https://theantistartup.com/i-stopped-advertising-everywhere-and-nothing-happened/

I turned off ppc when we had a staffing shake up in Aug last year. It's not been back on since and we've not missed it, as we have some pretty good momentum now.

I can't speak to recent practice, but about 4-5 years ago we changed agencies and they underspent* by about $25,000/mo for the first couple months and it was about a 5X reduction in revenue. We had always wanted to do the experiment to turn off PPC but didn't have the guts. So we got our experiment.

* - we had an ROI target based on profit margins and had fairly dialed how much we could spend and stay within that and so with the sudden drop, we got way better ROI of course, but lost a lot of revenue.

The ideal is to use your paid ads to reach relevant places where you don't already have an organic presence.

>> ROI target based on profit margins

Yes. The metric that matters is net profit at the end of a given time period. An "improved" ROAS might or might not be A Good Thing when viewed in the full context.

And that was part of what happened. The new agency wanted to wow with ROAS numbers. The problem is that we had decided based on the cost of other channels, that any sale with an ROAS of 6:1 was worth it. If ROAS fell below that, we could fill the hotel cheaper in other ways (e.g. promotions on Booking.com or whatever). But anything above that was "cheap" and we wanted it. More generally, we also had some guidelines on what each customer was worth based on 5-year "lifetime" value of a customer. So we could in some sense break even or even lose a little money on a customer's first stay. Not ideal, obviously, but it was at least on the table depending on conditions.

Of course, the easiest way to goose ROAS is to bring ad spend way down - underbid so your CPC goes down, but the number of clicks plummets. Only bid on, say, brand keywords. Things like that. So we had to teach them that we have an inelastic product and we are not optimizing for ROAS, we are optimizing for sellout. As such, a customer is worth a certain amount to us. As long as they are below that amount, we want to "buy" that customer.

As long as PPC was competitive with other channels, we wanted to keep spending, even if that additional spend drives down ROAS and other metrics they like to show in their monthly reports. If they ran out of budget and were still acquiring customers at our target price, we considered that a win and we would find more budget. If there were not customers to buy at that price and they could only spend half the target budget, they should spend half the target budget.

Our company eventually built a team and brought it all in-house, because it was hard to find agency people who could really understand the business and not optimize for the "habit" metrics. Of course, that's available for a price too, but at that point it was cheaper to have the full-time team of 3-4 where each person had only a handful of properties and collected a salary, not a percentage of ad spend (which also leads them to always tell you that you need to spend more).

Obviously, some things are industry specific. A hotel sells a product that expires every single day. If you have 500 rooms, you can't average selling 500 rooms by selling 250 one night and 750 the next. The goal is to sell out for as cheaply as possible. If you have a more elastic supply, I can imagine you would look at PPC spend very differently.