This is the key point, isn't it? A couple of questions:
1) For most moat-y businesses, do they have to pay their customers to choose their business?
2) For the ones which do, in fact, pay their customers (perhaps just initially) - what do those businesses look like?
In my opinion, for (1) most moat-y businesses do not need to be acquired. You are offering something so special that competition is limited. Either you have some special process to offer sustainably lower costs (i.e. best commodity producer) or you are so differentiated that you are offering a special product (brands, etc.)
Now, yes there are some businesses where you pay to acquire. The question then becomes, which are successful, and why? I would argue therefore the response to (2) is that you are building yourself into long-term customer processes, with high switching costs/specialized service, and a high natural relationship length. It becomes uneconomical to pay for customers if they can just hop to a competitor (low switching costs) in a month (low relationship length).
My problem with the VC world (and perhaps I am wrong) is that they over-estimate their competitive advantages. I am sorry, but food delivery is not a competitive advantage. And access to capital is almost never a competitive advantage. Real competitive advantages come from high skilled activities at least in most cases.
I think there is potential for significant competitive advantage in creating a network that makes delivery better, if it makes delivery better. Despite what some might argue, I am probably never going back to ordering a taxi because Uber/Lyft are so much more convenient that I would probably pay a premium to a taxi. Uber/Lyft made transportation a better experience IMO. They also made the driver experience better by making it a more flexible job, eliminating expensive medallions, etc.
The open question for me perhaps is how are the food delivery services making the experience better than the status quo? Is calling up my favorite restaurant and going to pick it up that much different than having someone deliver it? I see the two-sided network effects very clearly with something like Uber, but food delivery not so much, especially if restaurants are eating the delivery cost. It is at best another form of advertising for them in that world.
What makes Uber better than a taxi? Uber is a taxi service with an app. There is nothing more to it. Uber is just the app-ification of a service that's been around forever. The Uber business model wasn't related to network effects (even if they claimed that), it was about just subsidizing rides relentlessly to get to a scale that local laws didn't apply to them anymore. In fact, if you read about Uber, everyone can see the network effects never existed. Even small price increases (or smaller subsidies) causes Uber to lose market share.
I would argue Uber just took advantage of another Silicon Valley business model - ignoring regulations. Like how if you throw a Coke can on the ground, you are littering. If you have a scooter company where 10,000 scooters are thrown on the ground, you're a mobile transportation company.
I think it's less about the direct comparison between Uber vs. taxi services, and more about Uber allowing a massive number of underutilized assets (personal cars) to be monetized. Network effects come into play here (although likely weaker than the VCs originally believed), as it's a marketplace and not a fleet manager. The economic sensibility of this (what Uber charges vs. driver salary + vehicle depreciation + variable costs + whatever else) is another story altogether. In my last gig, we used to joke that the best way to help Lyft (one of our portfolio companies) was to take more Ubers.
