And besides, the field of competitors is getting smaller, not bigger. See: Atlassian throwing in the towel on HipChat and Stride.
The depth of those integrations serve as the depth of the moat. If your integration is basically "I send and receive textual messages", which from what I see the vast majority are, you've got an integration which could be ported to a competitor in less than a week. A lot of the integration work is in the thing being integrated rather than chat-system specific things.
As you use more Slack-specific features the moat gets deeper, but you do have to stretch a bit before you're really "slack-specific". Rich text formatting, conference rooms at a basic level, etc. isn't "slack-specific".
(Be sure to read that first sentence of mine a couple of times before replying. It does not say "integrations are always trivial to port and are thus irrelevant", for instance. If I wanted to say that, I would have.)
"Atlassian throwing in the towel on HipChat and Stride."
I strongly suspect that Atlassian threw in the towel not because it was hopeless, but because a highly profitable company has a business imperative to avoid things that will merely be sorta profitable to them, or where profit will be harder than their other highly profitable activities. It's why Google doesn't simply do everything; put over-simply, for something to be worth Google doing, it has to be more profitable than putting another engineer on AdWords. When a company is valued at $7.1B, it validates the market, and even if there was no competition today, there most assuredly will be in two years!
(I work for a company not quite as high as Google in the "necessary profit per engineer" metric, but we've also divested "merely" profitable businesses because they were basically a drag on the rest of the company, and lacked sufficient strategic value to overcome that disadvantage.)
Off topic, but when you write something like this the message you’re sending is “I’d rather insult the reader than take the time to communicate in a way that can be properly understood.” Hopefully that’s not the intention.
There is no other way to communicate nuance that I've found. If you claim that "some types of X are Y", you'll get angry replies about how some X aren't Y, angry replies about how terrible it is to claim that all X are Y, and on more than one occasion, angry replies about how some types of X are Y.
If you're insulted, well, then, it's not targeted at you. But abundant experience says it does have non-zero targets. And it's getting worse, not better, as more and more people are convinced(/programmed) to believe that reading people as harshly as possible and then being as angry and offended as possible is a virtue rather than a vice. As these people get more power, it's becoming simply self-defense to make sure you underline your point.
If I tried to write down all the other corporate-focused chat products that preceded Slack, there would be no way for me to write them all down. To say nothing of the chat products that weren't necessarily corporate focused, but could easily be used in a corporation. They were not first to market; they came into a crowded space, but one where nobody seems to have quite put all the pieces together the way the market wanted, and proceeded to dominate.
If I would have any concern about Slack as an investor, it would be that they don't have a terribly great moat, in part precisely because there are so many companies sitting on a code base that has 90% of the hard technical work done. Personal chat products had a certain amount of stickiness because you had to be on the service that your friends were on. A corporation can simply move wholesale by executive fiat, and the bigger the customer, the bigger the savings number will be and the less connected the buyer will be to the people demanding Slack in particular. I would expect Slack to be facing some pricing pressure in the medium-to-long term as the abundant field of competitors starts essentially copying from Slack what works and starts tweaking their products to fit.