I want to start a competition with my wife, but she won’t join in…
Last night was the science fair at my sons’ school. My wife and I are both nerds and geeks at heart, so you can bet our kids participate. (It’s only required in 5th grade.)
Because we want to foster their creativity, we’ve largely let them pick their own projects. They can do an experiment, a demonstration, or a display. This year, both picked displays (with a demonstration edge): How things dissolve, and How the earth was formed.
Both were interesting, and provided cool learning opportunities…
But both my wife and I would have loved for them to do more of an experiment. My wife, being the psychologist, has encouraged a social science experiment. Last night, as we were driving to the science fair, I got to thinking…
What if they did a scientific advertising experiment?
Now, if you’re a student of direct marketing, you know Scientific Advertising is practically gospel in our circles. Although he wasn’t the first to the party, it was Claude Hopkins that codified a scientific approach to advertising in his book by that name. (I recorded it as an audio book and give that away free here, you can also buy a copy on Amazon.)
Ever since then (and even more so now that online tech makes everything measurable and testable), a growing group of advertisers has relied on scientific experiments to make advertising decisions, instead of opinion.
We’ve discovered that you can run two ads simultaneously to segments of the same market, and one will generate more response than the other.
Sometimes the difference is tiny. Sometimes even a small change in an ad can lead to 10-times as many sales from mostly the same ad.
And here’s where this is relevant to the science fair last night: a good scientific advertiser will use the exact same experimental design principles as are required in any other scientific experiments.
In fact, if I had to venture a guess…
There’s probably more scientific experimentation and data collection going on every day in advertising than there is in all other fields of science combined!
I’m not saying this is how it should be. I don’t necessarily think that’s somehow morally wrong or right.
It’s just that there’s so much money on the line when it comes to advertising… And today’s tech has made it so easy to run one ad against another… That there’s a very high likelihood that at any given time the number of advertising experiments that are “live” in the world are probably 10-to-1 compared to other science.
So all of this is going through my head, when I start pitching to my kids that for next year’s science fair, they should consider running an advertising experiment.
They’d probably be the first kids in the history of the elementary school to do it…
It would be exciting because it’d expose people to a field of science that’s virtually unheard of in academia…
AND, maybe they could make a few bucks from running their test!
I told my wife my idea last night. She laughed at me. Then she said, “You better not get them to run an advertising experiment before I get them to run a psychology experiment!”
So I tried to throw down the gauntlet… And failed!
She (jokingly) said she didn’t want to compete against me, because I’d use my shifty marketer tactics to get them do to my experiment.
I laughed back, and admitted — that’s exactly what I’d do! (After all, very few psychology experiments actually lead to money in your pocket!)
But the more I think of it, the more I realize…
Understanding how to run business experiments is probably THE most valuable moneymaking secret!
Every business is experimental. But the most common experiment is testing SOMETHING against NOTHING.
We try running an offer, and if it works, we keep doing it.
We create a product or service, and if it sticks, we keep selling it.
We put out some kind of ad, and if it seems to bring customers in, we keep running it.
And so on…
But a few businesses are very scientific in their experimentation. Especially those in direct marketing.
We create a hypothesis about what will work. We test it. We measure results.
And here’s where it gets really exciting. Once we have something that works, we call that our “control” (note the language straight from scientific experimentation). Then, we create alternative test versions to run against the control.
In advertising, we split our market into randomized samples. We’re trying to create a generalized conclusion, based on a segment of the market.
We run the test and tally results from the multiple test versions, until we have statistical significance telling us which is the winner. That becomes our new control condition.
And like any other scientific discipline, scientific advertising has developed a body of best practices — principles that have proven themselves through experimentation, that can be reliably used to predict the results of future tests.
But also like science, we, the experimenters, are often surprised. We make assumptions about what will work, and are proven wrong by the data. Then we have to adjust our model to fit this new data.
I’ve described this in a way that all sounds very scientific, because I wanted to connect the dots.
Now, let me describe it in a way that will get your greed glands tingling…
You have an idea for a product. You find out you can source and fulfill it for $5 each if you buy 100 at a time. You decide to try selling the product for $10.
You plunk down $500. If you sell all 100, that’s $1,000 revenue.
Then, you start to run some experiments.
First, you manage to break even. You spend $500 on advertising, and sell the first 100.
Combine your advertising and fulfillment costs, and you spent $1,000 to get your first $1,000 in revenue.
But along the way, you did a little testing.
So you spend another $500 on products, but manage to sell all 100 with just $400 in advertising — giving you a net profit of $100.
Now, you roll that profit into more product, and buy 120 for $600.
More experiments give you better economics, and you sell all 120 for $300.
So now you made $1,200 in revenue off $900 expense — your net profit is $300.
You keep doing this, and eventually you’re able to spend $2 in advertising to sell a product that cost you $5 for $10.
Not only that, because you’ve found something that works well, you push your sales volume up to 10,000 units per month.
But then something interesting happens. Because of the volume, you’re able to shave $2 per product off your sourcing costs, and now you’re spending $2 on advertising and $3 on the product to turn around and sell it for $10.
And based on that volume, you’re making a cool $50,000 per month, and growing.
Now here’s where you say…
“But Roy, you made up all those numbers!”
Of course I did. But it was to prove a point. When you take a scientific approach to advertising, marketing, and selling your product… You can turn an unprofitable venture into a profitable one… And a modestly-profitable venture into something really big.
And the numbers just keep getting bigger, as long as you have the market to support it.
Because it’s most often the business with even a slight edge that has just what it takes to dominate their market. Usually, that slight edge doesn’t come via luck or happenstance, or even any major merits of the marketer. Usually, that slight edge comes through tests and experimentation to find little performance improvement after little performance improvement that eventually adds up to something big.
And of course, the proceeds of one success can fund more product creation, more offer experimentation, more tests, and more growth. Plus, as you build success on top of success, you’ll be building a customer base you can go back to at a lower (near-zero) cost to sell to again and again.
It’s all additive.
It all leads to more and more breakthroughs.
But it starts when you realize that advertising can be treated like a science, and run your first experiments.
Seems like that could be a cool science fair project, no?
Yours for bigger breakthroughs,
Roy Furr
