• by Laurie Sullivan  @lauriesullivan, February 22, 2017

Crealytics engineers are working to buildout a feed-based advertising platform to not only support Google and Bing's image-based shopping networks, but support marketplaces such as Amazon and eBay, Facebook, Instagram, and other comparison-priced platform.

"The idea is to create one platform to handle all," said Andreas Reiffen, CEO at Crealytics. "Marketers live by the illusion they can track and measure everything, but the industry isn't even close to solving all the challenges.

How are sales driven by advertising, from search to programmatic, valued and measured? Some companies track revenue and attach margins. It's common that 50% of a retailer's stock level runs low for products that sell well. The other 50% are in excess, products that typically need to sell at a "massively" low rate of return.

The cost of selling the overstocked product remains much higher, and it is not reflected in the reporting results of the Web analytic platforms, Reiffen said. "No one wants to talk about it but it remains a major problem," he said. "I don't want to say that we can solve this, but we are one in a few companies trying to take this into account."

Advertising campaigns should connect to stock levels, price, performance and the ability to stop and start bidding in real-time. The technology should run more than ad campaigns, and that requires better integration with the company's business technology.

These platforms include a retailer's out-of-stock technology and enterprise resource planning (ERP) tools, inventory, and customer retention rates. It puts technology at the crossroads between what makes the business run and how they use advertising to reach consumers.

ERP systems will soon integrate more closely with ad-serving platforms to programmatically identify when to serve an advertisement or allocate that media budget to another consumer based on CRM and inventory data.

Companies have tried to turn Reiffen's vision into a reality. In 2015, Vector Media Group introduced a programmatic system that links programmatic search advertising to product feeds by connecting a retailer's or brand's ERP and inventory control supply chain to automate the cycle. It is intended to adjust search campaigns automatically to the hundreds of thousands of products coming in and out of inventory. 

At the base of Reiffen's vision sits Crealytics technology that began with text ads and now Shopping Ads. It analyzes millions of search queries and breaks down each word into buckets based on their semantic meaning. If someone searches for "inexpensive Adidas women's shoes" the algorithms know "women's" equals gender. It also identifies "shoes" as a product category, "inexpensive" as a discount word, and "Adidas" as a brand. The company uses the same technology to offer targeting in Google Shopping.

If someone is browsing and typing in shoes, the technology can identify the action and reallocate the media buy to traffic with the higher propensity to convert into a purchase. Since it's not possible in Google AdWords, engineers found a workaround to slightly "violate the system" to bid on the specific search query.

The platform called Camato isn't actually designed for this, he said, but there's a workaround for Google and Bing. The idea is to boost performance for Shopping Ads.