How pick.gifts Works
Don't stress to find the perfect gift – our Intelligent Holiday Gift Finder will save you time and money. All you have to do is to answer a few questions about the person for whom you wish to find a gift, and we will give you the best matching Holiday present ideas based on what would make that person happy. We do this with the help of Artificial Intelligence and a bit of Santa's elf magic.
How do I know this will work?
pick.gifts is powered by Spark2, a set of our proprietary Machine Learning algorithms that produce gift recommendations based on multiple data points. Besides taking into account the information you provide, our system analyzes and constantly learns from the historical data of other users.
multiple data points
The magic begins the moment you start answering our fun quiz questions. Our system uses your answers to filter out the least matching products and then compares your data with that provided by other users who have bought similar items. Whether you are shopping for your significant other, colleague, family member, or best friend, you can be certain that our system will recommend the best gifts for everyone on your list.
Is it really free?
Yes! Our service comes completely free of charge and you can use it as many times as you like. In fact, the more people use pick.gifts, the better it gets. Each time someone uses our service, our Machine Learning-powered system teaches itself and gets even smarter in finding great gifts.
We only earn a little commission from sellers on a limited scope of products which our algorithm may or may not choose to recommend. However, it will never suggest something that doesn't fit the profile of the person for whom you wish to find the perfect gift.
What happens with my data?
We only collect impersonal data and every answer you provide is never attributed to you or any other specific person. The anonymous data that we journal is used solely by our algorithm for the purpose of improving the service we provide. We never share data with third parties, nor do we share it between individual users.
If you would like to dive deeper still…
Imagine a guy named Billy. Billy wants to buy a gift for his wife. "Something nice," says Billy. "Maybe a purse, or a pair of shoes." He knows nothing about purses but he thinks he has a rough idea of what kind of shoes his wife would love. "Something classy with high heels," according to Billy. Yet he's totally clueless about this season's women apparel trends or the best brands to pick from. He browses over dozens of online shops, reads tons of reviews and even watches a dozen YouTube videos only to get completely underwhelmed as he learns that a pair of high heels might cost anything between $50 and $1550.
Billy's problem is quite predictable: as high heels have different brands, manufacturers, dozens of materials, their cost depends on trends, seasonal demand spikes, and who knows how many more hidden factors unknown to our dear Billy. Naturally, he can't into account all the new information to consider the best option.
Puzzled and beyond any hope Billy asks for a piece of advice from his colleague Jane. Unfortunately, Billy is clueless about the fact Jane is a sneaker person, and that her taste differs completely from that of Billy's wife's. So the advice Billy gets results in a very different outcome than he expects–although his wife is moved by Billy's good intentions, she only wears the shoes once, only not to upset her caring husband.
Had Billy known of pick.gifts, this scenario would have played very differently. Billy wouldn't have needed to go through all the pains only to waste $300 on a pair that his wife will never wear. Based on the information Billy has provided, our system would have matched his wife's profile to the historical data of similar persons to come up with a thoughtful gift that would have made his wife excited.
People are only capable of taking into account a limited amount of data. For huge amounts of information, we need algorithms to do the maths for us so we could find all the hidden patterns.
Gift estimation can be treated just like any other Machine Learning problem. In our case, the goal of this problem is to get the optimal output (a personalized gift idea) given some inputs (your answers) that can be interpreted by algorithms. The key aspect of Machine Learning algorithms is that they are “trained” with real-world data and get better at their job as the amount of data grows throughout time.
To come up with the perfect gift recommendations, we use an ensemble of collaborative filtering (such as matrix factorization, nearest neighbors) algorithms, which work on the user interactions (such as product views or purchases), and content-based algorithms, which work on the gift item properties (such as prices, brand names, descriptions, user reviews, purchase trends and many more).
pick.gifts team would like to thank you for using our platform. We hope our service will make your Holiday gift shopping a little easier and much more fun!