best bet scenario for Mongo is that they don’t lose too many customers, but have to lower Atlas costs to stay competitive with Amazon, and keep their pace of innovation such that they still stay attractive to future customers vs. documentDB. But again, that just amounts to minimizing damage. There is no upside here for them.
I do think the market overreacted price movement wise. I will probably wait for a re-bound, reduce my position by half, and watch carefully for signs of Mongo losing revenue and shrinking margins.
Having worked in companies that usually depend on the Channel for the majority of sales, I do find what Amazon does strange. They do make money from any MongoDB customer that is running on top of AWS. Perhaps they make a bit more if they take the customer directly as opposed to the customer paying MongoDB but they also don’t have to incur the customer aquisition cost nor the support costs to the customer directly.
We sold through the channel, we often debated to take certain customers direct but there were a lot of hidden costs involved that would usually offset the increase margin on the deal. But the big risk was losing the partner that may decide to just migrate their customers to another companies hardware in the next cycle.
I would expect Amazon to offer a lot of Low level applications but to be careful not to compete or actively promote to move end users from one of their Partners to them directly which seems specifically what they have done. There would be a lot of SAAS companies that have to question themselves if it would be good for them to offer their solution on AWS if they believe Amazon is just going to turn around with a similar solution and tell the end users how easy it is to move. Maybe they just feel that it is impossible for any provider not to offer their application on AWS and don’t care.’
Maybe there is a business model for a cloud provider to applications providers that are completely agnostic and would never offer their own solution, just provide the infrastructure.
There would be a lot of SAAS companies that have to question themselves if it would be good for them to offer their solution on AWS if they believe Amazon is just going to turn around with a similar solution and tell the end users how easy it is to move. Maybe they just feel that it is impossible for any provider not to offer their application on AWS and don’t care.’
Maybe there is a business model for a cloud provider to applications providers that are completely agnostic and would never offer their own solution, just provide the infrastructure.
JDC,
This is perfectly put. I completely agree. The more I think about this, the more it irks me. And not just as a Mongo shareholder. It seems like Amazon is flying really close to the sun in a lot of ways. I don’t foresee its partners boycotting or anything – it’s a mutually beneficial relationship while it lasts, but it’s gotta suck doing business while always looking over your shoulder.
This is perfectly put. I completely agree. The more I think about this, the more it irks me. And not just as a Mongo shareholder. It seems like Amazon is flying really close to the sun in a lot of ways. I don’t foresee its partners boycotting or anything – it’s a mutually beneficial relationship while it lasts, but it’s gotta suck doing business while always looking over your shoulder.
Q. “Where does an 800 pound gorilla sleep?”
A. “Anywhere it wants!”
The only way to stop it is to bring in a 1600 pound trust breaker.
Not to worry, Bezos can’t have everything… like two wives.
I think this is a reaffirmation of the power of the Mongo database solution, and only furthers my conviction. I’m increasing my allocation to this name!
I would like anyone to tell me where an 800 pound gorilla ever disrupted a start up? I can’t think of one but I am sure there has to be a few. Everyone is worried about Amazon but the problem is that they are not disrupting MDB, it’s MDB that is disrupting the database area. Yes Amazon has disrupted a number of business’s but they were always starting from the bottom and moving up. I would rather put my money into the company disrupting the market, then the company trying to defend their market.
A fair point, but on the other hand. If an 800 pound gorilla disrupts a startup, it’s likely that the startup gets forgotten. I just wouldn’t trust my memory on this line of thinking. I think you’re right though that the velocity that a startup has is its key advantage.
I would like anyone to tell me where an 800 pound gorilla ever disrupted a start up? I can’t think of one but I am sure there has to be a few.
Netscape!
I dunno. The amount of discussion around this whole topic is making me feel like this is under Saul’s “Too hard to figure out”.
While AMZN is my largest holding, as I’ve been a rather enthused MDB shareholder, I do find this whole situation by AWS/DocumentDB pretty frustrating.
I get that MDB latest code has the ACID competitive edge and perhaps other features too ---- but DocumentDB 3.x will eventually have it. NoSQL momentum/gravity will eventually shift to DocumentDB. It’ll just be capitalist markets or physics at play.
Netscape is a great one. Microsoft did everything they could to take them out. Now does it seem that Amazon is going to put the same effort in to take out MDB? Seems like a side business not a major business to Amazon.
On the surface and in the near term it seems Mongo has every thing to loose. But on the longer term it may strangely actually help Mongo.
Let me explain, take the example of Oracle. Oracle’s revenue stream comes not just from selling it’s core database products but apps built on top of it. These apps are built by both Oracle and others.
In Mongo’s case example current example of apps would be Compass and Charts. I am sure there is a very long product map here possible. Over time I see that as the optionality in Mongo’s business
Amazon understands it is very hard to disrupt an entrenched database. A part of the problem is the ecosystem of tools that work with it. Typically the apps access the core products through the API.So they specifically point out that all tools that work with Mongo 3.6 with their DocumentDB. So I expect even Mongo’s apps should work with DocumentDB !
The giant portion of database market is still SQL with noSQL rapidly growing in significance. Amazon’s move has validated that Mongo is the clear leader in noSQL. Besides it may help spur further adoption of noSQL paradigm. That Mongo will lose some customers is a given. The market is big enough for multiple winners. We just need Mongo 200 million revenue to 1 to 2 billion. Even with Amazon’s move I believe Mongo will be bigger than 2 billion down the road - databases are simply everywhere. The market is just too big.
So in summary:
Amazon’s move may help accelerate shift to noSQL benefiting Mongo.
Even if Amazon Document DB is successful, Mongo can focus on monetizing it’s apps both on their core product and Amazon’s core product. Just as Microsoft has found it worthwhile to port office to other iOS and elsewhere.
I’m totally NOT a tech guy. I love the TAM that MDB was potentially addressing. I do see a competitive threat from AMZN’s AWS offering from all that has been written on this subject. So the question I have was how much does this news affect MDB’s TAM and what are the chances other cloud players will do the same thing. Is the on prem DB market still large enough for significant growth major cloud players aside? With so many fremium downloads, I would think that the on site market is quite large in of itself? If other major cloud players adopt a similar tactic as AWS has done; that would become more problematic to me. AMZN’s AWS is the major player but not the only large public cloud player. A list of the major cloud players by revenue:
I guess we will see how the numbers stack up. That will be the proof of the pudding to me. As long as MDB can continue to put up the numbers they have been doing the stock should do well. Meanwhile I will be alert to further developments of the other cloud players. Virtually no company is without competition and the bigger the TAM, the more likely some of the big guys will get in. They don’t always win. It has always been so.
Amazon’s move may help accelerate shift to noSQL benefiting Mongo.
Even if Amazon Document DB is successful, Mongo can focus on monetizing it’s apps both on their core product and Amazon’s core product. Just as Microsoft has found it worthwhile to port office to other iOS and elsewhere."
I think you make a very important point. Amazon has validated the NoSQL market and Mongo’s leadership. I remember years ago interviewing with a technology company that had been in the market for 7-8 years and the VP boasting to me how that had absolutely no competition. No one was doing what they do. I took a hard pass because if there is an addressable market with huge profit potential, at the least the big boys will try to play in it. This makes Mongo a great acquisition candidate and the future will be all about how far ahead they can push their innovation lead to address real business use cases.
The most likely outcome here is that the CTO Eliot Horowitz (Mongo is his baby) and his team will get pissed, and as competition will do, start pushing new product iterations on steroids.
Serverless (check), in database analytics - pipeline aggregation (check), ACID - multiple documents (check), mobile (check), VPN bypassing public internet (check), and so on…Kubernetes (goal for 2019).
All of these things are version 4.0., 4.2 was already planned 2 years ago and is forthcoming.
This only does not work if the giants emulation product is already “good enough”. I am sure it is for start ups w little budgets and small guy hackers, students, and hobbyists. There may also be legacy mission critical loads that large enterprises may find Document DB to be more than adequate for. However, anyone forward looking…not.
I would like anyone to tell me where an 800 pound gorilla ever disrupted a start up? I can’t think of one but I am sure there has to be a few.
Netscape!
I tho’t it was just me, but since you brought it up:
Twenty years ago, Massah Gates had proclaimed the www ‘thing and crowd’ was just a fleeting fancy… we’re not going to pursue…
Then, LESS than 12 months later, WE’RE IN. Formed group. And we’re going to charge for our product. Then, make it a part of the OS. Then, okay, no charge, and not boiled into Windows/OS.
(Mssr Gates was never the prescient oracle that many want to ascribe to him. He just wasn’t.)
I was heavy-holding Netscape at the time; made money okay, but total bucks were diminished in the train wreck.
