What you want to find is a predictable way to generate sales where the economics work. 1000x this. Test on a per-channel basis, ie. marketing spend per custom

What you want to find is a predictable way to generate sales where the economics work. 1000x this. Test on a per-channel basis, ie. marketing spend per customer acquisition, immediate customer value, (conservative) value over customer lifetime. Frankly anything but the initial sale is largely bullshit, but investors love it if you can show nice numbers here. At my first startup (China-wide multilingual hotel reservation platform) I had an angel work through this with me and come to the same conclusion I did: we'd tried pretty much everything, so customer acquisition was either going to be very slow organic or expensive and cashflow negative / capital intensive, save some kind of channel partnership deal.