The Trello Template We Use To Fix Our Marketing. (For You to Steal)
Half a year ago, we launched a free t-shirt marketing campaign at Kiwiship.
Just like other campaigns, we promoted it in different channels, and got a few hundreds of user signups after a while.
Everything was smooth, until recently, my co-founder brought this up and asked me “should we do that campaign again?”.
I didn’t know how to answer his question, and I was shocked by this fact.
Yes, we got some sign ups, but how many of these people visited our site again? How many of them eventually became our customer? Is this a good lead magnet? What did we learn from it?
I realized we got some problems here: we were testing different marketing strategies, but failed to optimize our learning from the test results.
(We eventually decided to launch this free t-shirt campaign again. Here’s the link if you are interested.)
Optimize for learning
So, we really need to “optimize” our way to test stuff, so that we can make sure we learn the most from it and move forward.
Here’s how I do it:
First, before you test anything, you need to set an expectation of the outcome. For example: “I expect 1000 views of this campaign and a signup rate of 20%.” It’s OK to guess the number, but you need to have something documented.
Second, make sure you can measure what you want to test. I normally add ref parameter to the links, then use tracking tools like Mixpanel to track user sessions and do data analysis.
Next, you need to set a duration for your test. If you let it run for an indefinite period, you will most likely never figure out whether it works or not. Just run the test for 2–4 weeks, then stop and compare the result with your expectation.
Lastly, you need to label each test: it “worked”, “didn’t work”, or “need to test again”. By doing this, you are forced to come up with a conclusion of each test, and make sure you get some meaningful insights after all.
The Trello template
I found Trello does a great job helping me organize our marketing experiments. Here’s the step-by-step setup:
First, create new board with 4 lists: “Backlog”, “On-going”, “Review now”, “Finished”
Then, create an item in “Backlog” list, name it “Template” and copy & paste below text to its description field.
- **[Test duration]:** (Fill in when create, set due date)
- **[Description]:** (Fill in when create)
- **[How to measure]:** (Fill in when create)
- **[What to expect]:** (Fill in when create)
- **[Actual result]:** (Fill in when review)
- **[Work or not]:** (Fill in when review)
When you come up with a new marketing idea, you can create a new item by duplicate this template in the Backlog list.
When you are ready to launch the marketing campaign, fill in the fields in the description and move the item to “On-going” list. Don’t forget to set a due date.
When the item is due, move it to the “Review now” list, analyze the test result and fill in the rest fields in description.
Finally, decide if it worked or not, and move it to Finished list.
Learning is everything
In the startup world, it’s fine to make uncertain decisions. You just need to figure out each decision works or not as soon as possible, learn the most from it, so you get a chance to turn the wheel to a better direction.
I hope this practice we use in Kiwiship can help you track your marketing effort easier, and eventually grow your business.