[zhyder]

We too have learned the hard way about 5.: customer acquisition is tough.

5.1. We'd read it before but hadn't quite internalized the cardinal rule: Cost of Customer Acquisition < Lifetime Value of customer (CCA < LV).

5.2. It's easy to think "hey my product costs me only $1, and the competition charges $100: I'll do great by just charging $10". But when you're charging $10 you won't be able to pay $30 per ad click, and so will (continue to) lose customers to your competition.

[patio11]

I could talk about reason 5.2 for a whole afternoon. It is one of the best reasons not to compete on price.

Additionally, if you're doing ads on Google, there are some weird interactions here that Econ 101 would not predict. For example, although Google describes AdWords as an auction, it is really "auction-like": the highest bid doesn't necessarily win. One factor which can cause lower bids to beat higher bids is if the lower bids have account history. You get account history by winning the auction lots of times.

If you're scratching your head thinking "How does that matter?", consider a software market where one vendor has a product priced at $30 and another has a competing product priced at $22. Their conversion rates and CPCs might very well be similar on day one, resulting in a cost-per-action of $25. One vendor might choose to participate in the auction while the other does not. As the participating vendor's campaign picks up positive history, Google may cause their CPC to decline due to improved account history, perhaps resulting in a CPA closer to $12 ~ $15. Vendor #2, seeing this, might think "Darn! I'd sure love to advertise if it cost $12 ~ $15", and then turn on their ad campaign... only to find that it still costs them $25. (In some cases, it could actually cost more as a result of their competitor's success.)

The above example is somewhat simplified. AdWords can get complex fractal-style.

[lsc]

The interesting part of this 'don't compete on price' advice is that in many markets (the market I am in included) marketing is often more than half what you are getting from the customer. In the example above, instead of paying $12-$15 per customer to goog, what about dropping your price by $10 and seeing what happens? If your product is comparable, it only takes one reddit article comparing you to the competition to get a whole lot of interest.

I can tell you from experience that if you can undercut the competition by half, you get a lot of slack when it comes to not having all the cool features the other guy has.

One of the things that I think many people miss about competing on price is that if you don't have the marketing muscle of the compititon, you need to be dramatic about your price. few will go with an unknown to save 10%; but if you can save 50%, well, that's a pretty good argument to go do some research on that unknown company, no?

(Overall, my position on the 'compete on price' / 'never compete on price' debate is "do what you are good at" - if you are good at selling to people with money, sure. charge a premium and take advantage of your connections and skillset. Rich people and large corporations need stuff, and there's nothing wrong with giving it to them. On the other hand, if you are good at being cheap and bad at selling, well, maybe you should think about attracting the attention of people willing to do research by lowering your price.)

[mbrubeck]

Since he politely didn't shill for himself above, I'll point out that lsc is the founder of http://prgmr.com/ which has crazy low prices for virtual hosting.

[staunch]

Details directly from Google: http://adwords.google.com/support/aw/bin/answer.py?hl=en&...

[zackattack]

If you were to write an AdWords ebook, I would buy it. Like a good one, I could easily rationalize spending lots of money on it.

[patio11]

While I'm happy for the vote of confidence, writing books feels a lot like work to me. Then, after I write it, I get to deal with a lot of email like "WTF dude I can barely afford my ramen and you want $500 for this? That's like $10 a page!? I can buy a sci-fi paperback and get a HARD COPY for only $6!", and that isn't my definition of a good time.

Instead, I prefer using what I know to make money directly (not necessarily limited to increasing my share of the teaching bingo market, hopefully) and possibly doing some consulting. Folks working on their next bowl of ramen can read my blog for free. I get emails saying mostly positive things. Everybody is happy.

[gridspy]

Or even 30c per click when 1/100 click-throughs convert to a sale.