The Growth Letter: Issue 003
“Science, my boy, is made up of mistakes, but they are mistakes which it is useful to make, because they lead little by little to the truth.”
― Jules Verne, Journey to the Center of the Earth
Scientific Marketing
How’d you like a simple, 6-step framework for testing... well, anything?
Today I’m going to share the same steps I follow with any experiment I run, marketing or otherwise. It’s how I’ve developed the confidence to take our campaigns from thousands of dollars of adspend up to tens of thousands of adspend.
You can apply this process to any experiment you’re running. It’s nothing new. I didn’t invent this. I just apply it to marketing and advertising.
But to follow it? You need data.
This means you need to know how to get your data and how accurate it is. If you don’t have accurate data, this won’t work with 100% certainty.
When you don’t have data, follow the steps as best you can and intuit the results. Its better to follow some kind of process at all, than not to be testing.
You can still follow the process, but you’ll be shooting in the dark with your conclusions.
Alright, these are the steps in Scientific Marketing
1. Observation or Question
The basic starting point. Start with a question like, “What would happen if…”
You can also focus on a goal like, “Increase conversions from 1% to 2%.” Or, you could base it off your previous experiment and iterate on those results, “Based on the results from our last experiment, what would happen if…”
Once you have your question or observation, do some research.
2. Research Topic
Dive into everything you know about this particular question.
What have you tried in the past?
What might previous data have to say about your question?
What are other people doing that seems to be working?
Can you find proof from someone else’s work that this might work?
3. Hypothesis
Use your question and your research to develop a hypothesis. Sometimes these first three steps happen all at once, and sometimes you need to do more research.
The basic idea is to come up with a sentence like this, “If we do X, we believe Y will happen, because Z reasoning.”
Z comes from your research in Step 2 and your previous experiment observations.
A simple hypothesis I’ve used is, “If I change the headline on the sales page, I believe we’ll get more conversions, because the #1 benefit is more clear to the reader as soon as they reach the sales page.”
It’s tempting to skip the reasoning part. Don’t fall into that trap. Document and record what you want to do, what you think is going to happen, and why. If you don’t document your reasoning, you’ll likely forget or be unable to effectively report your conclusions later.
4. Test and Experiment
Now you implement your test. Change the headline, swap out the images, change the button color.
You’ll want to set a specific timeframe to review results. Base this timeframe on the quality and quantity of data you receive back. You need statistical relevance.
For example if you only have 10 people hitting our page, you’re likely not going to get statistical relevance. Make sure you have enough data coming back to you that you feel confident the results wouldn’t change.
Another caveat often overlooked:
Only change one variable, leave the rest the same. You want to isolate the change so you can identify what created the result. If you change more than one thing you won’t know what created the result.
One of your goals with following this process is to develop your own set of rules and principles to follow in your business. You do that by isolating the changes you make that improve your results. If you change more than one thing, you’ll never know what created the improvement.
5. Analyze
Once your experiment is complete, gather data and analyze.
What happened? Be specific. “Conversion rate went up by 14.57% during the experiment.” “Sales went down by 10.39%”
Was your hypothesis correct? Why? Was it incorrect? Why?
If the experiment does win, test the same thing again in a different situation and see if you achieve the same result. If you do, you’ve likely discovered your own “rule”.
Also, remember to take into account any outside factors you think may have swayed the results.
6. Report Conclusions
The final step is to gather data and results and share them with the team.
This is likely the most important part of the entire process, yet it’s often overlooked. By gathering, analyzing and reporting on your results, you force yourself to absorb the insights from the experiment.
It can be as simple as sending a slack message to your team, or, if you’re a solopreneur writing the insight down in a testing journal you can review later.
Where to Follow the 6 Steps
Best places to follow this method?
- Everywhere
- Ad creative
- Ad copy
- Targeting
- Landing page hero section
- Landing page copy
- Email copy
- Social post copy
- Student support responses
Remember, constant experimentation leads to constant improvement. Always be testing.
Let’s Look at Data
We recently concluded our Black Friday sale.
We sold $121,088 over the weekend. Not the best sale we’ve ever done, but nothing to turn up your nose at.
In my opinion, the biggest win from the sale wasn’t the high-ticket sales from the big bundles. It was the massive increase in customers for the 30-Day Challenge. We discounted it by $20 for the entire week.
During that time, it pulled in 482 customers and a total of $16,173. For comparison, the week before it pulled 90 customers total and $4,980 in revenue.
Of those 482 customers, 268 were brand-new leads. Meaning, they didn’t exist in our database beforehand.
Image of the Week
It can be difficult to know what to test... so here's the general guideline I follow:
Test the biggest things first. Things you think will have the biggest impact. In general that's:
- Anything above the fold
- The lead
- The call to action
- The offer
Make big changes, don't just change the headline and say the same thing. Change the headline and try a completely different angle.