On the surface it was all backward Thursday night.

First, Amazon (AMZN) gave us one of the biggest misses I have ever seen, a wholesale retreat, almost comical in its lack of relation to the profitability people were looking for, a gigantic step backward for this $100 billion company.

Starbucks (SBUX), on the other hand, gave us a strong number with 7% comp sales that was able to mask the European problems that we knew were plaguing it. Looks like there's going to be hope of a quicker turnaround after all.

And Facebook (FB)? Where's Billy Mays when you need him? It was an infomercial for the company's sponsored-stories answer to the desktop-to-mobile gravitation that it was so not ready for that you would have thought they were selling a Shamwow or something, all new, all different -- yippee!

Of course, Amazon's stock got crushed immediately, as it should have. Facebook's stock hung in there. And no one knew what to do with Starbucks off the headlines. Didn't take long to figure things out.

First, Amazon reversed and headed up. Why? Because it is spending to build a nationwide chain of warehouses, because it is going to be the e-commerce backbone for the nation. The warehouse strategy is working. You put them everywhere your customers, including the incredibly fast-growing third-party customers, are satisfied, you lower your shipping costs rather dramatically and you increase your gross margins.

Amazon's biggest problem is a high-quality one. It has too much demand.

Starbucks? It all hinged on June and early July. Starbucks hit some sort of wall in the U.S. and it seemed somewhat inexplicable and the call got out of control, including one moment when Schultz called for a rewinding of the last five minutes because it started to sound like SBUX fell off a cliff in this country in June and the fall continues in July. The company didn't see it coming. The company seemed not to have a plan ready for it. And Schultz, while stressing that sales only "moderated" (a word subsequently used endlessly to correct the misimpression that it is bye bye Starbucks, instead of buy buy Starbucks) seemed worried. He said that he had checked in with other friends in retail and only Whole Foods (WFM) had a strong June.

Given that he's talking about an overexpansion in Europe that he now has to cut back like he did in the U.S. (except this time trying to figure out what happened in the U.S) it was a horrid conference call even as comp store sales were up 7%, which was pretty phenomenal.

Confusing. Bewildering. Not hopeful.

And then there was the non-call that was Facebook. This call was all about trying to prove to the company and to potential customers that the darned thing even worked on mobile, which is where the whole of the almost-a-billion people who use the thing are headed.

Reeling from Zynga's (ZNGA) blast, the Facebook folks presented evidence from two gaming clients as well as an NFL-affinity credit card program that it works. I was astonished and thought they were kidding. It was a missionary call, all about how they have to prove to clients that their ads are working at all. I started to think that advertisers feel they have to be in Facebook because they have to, not because they want to, and I have no idea if the sponsored stories and news feed with ads even worked. They are getting a million dollars a day from the ads, but funny thing, that's not a lot of money. Remember that Google will have $44 billion in revenue this fiscal year, and I would say that's a lot more revenue per day than you get from Facebook, as, shockingly, if you divide $44 billion by 365, you get a better ratio, like $120 million a day. What's the diff?

Plus, if you are doing missionary work, you are not going to make all that money off of the $365 million, as the shrinking gross margins clearly show. Of course, there are plenty of other advertising dollars around, and there is $600 billion in ad dollars up for grabs, but if they have to convince potential clients of their worth, they'd better hire a bigger sales force, to which I say wait a second, wasn't the big pull here that users generate content and you pay to be near it?

In other words, I thought margins would expand, not decrease, as more and more people became Facebook users. That is definitively not the case. This was not a prime-time story. These guys are rookies at making money and all pro at pleasing the users, and that's not necessarily simpatico with profits.

So it was a baffling night. You have a company that used the call as a selling point for potential news feed and sponsored story content on mobile. You have a company where sales moderately stumbled in June, although it sounded like a cliff jump, and a company that's taking over the e-commerce world -- all in one night.

Let's just say it was the strangest night of the earnings season and leave it at that.

Jim Cramer is a co-founder of TheStreet and contributes daily market commentary to the financial news network's sites. Follow his trades for Action Alerts PLUS, which Cramer co-manages as a charitable trust and has no positions in stocks mentioned.

More from TheStreet.com

• Facebook proves it stinks to be a long-term investor
• 5 dumbest things on Wall Street this week
• Cramer: Buy Google, not Facebook