Completing tests quickly will help you run multiple tests in succession
Your success will be a function of the learning cycles you fit into
your runway.
— Tom Hulme, Ideo
The best way to discover what makes your prospects tick is to run
an experiment. If you are pressed for time, the best way to
discover what makes them buy is to run many experiments over a very
short time period.
Assuming these two would cost exactly the same amount, which would
be more valuable to you as a product manager?
1 test, which runs for one week, which gives you 99% certainty
on your results
30 tests, which take one hour each, with 90% certainty each
Clearly the second set of tests is much more valuable. When you
finish the first test, you can form another hypothesis and test
that. One set of results directly feeds into the rest. If you want
to identify surprising results quickly, as this is most likely to
give you outsized results, option B is a much better way to enter a
market.
To some extent, it’s reversing the iterative nature of agile,
and applying it during pre-development when performing product
research. Feedback loops are enormously powerful when dealing with
prospects and customers too.
“Market Testing Velocity” can mean a lot to you in a
high-pressure situation if you want to get something going quickly.
If you complete tests quickly, you can run multiple tests in
succession.
In this case, you are doing marketing research on your
prospects or existing customers, in order to map out what they
need. Your goal is to identify what they need most, a
“bleeding neck”, as quickly as possible. You are trying
to discover a cognitive and emotional map of their needs, in order
to orient your own efforts most effectively.
If this tickled your fancy, and you’d like to read more about
market testing velocity, you’ll find out more in Launch Tomorrow.
A friend of mine from when I was growing up is a true blue professional hacker for the US military. He gets paid to hack security software. Specifically, he writes software which figures out how other security software works. He figures out what the requirements are. He prods the software, notes down its behaviors, and…
You may not have not realized it, but launching a new product-based startup requires that you navigate a number of chicken-and-egg problems successfully. Which comes first? While seemingly logical, there are a number of preconditions which you need to meet in order to stop being a startup and become a real business. I once visited…
Just wanted to privately share some tactics to help you counter Google’s aversion to opt-ins in your landing page MVP. It’s hard enough getting a startup idea off the ground, to have to deal with arbitrary Google Adwords funny business. If you are using the Launch Tomorrow process, email addresses have always been seen as…
Most startups fail. About 1 in 2 survive 5 years. In IT, that’s probably more extreme. How many companies do you deal with today, which are over 10 years old? Remember how many search engines there were in the late nineties, most of which don’t exist, or at minimum, don’t exist any more as a…