When Productsup was established in 2010, it was in the early days of Google Shopping and it barely had people actively interacting with it – whether advertisers or consumers.
As one of its co-founders had his own online shop listed in the Google Shopping system, one day he played around with the attributes in the feed. He discovered an instant traffic spike and after further review of the results, it was decided to launch Productsup as a digital agency that helped retail companies to get the most out of the channel.
For nearly 2.5 years Productsup ran as a specialized agency in Europe and optimized the product data of retailers for Google. Doing that, it had considerable success and sometimes increased traffic by nearly 900% in less than two weeks. In order to manage those amounts of products from its clients, the company built an internal software solution that would enable it to quickly change and optimize millions of products with one click and without having it done by tech-teams – but rather by non-tech account managers.
We talk to Volker Schmidt, Chief Revenue Officer at Productsup to get an insight into the genesis of the company, the most pressing concerns for e-commerce retailers, and the technologies that will transform the retail landscape in the coming years – including AI, machine learning and VR.
Could you talk a little bit about the early days, and how you’ve built the team along the way?
After having seen couple of hundred data feeds from customers, we learned to understand the internal processes of our clients well. There was a specific experience that occurred with one client, where he needed to drastically change the file he was giving us, but which took him almost four months to compile. To change it, he wrote IT tickets and waited for his IT to implement that change.
That occurrence sparked an idea. We had an internal solution to manage and customize product data without IT staff. Our client would need that as he would be able to change the data directly in the marketing department and provide us with a new file within a couple of hours. We started to develop a graphical user interface and switched from an agency model to an SaaS solution and gave clients access to our own system.
Today we manage and optimize millions of products for different marketing channels. You could say we put the marketing department back in the driver’s seat to manage and maintain a company’s marketing activities, rather than waiting for weeks for the IT department to change their own product output.
Where do you see automation playing a greater role in the coming years?
I personally believe in automation and it should be possible to automate all tasks that have to be completed multiple times, especially as applied to multiple layers or products. Looking at South-East-Asia, companies are starting to understand the importance of technology in day-to-day business. Truth be told, Productsup in South-East Asia (SEA) is currently more about education, rather than selling a product, as most of our discussions are ending with ‘I didn’t know this was possible’. If you look at the world, the push of innovation drives probably initially from the US, then to Europe, before trickling down to SEA a couple of years later.
I think there is also a correlation between the cost of living in a society and the process of automation. The easier it is to staff a large workforce for a lower salary, the easier it is to have them do manual jobs that could be automated, because the automation tool costs more than the salaries of those people.
In what ways, if at all, do you see the AI and Machine Learning revolution touching the e-commerce space?
AI and machine learning will open a whole new world in the e-commerce market.
It’s perhaps also one of the topics where you can find 10 different explanations from 10 different people. I view machine learning as part of AI or, to put it differently, AI would not be possible without machine learning occurring beforehand. While machine learning tries to answer the ‘how’ part of a question, the AI is executing the ‘what’ part of a question. To put it in perspective – a computer can learn to play chess via machine learning from a theoretical perspective, while the AI is executing the game on a move-to-move-level and also factors in things like game theory (what move might be relevant based on the opponent’s move) and decision theory (to make the decision based on a specific move). In the world of commerce there is a lot of potential for machine learning systems – because who wants to categorize a million products manually in an Excel sheet?
How has the e-commerce space in Europe and the US evolved over the past few years, and what lessons can Asia learn from those mature markets?
One thing that is different between the EU/US and Asia is the amount of automation already in place and the degree of analytics in place to manage and maintain marketing activities. While in Europe you think about how to optimize a specific channel in terms of ROI and other KPIs, most companies in Asia deal with the challenge of actually reaching clients and increasing revenue before optimizing based on KPI. Of course numbers also play a part, but the main driving factor is getting the word out and increasing visibility. The number of clients is directly factored by the amount of fast internet available in every SEA country.
Also, Western companies use a higher degree of innovation in their business. Automatic customer segmentation and targeting based on their interest in social media – so called automatic marketing orchestration – is still in early stages in Asian markets. There is no doubt that systems have to adapt to the behavior in the Asian markets as you can’t copy the targeting solutions 1:1.
What technologies on the horizon excite you that could change the way e-commerce merchants work in the future?
As mentioned before, machine learning is a very interesting field of development. Google has reached new heights with image recognition lately; identifying a product based on certain product information without manually looking at it can be quite interesting.
For me, I really look forward to VR in e-commerce. The one thing that is limited in this world is space – space to live in, space to build on. How about a warehouse which is completely empty but has VR-glasses from Nike, Puma and Adidas? Once you wear them, you will see the virtual store from those brands while someone else – who could even be in the same room – sees the other brands’ stores. This way, a certain space can have multiple purposes and can be used economically.
Another interesting topic is the development of automation while driving. At the moment you drive your car mostly by yourself. But consider a self-driving car while you get the latest products from Amazon placed on your front window to shop from. If you think about the term ‘Everywhere Commerce’, you will find different possibilities to reach your customers during various day-to-day activities where they will be able to shop in the future.
What do you think is the future of retail: pure e-commerce, bricks and mortar + e-commerce hybrid model, or some sort of retail showroom that uses technology in significantly new ways?
Physical stores will always have the right to exist. It’s human nature that not everyone thinks and acts in the same way, so there will always be people enjoying shopping in stores, where they can touch/try on the clothes directly. For this reason, I believe we will never see books fall out of favor simply because we have ebooks – certain people will always find joy in touching and reading a physical book.
From a technology perspective, I believe everything that encourages convenience is a strong approach – whether it’s a good recommendation engine or an Amazon Dash Button that allows you to directly buy a new product while picking the last one from your reserve shelf. Something that is convenient will always be chosen over something that has less convenience. Of course, that only applies for the majority of people, as you will always have some people enjoying a special treat by a local winery over a wine package from Amazon.
If you add a localized feature to the mix like the Google Local Product Ads, you have a pretty good idea about future commerce – everywhere, anytime, right when you need it.
Source: Google News