Below is an answer I recently put on Quora… Since I haven’t posted on my blog for so long, I figured some people not on Quora might find it useful.
Your marketing spend should be very minimal until you validate that you have created a product that people want or need (an important exception is for network effect products, which I’ll cover later). I would suggest 95/5 ratio between product and marketing. You don’t necessarily need a marketing person on the team to do this early validation.
Once you’ve validated that people want or need the product, you should spend as much as you possibly can on customer acquisition as long as the value of each user exceeds the cost of acquiring them. Often this requires raising additional funding, but if you can present proof of profitable, scalable marketing channels then it should be easy to raise the additional funding. Of course you should complement this paid customer acquisition with free sources if possible (and you can start these early in the validation process). If your product really does a good job solving an important need, you should also have strong organic growth. At this point the spend ratio generally tips toward marketing. I’ve seen it as high as 80% to marketing and 20% to product.
The exception for network effect businesses mentioned earlier is for the following reason… The user experience for a network effect product improves with each additional user. You may need to reach a critical mass of users before you can validate that the product is important for users.