Google’s Cloud NoSQL Database

More competition for MDB?

https://techcrunch.com/2019/01/31/googles-cloud-firestore-no…

TJ

2 Likes

It certainly doesn’t mean in any way that it’s all over for Mongo, but that was part of what I was worried about when I got out of Mongo this month.
Saul

2 Likes

On the other hand,

Mongo must really be on to something here. They are the ones that drove the market here. The disruptor, the Top Dog, the crusher, the winner, yadayada.

These Titans must sense a coming storm to all be trying to get in.

I like this for Mongo.

Darth

24 Likes

Here is a Gartner peer insight report on Mongo vs google. Note: you can add different vendors to compare which ever companies you want.

https://www.gartner.com/reviews/market/operational-dbms/comp…

If we assume the survey’s submitted from customers is representative (and proportional) of all customers, it can give some good insight into competitive advantage. (This could be a big assumption, but I think it is reasonable)

Mongo has about 8000 customers, with 537 surveys’s it’s about 8% of Mongo’s customers.

A couple of observations

Mongo has 7 times more reviews, so about 7 times more customers
Mongo also has about 7 times more large customers (above $1 Billion)

both have about equal customer rec. rates of 80% and 84%

when looking at the other customers considered section, you can see who wins when both are considered.

Mongo wins 50 times out of 67, or about 75% of the time when both are considered.

If you compare Mongo to Amazon, Mongo is a clear winner, especially with large customers.

Maybe this new Google product improves their competitiveness, time will tell, but right now Mongo is crushing them.

Side note:

I started looking at this site recently with AYX. I thought AYX had much less competition than MDB, and was a clear leader, but now I don’t think that is the case.

Here is AYX vs Rapidminer and Knime.

https://www.gartner.com/reviews/market/data-science-machine-…

Jim

18 Likes

I’m holding steady with Mongo.

These developments could certainly impact MDB, but I’m on the side that all of these developments are reinforcing the importance of NoSQL and will actually increase demand faster than what was previously expected.

As long as MongoDB keeps innovating and offering the best of breed service for their niche in this MASSIVE potential market (that is an assumption of mine) then they will be just fine.

I’m not even going to list all of the competing products that Amazon and Google (and other giants) have launched that have not taken over the world, or sent other, smaller, more intensely focused companies crashing, but here’s a few companies excelling in the face of competition from the giants.

TTD - advertising
Paypal & Square - banks/fin tech
Apple (original iPhone release) - every other mobile phone in the world
TWLO - many other contact center platforms
ANET - Cisco
AYX - every other data vis tool
MDB - Oracle, MSFT, and now Google/Amzn (Hello demand)
TSLA - other autos
Other grocery stores - AMZN buying Whole Foods (I think Kroger is up big since when this was announced, though I’m not sure…but the grocery shopping experience has improved)

the list goes on and on. The fact of the matter is that these giants do kill some companies… probably many smaller, private companies we never hear of and old, stagnant retail giants (Sears), but it has been prooven time and time again, that small, innovative, intensely focused companies, with technically superior products, that are usually founder-led generally stick around and stay leaders in their respective markets.

My bet right now is that NoSQL will be large enough for many winners, and MongoDB will continue to be a winner for a long time in that market.

If the numbers begin to show otherwise, or we see signs that others have caught up from a technical perspective, I’ll move that money elsewhere.

18 Likes

The question that comes up and is never answers is, why?

Why do you want to put your mission critical software on an inferior database, that costs as much or close enough to as much as to make no real difference, that is stuck on one cloud and for the most part even development has to take place on that cloud and not local?

There are some who just personally hate Mongo (software developers get that way); there are some who just don’t know the difference; and then there are some use cases where it may be beneficial to use the Google database.

Google is considered to be #1 in AI. Thus, if you are running AI in the Google cloud, then perhaps it may behoove you to run the Google database in their cloud. Perhaps not, but because you may not be an expert in databases you just go with it anyways.

Other than this, why would you want to use an inferior database that locks you in to one cloud and does not save you any money?

Finally, who cares. The NoSQL database market continues to gain marketshare in a market so large it is like a few drops of water in the ocean at present, and growing so rapidly it will probably be very difficult to show any signs of any one vendor taking any business from Mongo other than looking through the database ranking lists and see if google suddenly starts to move up the rankings like Mongo once did, like Postgres once did, like Elastic just did.

Otherwise, good to keep track of what is going on but it seems to have no more relevance than AWS or Azure doing the same.

Tinker

16 Likes

Also, the market seems not to care, as Mongo is up almost 5% today and near or at a new high…

Matt

2 Likes

They are the ones that drove the market here. The disruptor, the Top Dog, the crusher, the winner, yadayada.

These Titans must sense a coming storm to all be trying to get in.

That’s not a very good way of looking at it. I think you can make a very strong argument that MongoDB is neither “top dog” nor particularly a “crusher”.

The facts are:

  • There have been NoSQL options for decades (long before we had the term NoSQL). MongoDB was not the first document database, nor the first NoSQL database.
  • The NoSQL space is exploding. While NoSQL has been around a long time, until recently relational was the “safe choice”, especially for enterprises. Now NoSQL options have gone mainstream and you can now find the support/people/community/expertise/features you need to use these technologies a lot more effectively. This is a rising tide and it is floating many boats.
  • MongoDB is just one part of this NoSQL space. Yes, MongoDB is growing rapidly, but so is Cassandra. So is Cosmos. So is CouchDB.

My opinions are:

  • MongoDB is far from dominant in terms of usage, even in the NoSQL space. I definitely see them, but in my corner of the world Cassandra is much more popular. Yes Mongo is stealing a lot of marketshare from Oracle and other SQL databases, but so are all the other NoSQL players. I think you could also make the case that Amazon’s Dynamo is the big dog, as many AWS users just adopt Dynamo because it’s there and easy to use.
  • MongoDB is far from dominant technically. Mongo has had a lot of scalability and security problems. I guess this is normal growing pains, but the technology is not a “crusher”. I personally like MongoDB but in my part of the world MongoDB is not seen as a particularly good product and is perceived as technically inferior to much of its competition.

Why I think MongoDB is so hot right now:

  • Far more than other NoSQL vendors, Mongo has convinced companies to pay. Lots of people still use the community (i.e. free) version of Mongo, of course, but I feel like MongoDB has done a much better job of monetizing it’s product than much of its competition. It’s really walked that fine line of “free enough that it gets wide adoption” and “closed enough that enterprises are willing to pay money for it” well.
  • Far more than other NoSQL vendors, Mongo has executed well on the enterprise sales side. DataStax, in particular, has fumbled several times. Atlas is another good example of this. I’m not sure what % of their eventual business this will be, but it gives them an easy way to let customers get started and has been a good way for them to get businesses to pay money for an open source product.
  • MongoDB is a public company and I think that both gives them more visibility and credibility in enterprises. They’ve used capital well.

In short, Mongo has been successful because they have executed well. Perceiving them as the main disruptor isn’t a good view because others are more disruptive. Perceiving them as the crusher/winner isn’t a good view because they probably aren’t the most popular or best. Mongo is strong because they have been managed well, and because they have been effective at monetizing a market that others have struggled to monetize.

–CH

24 Likes

MongoDB is just one part of this NoSQL space. Yes, MongoDB is growing rapidly, but so is Cassandra. So is Cosmos. So is CouchDB.

This isn’t particularly helpful without quantification. Here’s the best I’ve seen:

https://db-engines.com/en/ranking

(Mongo #5 and rising, Cassandra #11 and falling, Couch #30 and falling, Cosmos #27 and rising.)

Or you could just look at document stores, where MDB is the overwhelming favorite.

https://db-engines.com/en/ranking/document+store

Bear

18 Likes

CH,

I appreciate an alternative view but I disagree completely. We have been discussing Mongo for awhile now. If you look at rankings, if you look at downloads, if you look at customer growth in other words usage of Mongo. Mongo most certainly is a Top Dog in their category. They most certainly are crushing that category. And they were certainly the First Mover in taking NoSQL and particularly Document NoSQL mainstream. There may have been others that were earlier on a calendar but that is irrelevant to what Mongo has accomplished.

As an investor, I can’t analyze what’s “in your corner of the world”. I’m trying to see the whole world. And for NoSQL and specifically Document, Mongo is crushing it. There is no argument to that. They themselves could be disrupted, as anybody can. There are alternatives for sure. And when ones corner of the world starts to go down a path they tend to stay on that path. And that’s what you’re comfortable with and you know so an organization continues to use it. Switching is always a problem for any software company.

Here’s Gartner’s list of Operational Database Management Platforms.

https://www.gartner.com/reviews/market/operational-dbms/vend…

#2 behind Oracle, Mongo with a customers choice award. Where’s Cassandra?

Database Rankings.

https://db-engines.com/en/ranking

Cassandra falling and Mongo continuing to climb. Mongo is the absolute King of the Mountain in NoSQL and Document Rankings.

Slides 10&11

https://investors.mongodb.com/static-files/b08b46b9-e74b-42b…

Look at the skills listed on LinkedIn and the NoSQL activity on StackOverflow. And then look at the Mongo Community Server download.

And how is being better at anybody else at monetizing your Database platform not an indicator of winning.

This isn’t my view or opinion, Mongo is the category crusher and the Top Dog according to the market.

Darth

26 Likes

We are way way way way way way way way way overcomplicating this.

It’s almost like this is our own (surely to be wrong) version of economic forecasting/predictions.

I think one of the things that has made Saul such an incredible investor for so long is actually his lack of industry (or lately, tech) specific knowledge.

Instead of creating some false sense of confidence around which products/services/companies are going to win, Saul’s been really good at finding mega winners by paying attention to the numbers. Once the numbers change (for better or worse) Saul has been able to quickly change his strategy and I think this is because he doesn’t fall for the “story” or have that false sense of ability to tell what’s going to happen in the future.

That’s where I’m at with my own MDB shares. All this stuff has come out. Who knows what’s right or wrong. I’m holding and going to let the numbers prove which side is right. If the numbers weaken, I’ll move to something else. Sure, I’ll lose a big chunk (10%, 20%, or 30%), but will still be up big overall as long as I don’t trick myself into thinking my silly prediction is right, when changing business results are staring me in the face.

35 Likes

We have been discussing Mongo for awhile now.

Yeah, I know. I don’t post much here, because I haven’t been doing as much active investing recently, but there some useful discussions that I can sometimes shed some technical light on so I often lurk.

And for NoSQL and specifically Document, Mongo is crushing it. There is no argument to that.

Well, I think for document databases, you are probably right. But that’s somewhat of a technical niggle, because there is a lot of overlap between the NoSQL categories.

But I think in NoSQL there it certainly is debate.

Look, I don’t want to be arrogant about this. But this is my job: I design highly scalable software systems, and that often involves NoSQL database. I’m not making the case against MongoDB. I’m not saying that MongoDB isn’t doing well. In fact, the opposite. I’m saying that MongoDB, as a corporate entity, is executing really well. And that helps with things like Gartner and creates a virtuous cycle.

But Mongo’s not completely dominant. The NoSQL database market is still pretty fragmented. There is no big dog, in the way that Oracle dominated relational databases. This is good news for MongoDB because it means that there is still market share to be gained.

I believe that MongoDB is a very strong investment right now. I’m just saying that, as someone who works with NoSQL for a living and who spends a lot of time with IT executives, I don’t see a single dominant player. If I were an investor, I’d be much more impressed by slides 17-20 of that presentation than slides 10-11.

If you are investing in MongoDB because it is outexecuting its competition, and is building a powerful business around an emerging market that is growing very rapidly, I think you have the right idea. But if you are investing in MongoDB because you feel like has some overwhelming technical advantage or some overwhelming market share, you might want to reinvestigate how fragmented the market is.

–CH

P.S.
On some of evidence you cite:

  • The StackOverflow survey stuff (which is essentially what was quoted by Mongo in the presentation) is the most convincing to me. But note that it shows Mongo at 18%. 18% is not a “big dog”. Also when you deal with surveys like this you end up oversampling the hobbyists. What you see on StackExchange is not always what you see in the Fortune500.

  • db-engines is better than no evidence, but’s largely just tracking social mentions. I find the fact that it ranks FileMaker higher than DynamoDB and Neo4J just kind of proves the point here. The recent controversy with Amazon could also be affecting things here.

  • The community downloads can be deceptive. Because they aren’t very indicative of usage. I could have a massive MongoDB database and not download a single new copy. (Or I could use Atlas and never download at all.) Conversely, if Mongo issues twice as many patch versions this month as last month they could double their downloads. I know this sounds silly for me to say, but 10mm downloads last year actually doesn’t sound impressive to me. The open source project that I personally work on had more than that last MONTH. It just highlights the fact that download numbers are meaningless because the project I’m working on has no where near the brand recognition as Mongo. If you had asked me to guess how many downloads community MongoDB had last year I would have guess a lot more than 10mm.

26 Likes

The iPhone was not the first smart phone…

For me it’s Atlas that makes MDB a good investment.

Denny Schlesinger

2 Likes

Nice post.

I imagine this is how i sound when i talk about nutanix. :slight_smile:

Darth,

I was looking at the slide presentation and it is old. I mean not old in conventional time but in internet time. Mongo has now doubled their number of customers, has multiple more consecutive quarters w ARR of > 120%, and has accelerated its growth rate.

The presentations timeline is really before version 4.0 was really in GA. General Availability. 4.0 really puts Mongo in rarified air to the point that there is no general purpose database that is now a true substitute w out severe compromises.

As for Cassandra, it is accurately stated that it has fallen down the ranks. The DB rankings may not be perfect but it has demonstrated to be very useful. Mongo ran up it, Postgres ran up it, Reddis ran up it and Elastic did as well, from nowhere. Real world usage supports the rankings and the timing of the rankings.

Cassandra has its use cases. Primarily w big data, and this will often be chosen for huge data sets that do not need a general purpose database and all its functions. And that is a lot of use cases. So in the corner of the world focused primarily on scalability as #1 it may look like Cassandra is right in there. It is not. It has its use cases that it is best for and does well in those niches.

What counts in the end is enterprise usage and enterprises paying to use. And that is a very small subset of all users. Many have very strong feelings about Mongo for many reasons, some related to past issues that no longer exist, some to bias, some just use something else, we even have those who thing SQL rules and can do nearly everything Mongo does.

Follow the numbers and if you do the slide show investor presentation became conservative as to where Mongo has gone since the release of 4.0.

Mongo owns the code for the DB. That gives them a huge developmental and marketing advantage over any other open source product and it shows.

Remember how we discussed, okay, what can substitute for Mongo. I even found a link that listed a dozen different database and except for couchbase they all had material compromises to be made or deficiencies and really could not be a true substitute as a general purpose database for Mongo. And Couchbase does not have most of the features in 4.0 much less upcoming 4.2.

Has AWS out out an emulator product for Cassandra or Couchbase? I don’t know. Probably not. They have for Reddis (another top 10 database). Speaks for itself.

Tinker

10 Likes

“I design highly scalable software systems, and that often involves NoSQL database.”

Are you seeing Foundation or Scylla starting to come up in competitions? Also, do you find distributed queue + stream processing - like Kafka and Spark/Flink - taking more of the big data architecture?

To your point about db-engines being flawed (and I am curious about your thoughts on this, if you have any you’d be willing to share) but I’ve been thinking that s3/Athena and Kakfa/ksql are emergent threats to document stores in simpler workloads, and - admittedly, understandably - they aren’t even represented there.

Great posts.

The open source project that I personally work on had more than that last MONTH.

Could you shed some light on this. 10M downloads to my non-insider self sounds like a lot of downloads, particularly in one month. Again I recognize this isn’t an all important metric for showing usage dominance, but I do find it relevant in building the market picture.

But note that it shows Mongo at 18%

That’s not mentioned in that slide that I could see. The slide is about activity on stack overflow activity. Over 100,000 items have been tagged for MongoDB on stack overflow.

https://stackoverflow.com/tags

If you put in MongoDB it shows 109,351 tags.
“Bid Dog” AWS DynamoDB gets 6,580.
Cassandra (which I think is th done you said you used) gets 15,576
Javascript get 1.75M (the opportunity)

That matches up to the slide graph.

Looking at the Stack Overflow Developer Survey

https://insights.stackoverflow.com/survey/2018/#most-popular…

Of 66,000 developers surveyed.

Mongo DB is #4 in use at 26%
The top three are MySQL, SQL Server, and PostregeSQL at 59, 41, and 33% respectively. All SQL databases.

The next NoSQL are Redis and Elasticsearch at 18% and 14%.

Amazon DynamoDB is way down at 5%.

And Cassandra at 3.7%.

As developers use more than one database, the numbers obviously don’t add up to 100%.

Under Most Wanted Database

https://insights.stackoverflow.com/survey/2018/#most-loved-d…

The numbers here add up to 100%

MongoDB is number 1 most wanted of all databases with 18.6%.
Elasticsearch is number 2 at 12.2%
Cassandra and Amazon DynamoDb both return about 6% on the most wanted list.

MongoDB is the most wanted database.

Stack overflows words, not mine.

Tech speak aside, As far as I know this is probably the best source for where the market is on databases. I’m not aware of any market share reports that aren’t behind pay walls. If there is other or better information I’m open to it.

Darth

7 Likes

Here is a Gartner peer insight report on Mongo vs google. Note: you can add different vendors to compare which ever companies you want.

https://www.gartner.com/reviews/market/operational-dbms/comp…

If we assume the survey’s submitted from customers is representative (and proportional) of all customers, it can give some good insight into competitive advantage. (This could be a big assumption, but I think it is reasonable)

Mongo has about 8000 customers, with 537 surveys’s it’s about 8% of Mongo’s customers.

A couple of observations

Mongo has 7 times more reviews, so about 7 times more customers
Mongo also has about 7 times more large customers (above $1 Billion)

both have about equal customer rec. rates of 80% and 84%

** Be careful with drawing too many conclusions about # of customers, average customer size etc. based on the PeerInsights data. I know from personal experience that many companies try to game the results on these sights. I worked for a small security firm that did just that, putting a campaign with their field sales together to go after all of their best customer base to submit reviews. The instructions were to send emails with links to the survey and call your customers that would be 5 star reviews and we were to do it relentlessly to up our rating. And it worked.

There is value in the ratings but you will see a lot of oddball companies in different categories jump the ratings with orchestrated marketing.

4 Likes