The order can bring in the money

Order in your online shop can make a difference. What the customer sees first matters. In a US test in 2010, SPAMfighter generated 25 per cent more revenue by changing which subscription customers were shown first.

Order in your online shop: What’s the first choice?

My colleague had tested one of our products, where you could choose a one-, two- or three-year subscription. Until then, one year was the first choice, while further down you could see the discount for buying more years.

Not many customers bought that product again the following year. Perhaps they’d cleaned up their PC or changed machines. So we were interested in whether more would choose a longer period from the start.

Fewer purchases, but higher total revenue

In the test, the three-year subscription was shown as the first choice. That meant customers saw a higher price first than before.

As expected, fewer people ended up buying. In other words, the conversion rate fell. But among those who bought, more chose three years, and total revenue in the test rose by 25 per cent.

The result came from the specific US test I discussed in December 2010. It’s not a guarantee of the same result in other online shops or markets.

Test both the price and what the customer sees first

We wanted to keep working with the figures and investigate whether the prices and differences between the choices could be improved further. That was something we wanted to test, not a result we’d already achieved.

My point is that you can experiment with what the customer is offered first. Perhaps a bundle of several products with a relevant discount makes sense. But look at the business as a whole, not just how many people click buy. Also read E-pusher about my experiences of online selling.

Also read how the order of fields can help first-time buyers in your online shop.

Christmas music in the car and a test from the US

I start with a little Christmas music on the stereo. There’s not a damn bit of Christmas atmosphere in Florida with the palm trees and everything, but I find the music cosy. At least it’s cold. Along the way, the music gets a bit much, so I change stations.

My colleague had been testing various things on one of SPAMfighter’s products. It was a very targeted test here in the US. I say from the start that it may not apply to everyone, although I personally think the idea could be used more widely.

From a download to a subscription

The product was SLOW-PCfighter. You could buy a one-, two- or three-year subscription. One year cost around 20 dollars, and I specifically mention 19.95 dollars as the previous first choice. Customers got a discount for two or three years, with three years costing around 49.95 dollars.

The process was that a customer saw an advert or something similar and downloaded the product. They were then offered the chance to buy the full version and get the machine properly cleaned up. Clicking the button took them to the purchase page.

By default, the page showed one year at 19.95 dollars. Further down was the option to get a discount by buying more years.

We wanted customers to buy more years from the start

Not many customers bought that product again the following year. Perhaps they’d fixed their PC or bought a new machine. Retention from year to year wasn’t particularly good.

So we wanted the customer to buy as long a period as possible from the beginning. That’s why my colleague reversed the choices.

Now the page defaulted to three years at 49.95 dollars. We were curious about that because there’s a big difference between being greeted by 19.95 and 49.95 dollars.

The lower conversion rate was only half the story

As expected, the conversion rate fell. Fewer people who reached the purchase page ended up buying.

What was interesting was what the remaining buyers chose. Many more bought the three-year option, and total revenue rose by 25 per cent in the US test I’m talking about.

Some may have seen 49.95 dollars and thought no without noticing the one-year option at 19.95 further down. Even so, the change generated more revenue overall. That’s why you can’t judge the result solely by whether more or fewer people click buy.

The next experiment would focus on the price difference

We wanted to work more with the figures and try to improve things further. I mention raising the one-year price from 19.95 to 29.95 dollars while keeping three years at 49.95.

My thinking was that the smaller price difference might get even more people to choose the longer period. It was something I wanted to try, not a result we’d already achieved.

I say in the video that prices aren’t, to a large extent, just a matter of costs. They also need to be optimised for customers so we get the most out of the business.

The idea doesn’t have to be just about software

I suggest experimenting with the first offer the customer sees. Where relevant, for example, you could suggest three items with a good discount instead of one.

It can be something other than software. But I also end with an open question: Will people accept it? You have to try to find out.

I was excited about the 25 per cent in the US. If we could make the same thing work in Europe and the rest of the business, I imagined an annual revenue increase in the millions of kroner. That was an expectation of a possible rollout, not a confirmed result outside the test.

So switching the one- and three-year options could turn out to matter a lot. I end by wishing everyone a merry Christmas, even though I hope we’ll meet again before then.

Best wishes, Martin