Business journalism really does a disservice when it pushes the "they have no revenue model" angle for these massively scaling traffic aggregators. It was a bit interesting when this was written of Yahoo, I guess, since we didn't have proof yet that online advertising was a viable business model.
In any case, the playbook is very well understood. Build a loyal audience numbering minimally in the hundreds of millions (Small audience pubs only achieve success with subscription-based models. Mid-sized audience aggregators may not have a viable business model at all). Demonstrate that the cost of building said audience is either free or pennies per user (e.g. unpaid viral audience growth). Make several dollars to tens of dollars per user per year by running advertisements to said audience. The more relevant or (lately) emotionally compelling the advertisements the more you'll earn per year per user.
Of those components, building an enormous audience is the hardest part, which is why every company that has followed this model, from google to snapchat, has focused on that long before building a corresponding revenue stream.
We've seen this model a dozen times and it's understood well enough that financiers are able to approximate the value of Bytedance future revenue streams rather confidently.