What I’ve done the last two days.

What I’ve done the last two days.

First The Trade Desk. Here’s what I wrote in Friday’s Summary:

The Trade Desk announced stupendous results Thursday after the market and rose $47, or more than 31%, on Friday. It was a new try-out position four months ago, at the end of October. Now it’s now a 7.1% position and is in 6th place in my portfolio. I posted a deep dive two and a half months ago. I’d rate them four and a half stars based on their results and their confidence in themselves, but they are an advertising company after all, which is a field that I have zero confidence in, even though I feel that this is a very innovative and creative company, so I doubt that I will ever let the position get very big. The Trade Desk seems to be a Leader in a Rapidly Growing niche Market within the larger field of advertising, which up to now is controlled by the behemoths.

However, the more I read over their results and the conference call, I decided that this company seems for real, and I wanted a larger position. It seems to really be the leader in its field, which itself is expanding rapidly, and The Trade Desk’s growth seems to be accelerating. However I didn’t have any other positions that I wanted to reduce, so what I did on Monday was sell 3% of the number of shares that I had in each position, leaving each 97% as large as it had been and keeping relative sizes. Bought Trade Desk, and my position grew to 9.6%. If you want to know why I changed my mind, just read the earnings release and the conference call.

Then on Tuesday, came the news that Lyft had left Mongo and moved to AWS. There were three bad news nuggets in that. First that they were dissatisfied enough with Mongo to undergo all the immense hassles of moving when they were already fully deployed, and second that AWS’s homegrown copy was “good enough,” and third, that this may give other companies the idea that maybe AWS really is good enough (or maybe even better) if Lyft moved to them. This bad news follows after that AWS announcement, and concurrent attack by Red Hat, which I’m sure most of you remember. And then yesterday there was also an Azure announcement.

I’ve had a checkered history with Mongo, selling out once last year just because I couldn’t understand what they did, and because they were losing such a large percentage of revenue each quarter. I bought back in at about the same price that I had sold. And then after the AWS announcement and the Red Hat attack, and after reading somewhat unproven stories that a lot of customers were dissatisfied with their service, I sold out again, but was convinced to buy back.

I must say now that I’m fed up with Mongo. I sold about two-thirds of my admittedly small position (3.5% before I sold) and added some of it to Trade Desk (now 11.4%, no more adding!) and a little to Coupa, my smallest position. I have a different take on Mongo than Steppenwulf, who I really respect, and it’s probably because I don’t have the tech expertise that he has, but there are just too many companies around that are simpler stories and don’t have Mongo’s problems. And I will definitely feel better investing in a company without all this stuff going on.

Please be aware that I make mistakes and wrong decisions all the time, and The Trade Desk may sell off after that huge rise, and Mongo may just keep on going up, so make your own decisions.




this may give other companies the idea that maybe AWS really is good enough (or maybe even better)

There are always going to be use cases where good enough and cheaper is a compelling argument, at least in the short run. But, in the long run, greater functionality is most often going to become compelling.

One should also note that moving an application from one document DB to another is a whole different kettle of fish than moving an application to or from document DB and relational.


Thanks for posting this, Saul. I just want to correct one sentence in your post. There is no evidence that Lyft migrated to DocumentDB, which is AWS’s clone of MongoDB. Instead, it appears that they’ve decided to drop MongoDB in favor of AWS’s DynamoDB, another NoSQL database, which they already had been using for several years now.

Chris Lambert, Lyft’s CTO was extolling the virtues of AWS DynamoDB way back in 2016, as you can see from this video:


So it seems to me that Lyft is a MongoDB customer who simply expanded their use of an older database on AWS that they were already using for several years. DynamoDB still has a relatively small share of the DB market, and it appeals to companies who don’t mind being locked into AWS.

I believe companies who don’t want to get locked in to AWS will continue to use MongoDB. I don’t see this move by Lyft as an existential threat to the future of MDB. But that’s just my opinion, and I respect your decision to sell.


I don’t see this move by Lyft as an existential threat to the future of MDB. But that’s just my opinion, and I respect your decision to sell.

Hi rdutt, for a guy like me who is not in tech, it’s just too complicated a situation. As I said, Mongo may do great, but there are companies where the situation seems much clearer to my non-techie perception.



I must say now that I’m fed up with Mongo. I sold about two-thirds of my admittedly small position…

In the investing world, buying shares directly from Saul may be the scariest thing possible. And, it’s possible that just (inadvertently) happened to me!

I also have had a reasonable sized position in TTD, so here’s to hoping both grow!


Thanks for the update. I sold my portion of MongoDB yesterday as well.

I worked with databases in Silicon Valley during the dot com ecommerce boom of the late 90’s, and continued with the company for another 6 or so years (we were profitable).

I’ve worked in the mortgage and banking industry, backstopped by Oracle, developing SaaS software for mortgage brokers in the US.

I currently do database and data-exchange work for a health-care software company, using SQL Server, in the oncology field.

After reviewing and playing with MongoDB, I thought back to the work I’d done in the past to see if MongoDB would have been useful. I could not think of a case where a document database would (or even could) be the primary data store for the problem I was trying to solve.

Maybe there would have been a case for storing mortgage applications for mortgage brokers as documents - it would have sped development up pretty significantly - but easy storage is only part of the picture.

I thought about data science - it’s very popular right now - taking data sets and doing regression on it to see if you can find trends that have gone unnoticed. A document database isn’t good for that - people use Hadoop to do regression, map-reduce, and other techniques to find patterns and analyze data.

I don’t understand what MongoDB customers are doing with the software. How big is that market? Is it growing? Is MongoDB the killer app for that usage? That’s why I’m staying away.

But what I don’t know could fill many books. I thought YouTube was a dumb idea. MongoDB could be the next Oracle. I’d rather invest in companies I understand.



David, just curious, why did you buy MDB in the first place, and what has changed since that point?


1 Like

Based on financial results, and the fact that they were the “new paradigm” of data storage. I’ve only held them about 6 months.

In the fall, and early winter, I installed the software, went through a few courses @ MongoDB university, and did some database development with the software.

I also did some reading - informal case studies - and discovered that many people try to use MongoDB like a relational data store, and that’s not it’s strength. As a result, many of the projects fail or transition (painfully) to a relational database.

I think it’s great for prototyping, and great for storing data where you only ever query the data from a single starting point.

Initially, I thought maybe there were a lot of uses for a document database. I still haven’t found that problem-set that it solves, or at least a problem-set large enough to validate my retirement funds being tied up in their stock.



MongoDB could be the next Oracle. (But) I’d rather invest in companies I understand.

Hi David, (HiTechGuy),
I couldn’t have said it better.

1 Like

Hey Saul,
Curious on COUP.
I have an increased appreciation for Enterprise Productivity software. Common names in the space or discussed here include: CRM, NOW, WDAY, TEAM, ZEN, PLAN, SMAR, and COUP. While not as sexy a stock, you could also include SAP.

The upcoming Slack IPO may also be interesting, but my guess is that it will be highly-valued out of the gate, as cloud/saas stocks aren’t sneaking by anyone right now, it seems.

I believe you were in ZEN before, and they just announced good results. I am a fan of NOW, WDAY, CRM and others from how my enterprise clients view them as mission-critical and how the infrastructure to support is almost an afterthought.

Currently I am trying a less-is-more strategy, and while all those listed are great companies and most are going to be market-beating stocks, if they aren’t materially better than stocks/companies I already own, I am doing my best to let them sit on the bench.

SMAR is the only one from that current group I have a small starter position in. They announce ER in March, and as of their last ER, they were growing the fastest of the bunch. They feel like a “Saul” stock due to that growth.

My question is: given you recently tried and cut PLAN, and now are in COUP…is there a particularly method to why you chose COUP over something like SMAR?

COUP is a 40% grower currently…pedestrian compared to your top holdings (which is amazing to say). SMAR is on track for full year 57-58% y/y growth, in comparison. Neither are profitable yet, and COUP slightly larger in rev and mkt cap.

COUP just seemed like an outlier compared to your other stocks, so was trying to understand your process in choosing them as a try-out position vs others you may have considered?




MDB up 2% on average volume. Mr. Market discounted the Lyft FUD.

Denny Schlesinger

1 Like

I have a deep respect for your analytical skills and what seems to be well honed intuition, but I just flat disagree with you about MDB.

Maybe it’s because I was there in the early days of the relational DBMS wars and witnessed first hand as the IT folks where I worked (as well as the non-IT employees in the divisions that performed IT like work).

We brought in DB2, Oracle, Sybase, Informix, Nomad (not relational), Ingress, Ashton-Tate, FoxBase, and probably others I’ve forgotten or didn’t know about. Out of all of them only DB2 and Oracle were in place a few years after the initial, experimental phase. Most of those others don’t even exist anymore.

You will see continued churn in the NoSQL DBMS space, but IMO MDB has already won the war. Take a look at my post on SteppenWulf’s post about MDB. I’m not very confident we’re even getting the real story about why Lyft went with Amazon, it doesn’t make much sense to me.


Saul never said this wasn’t a great investment, he just sees allocating his money a better proposition elsewhere at present and can’t own them all. I also sold out but when and if you are wrong, can always jump back in.

1 Like

Did Lyft really say they were dissatisfied with Mongo. Cause i only read it from the analyst, and i don’t listen to analyst.
I might have missed it but i didn’t see anything from lyft or mongo on the subject

That’s a good point. What I read was that they had issues with performance. I’m not sure if Lyft issued that statement or someone else speaking for them. And as I stated before, it was not clear to me if they had problems with vendor performance (support services) of DBMS performance (slow response time and such).

Thanks HiTechGuy;

As an old retired tech guy who still keeps up a rather elaborate personal financial db in MySQL your
post sums up my observations very well.
I currently have a position in MDB, but the more I try to understand the overall usage and potential
size of their market the more I question whether the price based on potential may be too high.
Not that MDB isn’t an excellent new product that fills some needs in the overall expanding db
environment, just that it isn’t going to be the answer to most problems.
One site that follows the growth and popularity of the different databases is:
Their ranking system uses number of websites, general interest, frequency of technical discussions,
job offers . . .is defined at: https://db-engines.com/en/ranking_definition
MungoDB is currently #5 with about 8% of the total for the top 10. It has been growing in popularity
but it isn’t the fastest growing db and hasn’t moved up from the #5 position that it was at in 2014.
MDB’s market cap is already 5.4B and they are only 1/3 the size of MySQL an open source which was
only valued at 1B 10 years ago. Obviously the db environment is larger than it was 10 years ago
but the question becomes how much more will they be worth in the next couple of years.




Supports whats already been said.

1 Like