How to make revenue and user acquisition projections for product which is not launched yet?

How can you make revenue and user acquisition projections before launching your product?

And are they worth anything?

You should take a look at this recent community wiki entry: How Do I Define and Research a Market and Estimate Market Share.

Estimating your revenue and user acquisition rates comes at the tail-end of estimating potential market share. For that, Jarie puts it best when he says:

Defining your marketplace, doing research and estimating market share are iterative processes that don’t really have any good algorithms to apply. A lot of the time, it’s trial and error and looking at what your competitors are doing.

Take a look at his entry to find some good resources to help you start the process!

You can make up numbers based on total market size forever, if your google/apple/microsoft then your look at these numbers because your going to have a chance at a serious percentage of these in the first year ... Unless you have deep pockets or a "facebook" style viral product then you need a different approach.

A realistic method is to work out "playing forward" from release for 2 years ... How many users will you get in the first month and how will you get them?

Then the next 3 months how many, how do you get them, how much do they cost on average to get, how much do you bring in from them?

You should also look at how many dev, support, sales people? How much in advertising? and other costs it will take.

Repeat forward quarter by quarter for 2 years ... Where are you at? Have a good and a bad version, take the aveage and you should be close to a realistic and workable number.

If you actually do better that's great but this is realistic numbers investors can believe. You can then say "is there that many people likely to bite?" and that is a much easier question to ask (well sometimes it is).

For extra credit do a with and without funding version so that you have a real story to talk to others about.

For the details :

User aquisition costings ... will come down to your product, barries to entry etc ... each product profile is going to be different ...

One of my companies sales are $50K to $100K per sale ... we do several $20K trade shows per year and run a sales team of 2 full time with the aim to get 8-12 customers per year out of them ... after the first few years the service payments back a good baseline reoccuring income.

If your at the smaller $5 per unit end of the scale then you could say "I will pay $1 per customer on average" ... Which sets your thinking around the fact that your not going to advertise on Facebook/LinkedIn, more likely adwords on Google or something in the blogging/viral/non-advertising space and hope to grow organically.

Additional bit because @Alex called me out on it (rightly so).

Looking at pricing models of comparable product models, it doesn't have to be competitors, just similar models to yours

• Selling propositions. Is it a $20 per month Saas or $100K per sale with SLA afterwards?
• Levels of engagement. Are you 24/7 support OR simply a forum/email when you can response times.
• Range of features / items. Does it run a large slab of someones business or does it do 1 thing really well.

Find a range of example companies online, taking a wild stab in the dark and saying your likely to be looking at a hosted/saas model then I would start with

• UserVoice
• ZenDesk
• Balsamiq
• Atlassian JIRA

And you will find many more ...

Others might be able to say how you get hold of their sales figures and costs per conversion ... I don't know that bit ... the article Alex references seems to have some pretty good starting points for that.

There are also a few "rules of thumb" about advertising, general advertising can expect 3-5% conversion, bad advertising can expect up to 1% conversion, spammers work on the 0.1% rule (they just hit enough to make it work), highly qualifed conversions can be upto 20%, qualifed introductions / recommendations from a trusted source can be much higher but I would be making a figure up. (I got these rules of thumb working with specilists in customer modelling and market segmentation who consult to fortune 500s) as I said each is different but you can start there and do lots of testing to work out where you sit.