As we wrap up a truly eventful 2022, I reflect upon our very complex multi-product SaaS platform that continues to evolve here at Particular Audience. What sets us apart from our competitors is not only our superior product offering (some of which is completely unique) and white glove services, but the cross product network we’re building within PA. Rather than just providing technology to our partners that services a single need, we’re focused on enabling a cross pollination of our products to also function as an intelligent way to serve sponsored placement across websites. Through our multi-product solution, PA is helping our partners optimise their website with technology to drive top line sales as well as enable an incremental ad revenue stream, propping up their bottom line.
For those who have followed PA since our early days, James set out with an intention to solve one of the greater challenges of the ecommerce industry, leveraging data to democratise big-tech’s advantages for the rest of retail. In doing so, PA took a privacy first approach, enabling retailers to personalise every customer journey on their websites, giving the industry one of the critical tools it needed to take on Amazon. This journey began with AI Recommendations. Particular Audience has come a long way since those early days, now a complex multi-product ecosystem equipping our clients with a sophisticated set of tools to not only remain competitive in today’s environment but actually get ahead.
2022 was a year of many exciting developments across our product ecosystem which I’d like to reflect upon with our community today.
What's your Vector, Victor?
Across the web we create relevance on our clients’ websites in product discovery through Recommendations and Search. Both products take a hyperdimensional approach, using both anonymised behavioural data as well as vectorised item data to accurately predict customer preferences. This year, vector embeddings have paved the way forward in our machine learning technologies.
For Recommendations, our vector approach allows us to further solve the problems of cold start items, lack of historical sales or small data sets, as well as selling more from a catalogue’s longtail. For Search, building on top of existing Natural Language Processing (NLP) and our legacy vector based Visual Search (VS), we’re developing a new approach using large language models to ensure contextual meaning of customer search queries is understood in returning and ranking relevant results. As a simple example, dogs and animals are not synonymous with each other, however the human brain associates them with one another. Through neural search and vector embeddings we’re solving this problem that traditional keyword search cannot solve for.
We’ve built the most intuitive customer shopping experiences with AI Bundles, a product that automates groupings of codependent or complimentary items, making it simple for a customer to add all items to their baskets in a single click. Bundles help retailers maximise profitability of each transaction by assisting with incremental UPT (units per transaction) and increased AOV (average order value) through cross sell; another win against the big retailers who rely on bundles as a key product feature within their funnel. In Bundles, our new vector approach translates strings of purchase data into literal sentences, embedding product attributes which cannot only accurately predict groups of items a customer has previously purchased together, but also help us make accurate predictions on new item groupings that include items never purchased ever before. Neat, huh!
Transparent dynamic pricing is now achievable with Similar Inc
In 2022 we led the dynamic pricing and incentives market with our Price Beat application of Similar Inc, solving the problem of price competition for retailers by allowing them to compete without slashing sticker prices. I would argue that Similar Inc’s dynamic pricing application across our client partners is smarter and more sustainable than Amazon’s race to the bottom pricing strategy and we’re the only product on the market around the globe actually doing this today via a front-end customer facing UI. The UI is critical to this product as it builds customer confidence on price via a trusted independent third party.
Similar Inc’s web scale pricing data keeps customers on site by allowing them to price compare without leaving the page and dynamically matching or beating on price when required. Not only does this solve the price comparison problem for a customer, it's a very powerful conversion rate optimization (CRO) tool for the retailer.
For retailers this actually means they discount less. Average discounts range from 2-4% per item not per transaction,and we only recorded highs of 4% discounting over Black Friday weekend. Also, retailers only discount when they absolutely need to, to convert the price comparison shopper.
I’ve worked in the retail and ecommerce industry for a long time and I can say, the results are insane, more than doubling conversions of participating items; yes that is a 100% lift!
The impact on loyalty appears to be significant as well. Offering customers transparency and fairness builds trust which keeps them coming back.
Similar Inc is rapidly expanding across many new client partners. Our engineering team have further expanded our product pricing dataset as well as developed and introduced new integrations packages for faster, more scalable and stable integrations.
Retail Media is most certainly taking over
Released this year, PA Retail Media empowers retailers to monetize selected slots on their websites via sponsored products, leveraging our market leading recommendations stack to ensure each ad is relevant to the consumer. Retail Media drives high margin advertising revenue for retailers and is becoming a more effective and very measurable portion of suppliers performance marketing budget.
According to GroupM, Retail Media now represents 18% of all global digital advertising and 11% of all advertising. Forbes also says it is on track to make up 60% by 2027, even more bullish than previous BCG forecasting.
As many would know, Amazon has long offset their retail losses through the Amazon Advertising retail media business, last year alone posting US$ 31 billion of revenue from advertising.
Since our release in June, initial results across our clients and their supplier partners have been extremely promising. We’re seeing strong engagement in sponsored relevancy campaigns, significant lift in brand share of voice for suppliers, and ROAS that is outperforming other performance marketing channels by a factor of 5.
We’ve rolled out new features to remove friction and build retention in the onboarding process for suppliers. We’ve launched new ad sets that target keyword queries and we optimised reporting, ensuring suppliers can have real-time access to their impression, clicks, budget balances, and attribution to always have a pulse on measurement.
Let's Go 2023!
So today we’re still on that journey James set out to solve, helping the rest of retail take on Amazon. The way I see it, our PA toolbox is mandatory for our clients to get ahead in the competitive ecommerce market today.
As we kickoff 2023 I look forward to further expanding our cross-product pollination and sharing more updates soon.