Our Technology department consists of Developers, Data Scientists, and Physicists, all working closely with our Digital Specialists who are executing the marketing campaigns together with our clients. This creates a whole new ball game for automisation and optimisation of Shopping campaigns, leveraging algorithmic approaches including machine learning. The outcome is dramatically improved results from our Shopping campaigns
Our developers and savvy product experts enable us to manipulate product feeds in refined manners, ensuring that our clients’ products are exposed for as many relevant queries as possible, providing our algorithms with the data they need and ultimately our clients with the profitability they seek (and more).
No matter how smart an algorithm is, it’ll only work towards solving the task you have defined. It needs the best possible KPI to solve for, which is precisely why we’re moving our retail clients from ROAS targets to profit targets (using Google Analytics and the measurement protocol to get actual profit data) as well as CLV using, e.g., Google Analytics 360 and BigQuery.
Technology and automisation are in Precis’ DNA, and we continually strive to raise the bar for what’s possible within digital marketing.
Precis’ campaign structure leverages automation, group search queries based on intent, and take control of which products get triggered. Precis’ tool is data-driven, using historical data to group search queries based on thresholds tailored to the individual client, and it dynamically promotes or demotes a search query if its behaviour changes, e.g., due to seasonality or other changes in demand.
Audience targeting isolates different groups of people with a particular behavior for improved bidding. RLSA lists are mutually exclusive and collectively exhaustive to ensure we are covering all previous visitors without any overlapping. The entire customer journey is taken into account, and bid adjustments are based on their conversion probability.
The automated campaign setup and RLSA hierarchy is the foundation for Precis’ proprietary bidding solution, which is built on predictive modelling and semantic clustering. The algorithm runs statistical modelling of historical data to maximise campaign results. Our portfolio optimisation derives the value from the next unit of spend for each product and changes the bids automatically. This way, each product is optimised to the optimal bid, with the lowest possible opportunity cost.
Feed manipulation can ensure a best-in-class product feed and improve Product Group structure. DataFeedWatch is our first choice when it comes to manipulating the feed, where we optimise the attributes available on product level as well as product titles and much, much more. Further, it allows us to treat products differently, e.g., when a promotion results in higher conversion rates.
Christopher Brixen & Christian Ejlertsen