At Shop It To Me, we are constantly experimenting. It’s one of our core values. In fact, as of this morning, we have 24 different A/B experiments running on different parts of our site. Each week we probably add anywhere from 1-5 new experiments to our testing pool (and take out 1-5 old ones). We experiment not only with small features but big ones too. At any given time have 1-2 completely new products that we are testing with a subset of our users.
But unlike a lot of companies who announce any new feature or product, we don’t announce even our biggest products when we first put them into the wild. Even though we have press asking for new things and we could get a great “usage spike” we actually tell our PR team not to talk about them until the time is right.
The problem is, until you have nailed the product, a press “spike” is just a spike.
Now don’t get me wrong, I’m not a big fan of the “stealth company” technique — companies that promote that they are doing something amazing but won’t tell you what they are doing. I think these companies actually set such high expectations it is really hard for the user to beat them.
But I am a fan of not announcing a new product until we are pretty sure it is a hit. Why? I am a big believer in the disproportionate power of first impressions. Humans are hard-wired as a species to take a tiny bit of data and extrapolate big decisions/impressions from it. It’s leftover from when our ancestors had to quickly decide if a creature is predator or prey, and it is still in almost all of our decision making.
We make significant long-term, big decisions based on only on a first impression all the time. Have a terrible first date and there’s a pretty high chance you will give up before going on a second one (even though even your “soul mate” and you are likely going to have bad days together). Try out a new restaurant and have a bad service experience and you likely won’t ever return — even if they later get great service reviews. And in business, start off a presentation poorly and much of your audience likely stopped listening (and will rate it poorly) even if remaining 80% is amazing.
Remember that before spending loads of money announcing your next product. If someone tries your product and just likes it (or thinks it is mediocre), you have likely given them the permanent impression that your product is just OK. They won’t return. On the other hand, get a user to have an “I LOVE THIS!” first impression, they often will still have positive impression years later. We’ve seen this time and again — our best users are the ones who found something they wanted to buy in their first emails.
So before we turn on the PR machine for a new release and send it to all of our users, we make sure that almost anyone who sees the product is going to have a great first impression — such a good impression that they want to keep using it (and hopefully tell their friends). How? We spend months usability testing the heck out of it — constantly tweaking it with user-testers and thousands of random users and friends and family-members and even random people we recruit off the street (yes we’ve done that). And we learn from them and make the changes that gets their reactions from “Meh” to “Like” to “Love” (a subject of a future post).
So while we may lose some “brand-awareness” and short term traffic from multiple product releases, we more than make up for it by having a product that for most people feels “pretty awesome” right from the start.