Phase 4 (2016-2017): going East
As we have seen in previous parts of this post, product matching was getting more and more complex, while clients were not ready to pay that much, nor to wait so long for the project to be delivered.
By this time, most vendors of pricing / assortment services have come to conclusion that product matching requires no skilled workforce. Instead, it requires an agent who would diligently go through product specs and compare them across various target websites. No powerful hardware was needed, a PC / laptop able to run a browser and MS Excel was all it took. Of course – Internet connection was required – but by this time, Internet was so well-spread across the globe that it was available practically anywhere.
Training such a agent was also not too demanding – 1 or 2 weeks of training proved to be sufficient. Further, agents working on product matching rarely worked from the office – in most cases they worked from home, on their own equipment. Depending on the country where they worked from, they even worked with no proper labor contracts. All of these factors significantly helped reduce the cost of product matching, especially in countries with lower wages. As we all know, Eastern Europe, Turkey and Latin America were, at that time, large source of such labor – and several companies like Profitero(operations set up in Belarus), Competera(Ukraine), Price2Spy(Serbia) and Net Rivals (Venezuela) took advantage, gaining the foothold in product matching market.
Speaking of the required skills – good command of English language was not even a prerequisite for such an agent – by that time Google Translate service became so good that just basic fluency in English was needed.
And what if the agent was not diligent enough? How do we ensure that client gets no wrong matches (in another post we have already elaborated why the accuracy of product matching was so vital). Well, that’s what the QA was there for. The final result was clear: many of such companies formed teams in countries where the cost of labor was significantly lower than in US / Western Europe. Soon enough, such teams started appearing in Eastern Europe, Balkans and Latin America.
The quality of their work was questionable in the beginning but got better over time. The business still was not that scalable (as large projects required hiring large teams), but overall, the costs of product matching went down, while the quality remained practically the same.
Phase 5 (2018-2019): going further East
In previous phases, many companies have tried establishing product matching teams in countries where the labor comes really cheap – India, Bangladesh, Indonesia, Philippines.
However, such an endeavor was rarely successful – either because of poor education, and lack of managerial workforce, or because of general conditions which were unstable for business.
Never the less, some of these attempts did bring some success, and product matching teams in various parts of Asia started getting traction – enabling companies that established such teams to make substantially higher profits, compared to companies who offered such services with European workforce.
Of course, working with far-East teams came with challenges of their own: language barrier (poor English skills, or English with a strong accent), different work ethics, time zone difference, poor Internet / power supply infrastructure. To overcome these challenges some companies went as far as sending their own managerial workforce to remote countries like Sri Lanka or Malaysia. In most cases Western managers stayed there for up to one year – their primary task was to establish a local office, identify locals who could act as managers, and who would run the office after the Western manager leaves.
For local companies (quite a few Indian companies like Sematics3 and Intelligence Node have entered the price intelligence market by now) these obstacles were not that difficult, and they have rapidly won large product matching market shares. However, no matter how cheap your workforce got and no matter how large the teams grew, scalability and project delivery time remained a problem (especially when Black Friday gets closer – all retailers want to make sure all their products are matched, and that their prices are competitive). Lack of automation was still the main problem with product matching – and something needed to be done about it.