We finally have an alpha. Even if it is held together by string and scotch tape, it’s really exciting for us. Up to now, we’ve been focused on learning consumer behavior, testing hypotheses, and focusing our ideas. Now we are out to get confirmation (or lack thereof) based on actual usage of a minimal viable product.
So what are we doing?
First let me tell you what we’ve learned. We have spoken to over 150 people of all sorts, but with some basic commonalities: over 21; mostly college educated; likes to shop online. More women than men. These are people who feel busy. They like deals, they want to save money, but interestingly enough, they don’t use “savings” sites like Nextag, Retailmenot or even Google/Bing. They’ve tried them, but were turned off by the sheer volume of information, friction, trust issues and more.
Our hypothesis is that less is more with these consumers — that it’s not about exposing 5 billion deals to them, but rather exposing the 5 or 10 that are relevant to them this day or this week. Search is not an every-time solution for these consumers. We want to filter out the noise, and be the ecommerce email / website / mobile app they are waiting to see every week. The (evolving) elevator pitch goes something like this:
Our mission is to painlessly provide personalized, extremely relevant recommendations that save you money when shopping online.
A few examples:
– a mom buys size 4 Pampers at store X; we point her to a lower price on the same item at store Y
– a cycling enthusiast is looking for a bike seat; we notify him when one goes on sale at a specialty store
– a busy professional can’t keep up with marketing emails; we show her the three most relevant promotions and markdowns this week
Our first version isn’t quite this pretty yet, but the below mockup gives you a sense of what we’re talking about. Obviously the items suggested would be different for every user.
Personalized savings / suggestions isn’t a radically new concept, but in past it has not worked very well. Previous approaches required the user to do too much work, and even then the underlying data was not strong enough. Personalization has been more of a marketing ploy than a reality. Amazon believes that the best indicator of interest is item-level purchase history. Our approach is to get item-level data from the email receipts stored in each of our inboxes, and we have multiple ways for you to share those email receipts with us. But it’s not just about receipts — we also believe that we can be even more effective by combining several other interesting data sets.
Building out the full system will be a fascinating technical challenge (thankfully I’m partnered with Liz Crawford, who, with her PhD from Carnegie Mellon, is clearly the brains of this outfit). However, right now our focus isn’t about charging ahead on some massive build, but rather testing our assumptions, even with a very crude product.
Obviously it’s a very big ecosystem in which we’re playing. There is some interesting competition out there, although we’re all taking different approaches. We know there’s a problem in the market and it’s a matter of finding the right way to crack it.
I’ve tried to keep this post from sounding like a marketing pitch, but here comes the “ask”:
If you know someone who loves to shop online, and who might be interested in being in our beta program, please let them know about us. Have them reach out to me (giff@ aprizi dotcom) or sign up on www.aprizi.com.
We’re starting small but will be letting people in as quickly as we can. And again, our mission isn’t to convince people of the brilliance of our concepts, but rather to learn and iterate our way to a brilliant product that speaks for itself. If you like to shop online, come help us learn how to help you!