Ben Thompson, one of the best tech writers, just put out an article on Mongo vs Amazon.
He says:
Basically, MongoDB sells three things on top of its open source database server:
Additional tools for enterprise companies to implement MongoDB
A hosted service for smaller companies to use MongoDB
Legal certainty
I believe Atlas is the hosting service, and if it is only for smaller companies, that leaves the door open for Amazon.
He thinks the Amazon risk is formidable.
Jim
I believe Atlas is the hosting service, and if it is only for smaller companies, that leaves the door open for Amazon.
Maybe some more tech-savvy folks can help, but is this true? If so, why wouldn’t Atlas appeal to larger companies? Why would they be more likely to want their data hosted with Amazon than MDB?
Bear
bear, just to correct one statement.
"Why would they be more likely to want their data hosted with Amazon than MDB?
nobody is hosting their data with MDB. Atlas is a service that is on top of either Azure, Google Cloud, or AWS. Think of it as a Amazon’s vendor lock in version of atlas. I agree that many people don’t want to be locked in, but there are already tons of people that are locked into AWS for other services so why not add a Database. Amazon has directly targeted why we invest in MDB…atlas.
-e
I mean, scroll down a page or so here: https://www.mongodb.com/cloud/atlas
…and see that Atlas has some pretty large customers including eHarmony, Sega, KPMG, 7-eleven…
Bear
I wasn’t very convinced by this article. While he did give some insights on the way open source works and made an interesting (also seemed a bit forced) comparison with the music industry, he didn‘t present any good arguments why Amazon will outcompete MongoDB going forward. (Especially, if you compare it to the deep discussion on this board and NPI) Basically, his argument is that Amazon will sell “performance, scalability, and availability”, which is much more about selling convenience than being a product, and therefore, even though MongoDB invested much in building the product, Amazon will eventually take the spoils… From what I read so far, this explanation doesn’t even seem to scratch the surface of the competitive situation in NoSQL databases.
Niki
his argument is that Amazon will sell “performance, scalability, and availability”,
Which also assumes that Amazon developers are better at providing this on a brand new database than Mongo developers are at providing this on their more mature database. Could be, of course, but so far all we have are unmeasured claims of improvement relative to a two year old product.
“Which also assumes that Amazon developers are better at providing this on a brand new database than Mongo developers are at providing this on their more mature database. Could be, of course, but so far all we have are unmeasured claims of improvement relative to a two year old product.”
It isn’t a new database. It’s very likely Aurora, which is 5 years old as a public AWS service and mature.
https://aws.amazon.com/documentdb/
“DocumentDB uses a distributed, fault-tolerant, self-healing storage system that auto-scales up to 64 TB per database cluster. In Amazon DocumentDB, the storage and compute are decoupled, allowing each to scale independently, and developers can increase the read capacity to millions of requests per second by adding up to 15 low latency read replicas in minutes, regardless of the size of your data. Amazon DocumentDB is designed for 99.99% availability and replicates six copies of your data across three AWS Availability Zones (AZs).”
https://aws.amazon.com/rds/aurora/
“Amazon Aurora features a distributed, fault-tolerant, self-healing storage system that auto-scales up to 64TB per database instance. It delivers high performance and availability with up to 15 low-latency read replicas, point-in-time recovery, continuous backup to Amazon S3, and replication across three Availability Zones (AZs).”
The implementation probably something close to a stateful shim. You map the open source definition of the MongoDB api function to an implementation in Aurora. It’s honestly not that hard to do on a basic level. Get’s harder the closer you are trying duplicate the exact MongoDB 3.6 server behavior at that API level.
In the end what it comes down to is two things:
(1) as we saw with Azure’s attempt to usurp MongoDB (and back then, with 3.4 APIs, Mongo was much less feature rich and easier to attempt to mimic - and amazing that even back then Microsoft decided they had to add the Mongoeeish stuff to try to be successful) is that despite the marketing claims by Microsoft, the database lacked sufficient stability, had potential data loss issues, and was not Mongo even with the ability to use the then current MongoDB 3.4 APIs.
So is this new Amazon database really any better like Microsoft proclaimed so loudly (and inaccurately)?
(2) DocumentDB, unlike the then DocumentDB, turned CosmosDB from Microsoft is now working two year old APIs (remember, Microsoft could not make it work with then current APIs at the time).
Thus, as a decision maker, do you decide to put your mission critical software on AWS with Amazon knowing the it will never have the newest features, and that it will diverge more and more and more from the standard Mongo - which is now the standard (look at the database ratings, you have Oracle, you have MSFT, you have Postgres, you have Mongo - Cassandra materially is dropping as an example). Do you want to build your company on that?
Well, only if there are really some big pain points. If the pain points were that big in the first place would not Cassandra or Couchbase or even Cosmos or Aurora be stealing Mongo’s thunder to begin with?
And even if there are real pain points, and for some god forsaken unknown reason no other piece database, until now, DocumentDB from Amazon can actually fix those pain points, don’t you think that Mongo is (a) well aware of those pain points, and (b) has been working to ameliorate them, and will continue to do on an on-going basis?
Thus follow the money. I hear many anecdotal tales from people claiming to be DBAs (not saying anyone on this board is not a DBA, but on comments to stock articles and to marketing pieces claiming to be such) talking about what a {female dog} it is to work with Mongo.
First it sounds like a Democrat talking about a Republican, or vice versa, simply personal animus against the particular product (I’m Windows, Apple sucks, I’m Linux Windows and Apple suck…), second if they were any good at their jobs it would be just working with another database, the leading database. For the leading database to be a {female dog} to work with means you really suck at your job because the majority of DBAs in the industry who actually work with NoSQLs have chosen to work with Mongo - {female dog} to work with or not.
Follow the numbers, follow the money. Oracle is hated, hated, hated, hated, hated, who cares, follow the money, they are #1 - end of story.
Then we have heard stories that SQLs can do just about anything, just depends on the skill of the DBA. Except then we see real world mainstream examples where despite years of effort with the best teams on the planet things that you would think would be just common for a database to be able to pull off the SQL could not do it. Thus their is professional bias going on. You know the stories old timers, with the rule of thumbs, tell to the next generation who think they are too smart for their own good.
Back in the early 1980s the same people were saying that the only thing personal PCs are good for is balancing your checkbook. As things stand now they are not even good for that anymore as the bank balances it for you online.
(3) Follow the money again. The stock is still above its December lows, and this is a stock that hit new highs in December as well as we were in a historical crash. Seems to indicate existential threat…not.
In fact Mongo has not even fallen far enough for me to call it a “buying opportunity”. I called the $60s as the last buying opportunity and the stock has not even fallen that far on the news.
But there are many stocks to invest in. In and out and in and out. On NPI one investor (and it is a fair strategy, good as mine, good as Sauls, just a heck of a lot less effective over the entire last decade) said he was going almost all cash and would buy after the market bottomed and started going up again.
Well, sounds reasonable and prudent. Except that was 35% ago for Zscaler. What is it, 40 some percent for Twilio? 25% for Nutanix.
Every time, I mean every time we hear panic on the boards, overthinking, overreaching, well each to their own.
We just had that big crash everyone was warning us about, every six months or so, as if it were a morality play and the “party is over” on our stock “market darlings”. Omitting of course that this is not the year 2000, anymore than we are no longer in the year 1914 either in foreign policy, or 1929 in economics, and thus anyone who is making money in the market must be doing it solely because they are in a bubble of momentum. Nevermind the BUSINESS FUNDAMENTALS.
In the end, it is the business fundamentals that are driving the share prices, thus why they bounce back. If you want a bubble, look perhaps towards cryptocurrency or marijuana stocks (both commodities until someone can brand them like Winston or Camel did with tobacco).
Oh heck, I’m just venting. :).
Tinker
Tinker,
It really boils down to the fact that regardless of the decision, the MongoDB investing thesis is based on Atlas and Atlas is exactly the market that AMZN is targeting.
Guidance will drive price now, revenues won’t reflect any significant changes for a few quarters.
It really boils down to the fact that regardless of the decision, the MongoDB investing thesis is based on Atlas and Atlas is exactly the market that AMZN is targeting.
“Una cosa piensa el burro y otra el que lo arrea.” Spanish saying, the donkey wants something and the herder something else. Just because Amazon is going after Mongo does not mean that Mongo is down and out. DocumentDB is not disruptive technology, it’s more of the same. Previous attacks on Mongo have failed. The decision is not made by Amazon but by Mr. Market. Try thinking like Mr. Market.
Denny Schlesinger
A couple of words of caution. First, if you are going to cite history, go back and look for more than just one example. And also, make sure to not call the game too early.
Microsoft had a stroke of brilliance with Windows. Just about every single thing since them has been them trying to catch up and overtake some visionary company, usually by throwing gobs of money at the problem (or giving product away for free!). While they might be hated for it, they are very well practiced at this game, and they know it’s a long haul type of play. You say CosmosDB failed to dislodge Mongo? I say way too early to tell. Remember WordPerfect? Remember how much better it was than stinking Microsoft Word? Well, where is it now? That’s right. Sure, it took years. But it’s the final result that matters. Remember the first Internet Explorer? Netscape was so much better! We know what happened there. SQL Server? Re-branded Sybase. Xbox? Built on Playstation envy. Azure? Amazon did first. There are literally hundreds of examples.
Microsoft missed the internet. They missed the cell phone. The relational database. The game console. And yet, today, they are successful players in most of these spaces, taking significant market share from the original visionaries. Why? Not because they have the best, innovative engineers. But because they have a cash cow in their belly, and very little in the way of pride. They will throw money at a thing until it’s theirs. People will hate them for it, and they will fight it, and it will take years, but in the end, Microsoft will have its way. Because money talks, and you can keep undercutting the better, smarter, more innovative and more honerable competition on price until they weaken.
Now what’s the difference between Microsoft and Amazon in this scenario? Only one thing. Amazon actually has far better engineers. They are actually the ones who made cloud real, after all.
I say way too early to tell…Sure, it took years. But it’s the final result that matters.
Well, that is unless you’re buying Mongo stock with the intent of following the company’s results and acting, if necessary, if their progress or competitive thesis changes. In that case, you can ride this current market leader for potentially a long time, make a lot of money, and exit with profits intact, even if Amazon eventually ends up being the winner (a big “if”). That’s my plan, at least, doesn’t always work out, but you can’t fault me for trying… 
Just saying, as agile investors, we’re not making an unchangeable bet until Mongo either takes the entire NoSQL market or goes under, we can exit anytime it looks prudent to do so, hopefully with a profit under out belts. So in that case, the “final result” doesn’t really matter.
Some see this as such a time (to exit), I do not, and continue to hold, and even add to my position as the market allows.
Well, that is unless you’re buying Mongo stock with the intent of following the company’s results and acting, if necessary, if their progress or competitive thesis changes. In that case, you can ride this current market leader for potentially a long time, make a lot of money, and exit with profits intact, even if Amazon eventually ends up being the winner (a big “if”). That’s my plan, at least, doesn’t always work out, but you can’t fault me for trying…
That is certainly a fair point. The point of intersection between these two things is valuation - if current prices are based on the assumption of long term market dominance, there may still be a near-term hit as that theory is re-evaluated, even though the actual damage won’t be directly observable for a couple of years or more.
