The Beauty of PureClarity

PureClarity’s power stems from the flexibility of allowing you to choose to let the AI run all of your personalization, and offering you the option of manually enriching if you choose creating Personalized Campaigns to layer onto the automated AI for a truly personalized experience.

A truly personal experience

PureClarity is built on AI (Artificial Intelligence) and Big Data technology. It collects data about every one of your visitors, tracking their interactions with you onsite and offsite to build up a picture of their behavior, their dislikes, likes, interests and habits.

This information is then fed through Machine Learning algorithms in order to build a picture of customers behaviours. Data that is collected can be explored and drilled into. You can time slice and segment the information to get a deeper understanding of your visitors’ behavior and the performance of your site.

As visitors arrive on your site, PureClarity will change the recommended products and strategies based on their behavior during that very visit dynamically in real-time. PureClarity will show recommendations that it knows have worked on other visitors to maximize the chances of the current visitor seeing relevant results and buying on that visit.

PureClarity’s AI is as unique and
as personal as your fingerprint

With PureClarity, website visitors and customers experience a completely personalized and optimized recommendation strategy that changes in real-time as PureClarity collects more information about them, always optimized to maximize online conversion and average order value. One of the reasons PureClarity’s AI is successful is the fact that Big Data is so easy to process with PureClarity’s suite of intelligent algorithms.

Hybrid Algorithms

PureClarity AI uses a full suite of hybrid algorithms including collaborative filtering, pattern mining and Machine Learning algorithms to determine which recommender to present based on any dataset. This means that the large amounts of data are analyzed to find the optimal cross-sells and upsells based on what your customers have looked at, are looking at, previously bought, looking to buy right now and predicted to buy next. Recommendations are highly personalized, change in real time and effective as the customer’s journey progresses.

AI Recommender Engine

PureClarity’s AI Recommender Engine will never show two of the same recommender strategies on the same page and will work to minimize product duplication. Visitors can enjoy relevant, specific product recommendations without seeing any of the same content, guaranteed to increase your average order value and conversion rate.

Conditional Probability Algorithms

Within PureClarity there are also algorithms that work on Conditional Probability which ensures that recommenders do not include products which are bought on an extremely regular and frequent basis. An example of this is seen with PureClarity’s client The Royal British Legion. Purchases on the RBL online shop almost always includes a Poppy Pin Brooch. Given that these items will almost always be included in the cart, PureClarity’s algorithm ensures that Pin Brooches are not boosted as much in Personalized Recommenders so that more meaningful relationships between products can be displayed.

Automated A/B Testing

PureClarity’s AI engine continually runs automated A/B testing in the background, testing which strategy works best. If your customers respond best to cross-sells, PureClarity will favor those algorithms. If they prefer strategies which show more of the products that are popular on your site in real-time, PureClarity leans more towards that. Every recommender is optimized to be right for the customer, the page it’s on, and the overall trends on the site.

Sparse Data Analysis

PureClarity offers sparse data analysis enabling the AI to implement a content based filtering recommendation system which finds products with similar product properties within the core data set. This algorithm is highly effective to personalize recommendations for businesses with low customer data volumes.

An example of this is with JewelStreet – a high end Jeweller who came to PureClarity looking for a personalization solution.  JewelStreet have low data sets as their products are high end and purchases tend to one off or infrequent.  PureClarity’s content-based filtering recommendation algorithm works very effectively helping them to increased their total order value by 83% and tripled their average conversion rate in the last 6 months.  If a visitor looks at an emerald necklace on their store the AI can retrieve and recommend other products based on the title attributes in this case ‘emerald’ first as a high scoring unique term and necklace second which has a lower score due to the generic nature of the product title within the store.