Skip to main content
All CollectionsFeed Octopus Introduction
Introduction to Data Octopus
Introduction to Data Octopus
R
Written by Robert Stolarczyk
Updated over a week ago

Hi!

My name is Robert Stolarczyk and I am the Co-founder and CEO of Data Octopus.

Here I would like to introduce ourselves to you and help you start our cooperation.

I believe that in this way I will increase the chance of effectively understanding how we operate and addressing your business needs.

Let's go!


Who are we and what do we do?

Data Octopus is Product Data Management platform.

What exactly does this mean?

  • Product because we focus on product data, its aggregation, processing and activation in order to improve e-commerce sales. It sounds mysterious? Don't worry, everything will become clear soon.

  • Data because we focus on data. Our goal is to build on them and use them.

  • Management because we manage data, on the one hand, bringing it closer to each other, and on the other hand, processing it.

Is this the same as PIM (Product Information Management)?

NO.

PIM are systems that focus on working with product data mainly for the e-commerce engine. We focus on product data for output channels such as Google Merchant Center, Meta Ads, marketing automation systems, price comparison websites, marketplaces, etc. For example, a product description prepared within PIM most often goes to the e-commerce engine and product feed , which it generates. Data Octopus treats such a file as input from which it starts its work.

So what do we do?

Simply put, we use product data to improve e-commerce sales.

How? You will learn this later.


Where did we come from?

We will not be original because the idea for Data Octopus was created by real needs that we, as digital marketers on behalf of the agency, saw among our clients and on the market.

We noticed:

  • Growing amount of data

  • Problems with data being scattered in many places and in many formats

  • Inconsistency of systems when it comes to data standards

  • Development of advertising systems in an increasingly strong direction towards "black box"

  • Advertising systems have a growing appetite for good input data

  • The quality rather than the quantity of data is becoming more and more important

  • Lack of multi-channel analytics at the product level, which makes it difficult to decide which product to promote where

  • One tool to handle all feeds on all markets

  • GA4 limitations due to not connecting it to Google Big Query (short data access, cardinality, sampling, narrow reporting)

  • It is not possible to add information from Google Analytics, ERP/CRM system, warehouse system and others to the feed

  • Limited IT competences

We decided to use our knowledge, experience, access to the market and data to create a product that will address most of these trends and challenges. This is how Data Octopus was created.


Why do we identify with the octopus?

Did you know that the octopus is one of the most intelligent animals in the world? We identify with it not only because we also care about intelligent data processing and connecting the dots. The octopus has 8 arms and in fact its shape fits the Data Octopus data flow pattern perfectly. Sounds complicated? See for yourself.

Input data is all data that constitutes input to the system. They may come from sources and formats supported by the system, which you can read more about here.

The output data is all data that is already processed by Data Octopus at the BigQuery level using additional components such as Vertex AI.


What problems do we address?

Below are the main e-commerce problems that we address in cooperation with their digital departments:


Why are the problems we address important?

  1. Advertising systems are becoming more and more intelligent and automatic, but to work effectively they need good data, which they must receive in the form of a feed. No super car can go fast without super fuel. No AI will work well without a good prompt.

  2. Product data should come from other sources, such as a warehouse system or an e-commerce engine only 15% of marketers use more than 1 source of product data to create feeds.

  3. Ultimately, what matters for e-commerce is profit. Only 2.5% of marketers who segment products in their offer use the margin context to optimize sales not only in terms of COS/ROAS but also profit.

  4. More and more data and sources generating it will force a modern cloud-based data storage architecture. Already 78% of companies in the US have implemented cloud solutions in their organizations. In Poland it is still 38% of companiesm.

  5. 3rd party date restrictions resulting from min. the growing importance of data privacy, cookie restrictions and the increasing importance of 1st party data in the personalization, targeting and optimization of advertising require the construction of a product data repository.


What do we believe in? What is our mission?


For which segment of the e-commerce market is Data Octopus the most valuable?

The value increases with the scale of several factors:

  • SKU

  • revenues

  • advertising expenses

  • sales breadth (number of advertising channels, markets)

  • the number of systems that retrieve product information

  • the importance of Google BigQuery in the e-commerce data ecosystem

This makes Data Octopus best for the mid-to-high end e-commerce segment.


How do we use product data to improve e-commerce sales?

Until now, segmenting products using knowledge about margin or inventory required involvement and internal IT development. This involved longer time, greater costs, resources, complexity and often limited effectiveness.

The result of this process was the traditional arrangement of products in the advertising account, most often depending on the category tree available in the store. Men's shoes were included in the men's shoes campaign, they often had a similar optimizer and goals as products from another category, and their performance was assessed based on the same metrics.

With Data Octopus the process is different. We start by reducing the need for IT and Data Analyst resources. Much of the necessary data work can be done without additional IT infrastructure at the Data Octopus level, mainly in the section Data&Feeds, and the reporting environment in the Reports section that answers a number of questions that, combined with product segments, reduce the need to engage a Data Analyst working with product data.

As a result, products are placed in campaigns that are best suited to them based on the data. Additionally, they are "trained" for specific goals, such as increasing the number of views or conversions. Not every product has the same goals, optimizer, budget strategy. Not every product is assessed on the basis of ROAS set at the same level.

As a result, the system supports better matching of the strategy to what the product represents in the database, increasing the chance of effectively activating the sale of a larger part of the SKU. What is very important, however, is that by using the margin context, we increase the chance of managing and increasing profit, even if revenues and profit remain unchanged.


What does data flow look like and what are the benefits?

Below is an illustrative data flow diagram.

Data Octopus has 3 important functions:

  1. Product segmentation based on data from the material model.

  2. Selection of products before they go to specific outputs based on the database.

  3. Reporting based on data stored in Google BigQuery.


What is a Master model and what does its diagram look like?

Master Model of product data, consists of 40-50 columns of data. It contains information from the product feed, but also data coming directly from Google Analytics 4, Google Ads, the e-commerce engine and from additional sheets sent by you, imports with additional data.

Master model includes:

  • data from the product feed - basic information about products

  • data from GA4 - knowledge of how products engage and convert

  • data from GAds - knowledge of the effectiveness of product sales

  • data from other sources - knowledge of the margin on the product and the final profit on its sale

Below is an illustrative diagram of the data structure of the master model, which is created at the Google BigQuery level and then processed by Data Octopus.


Stories of cooperation with selected clients.

Get inspired by some of our past success stories. You must know that for this to happen, cooperation between e-commerce digital marketing resources and the Data Octopus tool and team is necessary. Even good input data will not make a difference on its own if it is not properly directed by a human.

  • Intersport

  • 4fizjo

  • Unitrailer

  • Butosklep

Did this answer your question?