Many fields have standard guideline ratios for numbers. For example, in the catalog industry the rule I hear is that all costs excluding labor should never exceed 70% of revenue for a product.
Does anyone know of any such guild lines for the software world, where alot of cost are fixed.
Guild-lines regarding monthly subscription and LTV would be great as well. For instance, I would love to know averages for how long monthly auto-billed customers stick around for. It would be good to bench mark my business against it.
For SaaS the answer is tricky, because the thing you're selling is mostly sunk costs (previous development) and fixed costs (maint, new features), whereas the catalog industry has a lot higher variable costs. Also, rules of thumb will vary wildly based on deal size of the company, how many deals you have to sell for breakeven, support structure, etc. (e.g. highrise vs. salesforce).
Here are a couple of rules of thumb I've run across in the world of deal sizes between $15k-$100k/year:
- Keep per-sale sales costs (including commissions / bonuses) below 70% of first year's gross profit (revenue - truly variable costs like variable bandwidth fees). Stay below 40% and you're doing well.
- One customer service rep can handle 20-30 customers for a fairly high-touch, dedicate customer support setup
Maybe the most important question is: how many customers do we need before we can reliably cover our fixed costs with their recurring subscription? (no rule of thumb here)
The problem here is software == technology, software != product. The catalogue industry may have guidelines for their margins, but with SaaS -- are you selling entertainment (Facebook, Youtube), expense reduction (Mint), profit growth (Salesforce)? All will have different figures...
One number that I did hear a few years back was 15 USD annual revenue from well-targeted ads to a known user (i.e. per-user value of targeted ads on social networks with good data mining possibilities). This was 2 years ago however, so the number may be lower today, or more sophisticated systems may be required to get that revenue per user.