There are many shenanigans that startups can play to juice ARR to render it meaningless. The idea isn't terrible, but that's why I invoked Goodhart's Law.
> After all, sales is sales and if you don't have repeatable sales you probably don't have much of anything.
That's right, but you're missing the key that it's a necessary, but not sufficient requirement. When the pool of investments is 90% with ARR are honest and straightforward, then it's good to use ARR as a metric. But when the pool changes to 50% or less who are honest (as Goodhart would predict), the metric loses value.
I can think of a few; 1) make 2 year contracts cost the same as 1 year contracts (so you're lowering the price by 50% but get the number you want for your report today); 2) give incentives equal to the value of the contract (our software costs $30k but we'll give you $30k of AWS credits).
Two companies pay each other absurd amounts of money for their (nearly) zero marginal cost services, netting no money transferred, but loads of recurring revenue for each.
Your #2 hint at this, but paying more than lifetime value of a customer to acquire the customer.
> After all, sales is sales and if you don't have repeatable sales you probably don't have much of anything.
That's right, but you're missing the key that it's a necessary, but not sufficient requirement. When the pool of investments is 90% with ARR are honest and straightforward, then it's good to use ARR as a metric. But when the pool changes to 50% or less who are honest (as Goodhart would predict), the metric loses value.