Sep 16, 2021

How Party City is Using Increasingly's AI and Machine Learning to Drive Revenue

Bethany Cowan

At its core, artificial intelligence is a term used to describe computer systems that are actually able to perform the types of jobs that would normally require human intelligence - all without the intervention of actual people in any way, shape or form. Examples of this include but are certainly not limited to speech recognition, language translation and even decision-making in some situations.

Machine learning is the natural extension of this idea, and it describes a situation wherein those same computer systems can "learn" to do their jobs better over time - all without being specifically programmed to do so.

Neither of these concepts are anything new, necessarily - indeed, they've been around for years. But the modern day applications of them are nothing if not impressive - as Party City has recently gone a long way towards proving.

Party City partnered with AI technology platform Increasingly to encourage customers to buy more. Increasingly's technology displays hyper-relevant product combinations via a 'Build your own bundle' module on product pages and at various points across the customer journey. The bundles are automatically populated via Increasingly's AI platform analyzing past purchase behavior patterns and customer intent data. This provides an unparalleled shopper convenience, helping Party City customers find everything they need for their special occasions. Check out the results below.

PartyCity + Increasingly's AI and Machine Learning: A Match Made in Heaven

Based out of New York, Party City is the largest retailer of party goods in North America. Originally founded in 1986, it currently has over 900'locations'all over the world and employs more than 10,000 people by most estimates.

Although the technology behind the Party City AI platform is admittedly quite advanced, the concept at the heart of it all couldn't be more straightforward. It begins by analyzing both the purchase history and behavior of real customers - all in an effort to learn as much about them as possible.

Using Real Data to Determine Likelihood of a Sale

At that point, based on what someone is viewing on the website at the time, the system can detect their purchase probability - all by using real data to determine whether or not someone is just browsing or whether they're likely to actually make a sale.

If the determination has been made that a sale is likely, the system can then identify user level product combinations that make the most sense given user actions at the time. In other words, based on exactly what you're looking at and how likely you are to buy, it can identify cross-sell opportunities in real-time - all without needing to involve a human employee in any way.

A More Personal, Relevant Experience

Truly, this is a win-win situation in every sense of the term. Users of the Party City website get the more personal, relevant experience that they've been clamoring for. Party City gets a sure-fire way to separate themselves from their competitors. But more than that, the organization also gets a massive boost in revenue. They've indicated that this approach has already netted them a $6.5 million incremental revenue boost, amounting to about 8% of the total site revenue. It's also generated a 1.8x increase in items per order - a perfect storm if there ever was one.

More than anything, Party City's application of Increasingly's artificial intelligence and machine learning has been a way to generate more insights, more opportunity and - most critically - more sales during a time when they're desperately needed. That last point in and of itself may be the most important benefit of all.