A few thoughts on MDB

Investment Analysis Clubs Saul’s Investing Discussions
21

SQL is very good at what it does but it has lots of problems because it is a straitjacket when you need to make changes to the data structure. Document storage is based on the principles of OOP which, if properly implemented, is much easier to update.

I beg to differ. Any data you want to search by needs to be accounted for up front. If you forget something or requirements change, you’re in the same boat as SQL. I don’t have direct experience with Mongo, but with the Amazon equivalent, such changes were a much bigger headache than with a SQL database.

In fact, I didn’t see what Amazon offered in terms of data structure that SQL didn’t by just using a CLOB field. You still need to pull some data into individual fields to make it searchable. What it did offer was a way to more effectively distribute data across storage, but for a vast majority of the market, this is a non-issue.

SQL is being disrupted by Document storage. If MDB Revenue’ is 0.7% of the total market, that’s FANTASTIC news! A huge market to grow into.

Because of the above, this is misleading. Mongo solves new problems, or problems that SQL wasn’t good at solving. But it will not replace SQL, I’m not sure it will even erode SQL that much (except among IT trend followers who inevitably overreach). I think they are less disruptive than an engine that grows the entire database market.

I still think Mongo’s prospects for growth are good, so I’m still in.

5 Likes
22

By that I mean that some day SQL might be obsolete.
NoSQL databases in some form are in existence from 1970’s. So this is not a new technology as many would want to believe. However, many technologies had to wait for the right conditions or right challenges to thrive. With internet, came large number of transactions and it created the problem of storing these records quickly.

A key difference between RDBMS and NoSQL is, the data is organized, structured, classified and stored, that means there is lot of work before storing the data. Whereas NoSQL just stores the data and does classification when you need to access the data. This solved two issues, one very fast data recording, no need to organize the data, just store it, secondly because no need to organize data do not have to be similarly structured, you could have two customer documents in completely different format. It allowed different formats, rapidly changing data structures all grouped together.

There are more unstructured data than structured data. But, I do see RDBMS and SQL definitely have a place (you can enforce very fine grained security controls on who can create data or what data can be stored with RDBMS and today you cannot do that with NoSQL) and going to be there for sometime or even better the database technologies going to evolve such that it can handle both structured and unstructured data simultaneously.

With useless technical details aside, the growth for NoSQL is going to be good relative to RBDMS, yeah pun intended. But RDBMS is also growing and Mongo to gain a market share even in NoSQL market needs to be seen. There are other competitors who are gaining market share. This is a topic for some other day.

I wouldn’t invest in MDB if they didn’t have Atlas-aaS.
I hear you. While you talk about LTBH, often I get the opinion that this board and market in general is more focused “current” growth and could drop the stock and move on, if you slip two quarters in row. But that is a not the main point.

The fact is Atlas runs on others cloud platforms. I know the board has looked at some reinvent sessions and concluded Atlas is far superior to AWS and that essentially eliminates the competition. One thing I know about AWS is, that is not the end of the story, but the beginning. Now, Azure has a long database experience and they own the cloud platform. Eventually they will have a product that will be at par with MDB and they can invest far more than MDB. Now GCP, Mongo’s origin is Google and it surprises to see that GCP is not in the mix as much as AWS and Azure when people talk about cloud platform. In other words all the three main cloud platform vendors are serious competitors, and time is their friend. One should be careful to extrapolate current success into future (as in 3, 5, 10 years) market share.

Lastly, if the market falters and share price suffers serious decline, someone might takeout MDB. So don’t assume long-term is given.

5 Likes
23

Lastly, if the market falters and share price suffers serious decline, someone might takeout MDB. So don’t assume long-term is given.

As I said somewhere above, “Sell when the story has changed for the worse or if you discover you have made a mistake.” That also applies to AWS, Azure, or Google eating Mongo’s lunch. So NO, I don’t assume long-term is given. Right now I consider MDB one of the safest SaaS stocks. The story might change. :wink:

https://discussion.fool.com/see-my-other-post-on-the-tam-for-nos…

Saul started this board because he became frustrated with strategies that didn’t allow for changing one’s mind. As I have stated more than once, one of Saul’s strengths is that he is nimble. Just because the story might change someday, or because one might be making a mistake today are not good reasons for paralysis. I presented my technical analysis only to support my argument. I’m happy to have people disagree with me. That’s what boards are for.

I have nothing more to contribute to “Thoughts on MDB” at his time.

Denny Schlesinger

8 Likes
24

Any data you want to search by needs to be accounted for up front.

I forgot that I meant to mention this in my “quibble” … at least some of the top relational databases are increasingly capable of a variety of changes while on-line, including schema changes and adding indices.

1 Like
25

are increasingly capable of a variety of changes while on-line

The key difference is such a change is “administered”. In NoSQL, if the income data changes, it has no impact, and the subsequent query or report can utilize the additional data without any “administration”.

26

Sell when the story has changed for the worse

Denny,

The share price change may or may not have anything to do with the story fundamentally changing. One of my biggest frustrations is, when a stock gets mispriced, it is taken out of your hand, and in the process you have a loss. To give an example, for some reason (like a breach), MDB shares crash to $50 and someone takes it out at $75, If you are someone entered around $100 (which is still lot less than today’s price) has a permanent loss.

This is a hypothetical scenario, but one I have experienced on many stocks over the decades.

In any case, safety is one’s perception. Good luck with Mongo.

1 Like
27

In any case, safety is one’s perception.

I promise this is my last reply to this thread! :innocent:

“The only safe place for a human is in a six foot box six feet under.”

– Dr. William Thomas, The Eden Alternative
https://www.edenalt.org/our-team/

To anyone looking for a retirement home I strongly recommend taking a look at the Eden Alternative. I had the good fortune to attend one of Dr. Thomas’ workshops and I can personally recommend them.

Good luck with Mongo.

IRR 102.6% and it’s not luck. A year or two ago Saul posted that he had finally bought MDB after reading a bullish thread right here. I was one of the bullish posters in that thread. Before that there were bullish MDB threads at NPI and I consistently refused to join that crowd based on a Freemium business model because MySQL never was profitable as an Open Source business. They eventually sold out and it became part of Oracle. It was only after Atlas showed its mettle as a SaaS business that I became bullish on MDB.

I know that wishing luck is done with the best intentions but I find it condescending in a discussion such as this one. “Investing” is work, not luck, and stock picking has been the most difficult job I have ever tried.

Denny Schlesinger

Please do NOT reply to this post.

18 Likes
28

Yes, the document database can accept a document in a new format. That doesn’t mean every field in it is indexed, though, does it?

29

That doesn’t mean every field in it is indexed
I think you are looking it upside down. Say you accept user details as a document and sometime later you introduce a field (on your web form) to upload user profile photo and now you store the document with photo and a column showing whether the photo is uploaded or not.

So you can choose to index or query based on whether a photo is attached to a user profile. This requires no change on the NoSQL side, unlike RDBMS, where you have to make changes to the table structure to capture the new data. Until the change is made you cannot capture that data.

Indexing is something you choose to do for faster access or create some other relations but not required.

30

And, with an RDBMS one can add the two fields on-line as well. The point about indexing is that one has to do it in either case if one is retrieving on large data sets. No?

The RDBMS is certainly more formal. I am just questioning the idea of reporting on a data item in a large data set in a performant way without having to do some work.

31

Values from their last 5 years of quarterlies:


                    2019 7/31 2019 4/30 2019 1/31 2018 10/31 2018 7/31
Sales/Turnover (Net)    99,368	89,388	85,484	71,782	59,611
Cost of Goods Sold	25,477	24,880	23,109	16,047	15,688
Gross Profit (Loss)	73,891	64,508	62,375	55,735	43,923
**SG&A 	               106,607	91,447	83,738	72,285	70,221**
32

Flying Circus, true, MDB is still unprofitable. But if you look at the numbers and percentages, sales are growing much faster than expenses, percentage wise. This is all part of their strategy to become a de facto standard as quickly as possible.

6 Likes