Robert Beerworth Team : Web Strategy Tags : Clients Performance Analytics Rants

Metrics and benchmarks for startups: pity they probably don’t apply to you

Robert Beerworth Team : Web Strategy Tags : Clients Performance Analytics Rants

I get asked a lot by startups about the sort of numbers they should be targeting and/or expecting.

This is almost an impossibly difficult question to answer because there are simply so many variables:

  • The effectiveness of their traffic process and strategies.
  • The competitiveness/effectiveness of their product and/or service.
  • The client themselves.
  • The exchange rate of the Thai Bhat and what phase the moon is in.


It is of course entirely possible and entirely recommended that a model be developed so that different scenarios can be modelled; from the number of customers on the website, to the average margin per customer, to the churn rate of the website.

But what is a good number? And what are the right numbers to be putting into the model?


Real numbers from real startups

I came across these numbers this morning, from a survey of over 1,500 startup, SAAS businesses. Wow – real numbers from real startups.

This is gold!

The data is from 2013, though that is not to say it wouldn’t still be accurate. Key findings:

  • The marketing mix and what worked and didn’t work is totally varied. No surprise there.
  • For a B2B website, the average monthly revenue per user was $140.41. For a B2C website, it was $16.90.
  • The average, monthly churn rate was 3.2%. I have consistently heard churn rates around 8 – 10% so not sure if the B2B websites were pulling this reported churn rate down or what.
  • The ‘Visitor to Trial’ conversion rate was 8.4%. The ‘Trial to Customer’ conversion rate was 11.0%. This second number seems high though again, this might be a reflection of the B2B respondents to the survey.
  • The average non-trial conversion rate – where a user became a customer without every undertaking a trial – was 2.2%.


I prefaced the list of key findings by saying that just because the data was from 2013 didn’t mean it wouldn’t still be ‘accurate’.

And herein lies the issue with the data.

Almost 80% of the respondents were B2B websites and so the numbers are skewed. B2B websites performance is quite different to B2C in all regards.

If the data could be provided without aggregation, startups could better analyse their segment and get much closer to the sort of outcomes they should expect and target.

It is great that someone made the effort and having a number, any number is a good start. My feeling however is that is all it is.

Sorry folks. We all need to keep looking.